Electronic Tuning

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

James, since I am still a newbie with the device I don't want to project too much on concepts like tuning speed because it always takes a while to get used to new techniques and get literally up to speed on them. Things that might seem a little clumsy at first eventually get into muscle memory and then you are zipping along. But I kind of doubt that all told it will be significantly faster or slower on a grand. And speed is not my goal. To me the question is can I get at least a little better tuning than what I can get without it, and the answer to that is yes. It is not as big of an improvement as going from aural to Verituner, but it is a difference I can hear, and I don't mean just tuning tests. I mean the overall timbral experience of hearing a piece of music played. But again, that difference is mostly at the very top. My guess is a lot of people with good ears would admit to the improvement if it's pointed out, or they could hear an A/B comparison. The cleanness of the top octave especially. 

As to your last Q I am not sure exactly what you mean, but I would say it made the the Verituner readings more steady and locked in, even at the extreme ends of the piano, so overall the PianoSens seemed "easier" for the Verituner to handle than the default mic input mode. Not as jumpy or jittery. Not to mention that extraneous noises in the room mean nothing to PianoSens. This alone would have been an absolute deal maker for me in certain tuning situations (noisy public pianos, tuning with other pianos or other instruments playing nearby, and even other tuners tuning in the same space). 

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

Original Message:
Sent: 3/20/2024 7:08:00 PM
From: Steven Norsworthy
Subject: RE: First tuning experience using PianoSens

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-20-2024 15:15
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 8:03:00 AM
From: Peter Grey
Subject: RE: First tuning experience using PianoSens

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

------------------------------
Peter Grey
Stratham NH
(603) 686-2395
pianodoctor57@gmail.com

Original Message:
Sent: 03-20-2024 02:24
From: Steven Norsworthy
Subject: First tuning experience using PianoSens

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

(Again, non commercial interest, no product or app promoted or represented).

The false beats of the low notes are slow enough to be mostly inaudible but they are smearing up the app's ability to measure them. If the app uses zero crossings, it would produce lots of false crosses at the double peak frequencies that are so close. The app would not have the ability to separate them unless it used super high resolution FFT's. Here is an example of the note G1 on a concert grand. Notice the double peaks are only a couple Hz apart and only a few dB different in magnitude. This is a real problem and the errors are huge on most apps.  A very long FFT is the only practical answer.

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-20-2024 20:39
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 7:08:00 PM
From: Steven Norsworthy
Subject: RE: First tuning experience using PianoSens

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-20-2024 15:15
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 8:03:00 AM
From: Peter Grey
Subject: RE: First tuning experience using PianoSens

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

------------------------------
Peter Grey
Stratham NH
(603) 686-2395
pianodoctor57@gmail.com

Original Message:
Sent: 03-20-2024 02:24
From: Steven Norsworthy
Subject: First tuning experience using PianoSens

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

Except that no ETD is ever going to look at the 1st partial of G1 down at 50 Hz. If it's Cybertuner being discussed, I just pulled up their manual and it looks like for G1 they'd be using something from the 6th to 12th partial. (See Page 19.) Zero crossings should be significantly more accurate in that 300-600 Hz range. 

(Again, non commercial interest, no product or app promoted or represented).

The false beats of the low notes are slow enough to be mostly inaudible but they are smearing up the app's ability to measure them. If the app uses zero crossings, it would produce lots of false crosses at the double peak frequencies that are so close. The app would not have the ability to separate them unless it used super high resolution FFT's. Here is an example of the note G1 on a concert grand. Notice the double peaks are only a couple Hz apart and only a few dB different in magnitude. This is a real problem and the errors are huge on most apps.  A very long FFT is the only practical answer.

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-20-2024 20:39
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 7:08:00 PM
From: Steven Norsworthy
Subject: RE: First tuning experience using PianoSens

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-20-2024 15:15
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 8:03:00 AM
From: Peter Grey
Subject: RE: First tuning experience using PianoSens

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

------------------------------
Peter Grey
Stratham NH
(603) 686-2395
pianodoctor57@gmail.com

Original Message:
Sent: 03-20-2024 02:24
From: Steven Norsworthy
Subject: First tuning experience using PianoSens

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

Hi Fred: 

I have been using the Accutuner since I began tuning in 1988, so I know what you mean about the slower display.  I can't imagine it being any jumpier.  

As far as using Pianoscope, I started using it about a year ago, and I've seen it make improvements along the way, especially the Freeze concept.  I wish all of the ETD's had this feature.  Anyway, the Freeze line helps you make better decisions because it only measures during a short duration of time after the initial blow.  The pitch is more accurately determined that way.  As you know, the pitch is constantly changing as the decay period goes on.  If you measure the pitch during the attack phase, in a limited time frame, you'll get a better reading.  Since PIanoscope (and other ETD's that use multiple partials) require measuring them to calculate the pitch, if they fade away, they can't be used.  So, the measurement needs to be done while the partials are sounding their loudest.  Waiting seconds later, the pitch is falling and the partials are fading.  All of the pitch information needed is in the first second of string vibration.

Here's the thing- this feature would not have been possible to create until recently because there wasn't enough processing power in our IOS devices to run it.  It is entirely a new concept.  Of course, we have always advised to "tune to the decay" except in the treble where we "tune to the attack".  Well, obviously, there is not other option up there.  But creating a "time window" where the pitch measurement is done, the reading is more consistent.  That's why it was invented.  The Freeze line appears after a specified time measured from the initial blow.  You can set this time window from 100 miliiseconds to 999 (1sec).  During this time, an average measurement is made, and then displays the pitch as a grey line.  Play the note again to reset the grey Freeze line.  Be sure to release the key so that the string stops vibrating for a half second so that the program knows you're resetting the Freeze.  Otherwise it gets confused and gives a false reading.  Very important point.

You can use the red line, and the "strobe" in the background, to get close to the pitch, but then use the grey Freeze line to land as close to zero as possible.  If you're in the high treble, this is more tricky especially with false beats and such.  Sometimes I'll watch the strobe in the background and try to get it to slow down and stop, or oscillate back and forth in one place.  I look at all the indicators and decide ultimately with my ears where I want to leave it.

As far as unisons go, if I use the Freeze line, I can tune beatless unisons consistently.  My tunings are much cleaner than before, and I can correct them later by simply isolating the strings and retuning them.  Bad unisons stand out like a sore thumb when the rest are quiet.  If you like smearing unisons, that's a choice too.  There is a reason that all of the ETD's suggest using your ears to tune unisons, and that is because they don't have the necessary consistency to make them perfect.  Of course, you can get them close, but by ear getting them closer than .4 cents is hit and miss.  Adding the Freeze to Pianoscope makes it easier and more accurate, down to .1 cent. Doing this requires a little extra time and finesse, but the result is worth it, IMO.

Using the Freeze feature makes it much easier to tune than watching the red line jumping all over the place.  I use them both while tuning, but the Freeze is where the pitch is most accurate.

I hope you'll give it another shot, and I hope I've explained things a little better.

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

Original Message:
Sent: 3/20/2024 11:52:00 PM
From: Paul McCloud
Subject: RE: First tuning experience using PianoSens

Hi Fred: 

I have been using the Accutuner since I began tuning in 1988, so I know what you mean about the slower display.  I can't imagine it being any jumpier.  

As far as using Pianoscope, I started using it about a year ago, and I've seen it make improvements along the way, especially the Freeze concept.  I wish all of the ETD's had this feature.  Anyway, the Freeze line helps you make better decisions because it only measures during a short duration of time after the initial blow.  The pitch is more accurately determined that way.  As you know, the pitch is constantly changing as the decay period goes on.  If you measure the pitch during the attack phase, in a limited time frame, you'll get a better reading.  Since PIanoscope (and other ETD's that use multiple partials) require measuring them to calculate the pitch, if they fade away, they can't be used.  So, the measurement needs to be done while the partials are sounding their loudest.  Waiting seconds later, the pitch is falling and the partials are fading.  All of the pitch information needed is in the first second of string vibration.

Here's the thing- this feature would not have been possible to create until recently because there wasn't enough processing power in our IOS devices to run it.  It is entirely a new concept.  Of course, we have always advised to "tune to the decay" except in the treble where we "tune to the attack".  Well, obviously, there is not other option up there.  But creating a "time window" where the pitch measurement is done, the reading is more consistent.  That's why it was invented.  The Freeze line appears after a specified time measured from the initial blow.  You can set this time window from 100 miliiseconds to 999 (1sec).  During this time, an average measurement is made, and then displays the pitch as a grey line.  Play the note again to reset the grey Freeze line.  Be sure to release the key so that the string stops vibrating for a half second so that the program knows you're resetting the Freeze.  Otherwise it gets confused and gives a false reading.  Very important point.

You can use the red line, and the "strobe" in the background, to get close to the pitch, but then use the grey Freeze line to land as close to zero as possible.  If you're in the high treble, this is more tricky especially with false beats and such.  Sometimes I'll watch the strobe in the background and try to get it to slow down and stop, or oscillate back and forth in one place.  I look at all the indicators and decide ultimately with my ears where I want to leave it.

As far as unisons go, if I use the Freeze line, I can tune beatless unisons consistently.  My tunings are much cleaner than before, and I can correct them later by simply isolating the strings and retuning them.  Bad unisons stand out like a sore thumb when the rest are quiet.  If you like smearing unisons, that's a choice too.  There is a reason that all of the ETD's suggest using your ears to tune unisons, and that is because they don't have the necessary consistency to make them perfect.  Of course, you can get them close, but by ear getting them closer than .4 cents is hit and miss.  Adding the Freeze to Pianoscope makes it easier and more accurate, down to .1 cent. Doing this requires a little extra time and finesse, but the result is worth it, IMO.

Using the Freeze feature makes it much easier to tune than watching the red line jumping all over the place.  I use them both while tuning, but the Freeze is where the pitch is most accurate.

I hope you'll give it another shot, and I hope I've explained things a little better.

------------------------------
Paul McCloud, RPT
Accutone Piano Service
www.AccutonePianoService.com
pavadasa@gmail.com

Original Message:
Sent: 03-20-2024 20:39
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 7:08:00 PM
From: Steven Norsworthy
Subject: RE: First tuning experience using PianoSens

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-20-2024 15:15
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 8:03:00 AM
From: Peter Grey
Subject: RE: First tuning experience using PianoSens

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

------------------------------
Peter Grey
Stratham NH
(603) 686-2395
pianodoctor57@gmail.com

Original Message:
Sent: 03-20-2024 02:24
From: Steven Norsworthy
Subject: First tuning experience using PianoSens

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

No one out there seems to understand the Freeze frame concept. It's up to the inventor to explain.

No, the Freeze is not a snapshot, it is window of time from the hammer strike to about 700 milliseconds thereafter, where that 'average frequency' is measured over that duration of time window where the signal power is highest. There is no such thing as an instantaneous frequency measurement in zero time.

The real 'driving force' for the invention of the Freeze frame window is that the 'pitch' also decays as the amplitude decays. Can any human watch a continuous reading and remember were the pitch was consistently? No. The Freeze indicates that, and it is the ONLY consistent way that has ever been described or reported.

The piano decays in amplitude about 20 dB / sec in the mid range, and about 20 dB in 300 milliseconds in the upper register. Tune where the power is the highest, in the first second. The piano not only decays quickly but the human absolutely cannot perceive the pitch accurately in the first sec, nor can most any ETD, 'unless' it integrates it over that time using an FFT. 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101
------------------------------

Original Message:
Sent: 03-21-2024 14:41
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 11:52:00 PM
From: Paul McCloud
Subject: RE: First tuning experience using PianoSens

Hi Fred: 

I have been using the Accutuner since I began tuning in 1988, so I know what you mean about the slower display.  I can't imagine it being any jumpier.  

As far as using Pianoscope, I started using it about a year ago, and I've seen it make improvements along the way, especially the Freeze concept.  I wish all of the ETD's had this feature.  Anyway, the Freeze line helps you make better decisions because it only measures during a short duration of time after the initial blow.  The pitch is more accurately determined that way.  As you know, the pitch is constantly changing as the decay period goes on.  If you measure the pitch during the attack phase, in a limited time frame, you'll get a better reading.  Since PIanoscope (and other ETD's that use multiple partials) require measuring them to calculate the pitch, if they fade away, they can't be used.  So, the measurement needs to be done while the partials are sounding their loudest.  Waiting seconds later, the pitch is falling and the partials are fading.  All of the pitch information needed is in the first second of string vibration.

Here's the thing- this feature would not have been possible to create until recently because there wasn't enough processing power in our IOS devices to run it.  It is entirely a new concept.  Of course, we have always advised to "tune to the decay" except in the treble where we "tune to the attack".  Well, obviously, there is not other option up there.  But creating a "time window" where the pitch measurement is done, the reading is more consistent.  That's why it was invented.  The Freeze line appears after a specified time measured from the initial blow.  You can set this time window from 100 miliiseconds to 999 (1sec).  During this time, an average measurement is made, and then displays the pitch as a grey line.  Play the note again to reset the grey Freeze line.  Be sure to release the key so that the string stops vibrating for a half second so that the program knows you're resetting the Freeze.  Otherwise it gets confused and gives a false reading.  Very important point.

You can use the red line, and the "strobe" in the background, to get close to the pitch, but then use the grey Freeze line to land as close to zero as possible.  If you're in the high treble, this is more tricky especially with false beats and such.  Sometimes I'll watch the strobe in the background and try to get it to slow down and stop, or oscillate back and forth in one place.  I look at all the indicators and decide ultimately with my ears where I want to leave it.

As far as unisons go, if I use the Freeze line, I can tune beatless unisons consistently.  My tunings are much cleaner than before, and I can correct them later by simply isolating the strings and retuning them.  Bad unisons stand out like a sore thumb when the rest are quiet.  If you like smearing unisons, that's a choice too.  There is a reason that all of the ETD's suggest using your ears to tune unisons, and that is because they don't have the necessary consistency to make them perfect.  Of course, you can get them close, but by ear getting them closer than .4 cents is hit and miss.  Adding the Freeze to Pianoscope makes it easier and more accurate, down to .1 cent. Doing this requires a little extra time and finesse, but the result is worth it, IMO.

Using the Freeze feature makes it much easier to tune than watching the red line jumping all over the place.  I use them both while tuning, but the Freeze is where the pitch is most accurate.

I hope you'll give it another shot, and I hope I've explained things a little better.

------------------------------
Paul McCloud, RPT
Accutone Piano Service
www.AccutonePianoService.com
pavadasa@gmail.com

Original Message:
Sent: 03-20-2024 20:39
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 7:08:00 PM
From: Steven Norsworthy
Subject: RE: First tuning experience using PianoSens

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-20-2024 15:15
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 8:03:00 AM
From: Peter Grey
Subject: RE: First tuning experience using PianoSens

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

------------------------------
Peter Grey
Stratham NH
(603) 686-2395
pianodoctor57@gmail.com

Original Message:
Sent: 03-20-2024 02:24
From: Steven Norsworthy
Subject: First tuning experience using PianoSens

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

No one out there seems to understand the Freeze frame concept. It's up to the inventor to explain.

No, the Freeze is not a snapshot, it is window of time from the hammer strike to about 700 milliseconds thereafter, where that 'average frequency' is measured over that duration of time window where the signal power is highest. There is no such thing as an instantaneous frequency measurement in zero time.

The real 'driving force' for the invention of the Freeze frame window is that the 'pitch' also decays as the amplitude decays. Can any human watch a continuous reading and remember were the pitch was consistently? No. The Freeze indicates that, and it is the ONLY consistent way that has ever been described or reported.

The piano decays in amplitude about 20 dB / sec in the mid range, and about 20 dB in 300 milliseconds in the upper register. Tune where the power is the highest, in the first second. The piano not only decays quickly but the human absolutely cannot perceive the pitch accurately in the first sec, nor can most any ETD, 'unless' it integrates it over that time using an FFT. 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-21-2024 14:41
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 11:52:00 PM
From: Paul McCloud
Subject: RE: First tuning experience using PianoSens

Hi Fred: 

I have been using the Accutuner since I began tuning in 1988, so I know what you mean about the slower display.  I can't imagine it being any jumpier.  

As far as using Pianoscope, I started using it about a year ago, and I've seen it make improvements along the way, especially the Freeze concept.  I wish all of the ETD's had this feature.  Anyway, the Freeze line helps you make better decisions because it only measures during a short duration of time after the initial blow.  The pitch is more accurately determined that way.  As you know, the pitch is constantly changing as the decay period goes on.  If you measure the pitch during the attack phase, in a limited time frame, you'll get a better reading.  Since PIanoscope (and other ETD's that use multiple partials) require measuring them to calculate the pitch, if they fade away, they can't be used.  So, the measurement needs to be done while the partials are sounding their loudest.  Waiting seconds later, the pitch is falling and the partials are fading.  All of the pitch information needed is in the first second of string vibration.

Here's the thing- this feature would not have been possible to create until recently because there wasn't enough processing power in our IOS devices to run it.  It is entirely a new concept.  Of course, we have always advised to "tune to the decay" except in the treble where we "tune to the attack".  Well, obviously, there is not other option up there.  But creating a "time window" where the pitch measurement is done, the reading is more consistent.  That's why it was invented.  The Freeze line appears after a specified time measured from the initial blow.  You can set this time window from 100 miliiseconds to 999 (1sec).  During this time, an average measurement is made, and then displays the pitch as a grey line.  Play the note again to reset the grey Freeze line.  Be sure to release the key so that the string stops vibrating for a half second so that the program knows you're resetting the Freeze.  Otherwise it gets confused and gives a false reading.  Very important point.

You can use the red line, and the "strobe" in the background, to get close to the pitch, but then use the grey Freeze line to land as close to zero as possible.  If you're in the high treble, this is more tricky especially with false beats and such.  Sometimes I'll watch the strobe in the background and try to get it to slow down and stop, or oscillate back and forth in one place.  I look at all the indicators and decide ultimately with my ears where I want to leave it.

As far as unisons go, if I use the Freeze line, I can tune beatless unisons consistently.  My tunings are much cleaner than before, and I can correct them later by simply isolating the strings and retuning them.  Bad unisons stand out like a sore thumb when the rest are quiet.  If you like smearing unisons, that's a choice too.  There is a reason that all of the ETD's suggest using your ears to tune unisons, and that is because they don't have the necessary consistency to make them perfect.  Of course, you can get them close, but by ear getting them closer than .4 cents is hit and miss.  Adding the Freeze to Pianoscope makes it easier and more accurate, down to .1 cent. Doing this requires a little extra time and finesse, but the result is worth it, IMO.

Using the Freeze feature makes it much easier to tune than watching the red line jumping all over the place.  I use them both while tuning, but the Freeze is where the pitch is most accurate.

I hope you'll give it another shot, and I hope I've explained things a little better.

------------------------------
Paul McCloud, RPT
Accutone Piano Service
www.AccutonePianoService.com
pavadasa@gmail.com

Original Message:
Sent: 03-20-2024 20:39
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 7:08:00 PM
From: Steven Norsworthy
Subject: RE: First tuning experience using PianoSens

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-20-2024 15:15
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 8:03:00 AM
From: Peter Grey
Subject: RE: First tuning experience using PianoSens

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

------------------------------
Peter Grey
Stratham NH
(603) 686-2395
pianodoctor57@gmail.com

Original Message:
Sent: 03-20-2024 02:24
From: Steven Norsworthy
Subject: First tuning experience using PianoSens

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

About the freeze line. There's a setting to change it. You can specify how many milliseconds it will average to set the freeze, and that number can be set for C0, C1, each C up to C8. On my iPad Pianoscope app, C0 is currently set to 500 ms, C1 to 400 ms, C2 through C8 all 300 ms. Nothing is set to 700. (I have not changed these settings.)
What difference would it make to change the milliseconds? I don't know. 
Test blows produce a sharper freeze line than normal tuning blows, so whom do you believe? Its erratic behavior makes the freeze line a nuisance distraction from the ear. 
Personally, at the moment, I ignore the freeze line. Instead I stop the strobe blocks. When the blocks lie still, I'm good. One of the pleasures of this is that I can gaze at something pleasant and keep the strobe blocks in my peripheral vision. Try it - it's profoundly different from focusing on a small screen, and promotes the practice of listening with your whole body. 
My pianosens is in the mail. I'll be tuning my own Kohler & Campbell studio in a few days, if I can locate the piano under the books and papers and boxes and bags.

No one out there seems to understand the Freeze frame concept. It's up to the inventor to explain.

No, the Freeze is not a snapshot, it is window of time from the hammer strike to about 700 milliseconds thereafter, where that 'average frequency' is measured over that duration of time window where the signal power is highest. There is no such thing as an instantaneous frequency measurement in zero time.

The real 'driving force' for the invention of the Freeze frame window is that the 'pitch' also decays as the amplitude decays. Can any human watch a continuous reading and remember were the pitch was consistently? No. The Freeze indicates that, and it is the ONLY consistent way that has ever been described or reported.

The piano decays in amplitude about 20 dB / sec in the mid range, and about 20 dB in 300 milliseconds in the upper register. Tune where the power is the highest, in the first second. The piano not only decays quickly but the human absolutely cannot perceive the pitch accurately in the first sec, nor can most any ETD, 'unless' it integrates it over that time using an FFT. 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-21-2024 14:41
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 11:52:00 PM
From: Paul McCloud
Subject: RE: First tuning experience using PianoSens

Hi Fred: 

I have been using the Accutuner since I began tuning in 1988, so I know what you mean about the slower display.  I can't imagine it being any jumpier.  

As far as using Pianoscope, I started using it about a year ago, and I've seen it make improvements along the way, especially the Freeze concept.  I wish all of the ETD's had this feature.  Anyway, the Freeze line helps you make better decisions because it only measures during a short duration of time after the initial blow.  The pitch is more accurately determined that way.  As you know, the pitch is constantly changing as the decay period goes on.  If you measure the pitch during the attack phase, in a limited time frame, you'll get a better reading.  Since PIanoscope (and other ETD's that use multiple partials) require measuring them to calculate the pitch, if they fade away, they can't be used.  So, the measurement needs to be done while the partials are sounding their loudest.  Waiting seconds later, the pitch is falling and the partials are fading.  All of the pitch information needed is in the first second of string vibration.

Here's the thing- this feature would not have been possible to create until recently because there wasn't enough processing power in our IOS devices to run it.  It is entirely a new concept.  Of course, we have always advised to "tune to the decay" except in the treble where we "tune to the attack".  Well, obviously, there is not other option up there.  But creating a "time window" where the pitch measurement is done, the reading is more consistent.  That's why it was invented.  The Freeze line appears after a specified time measured from the initial blow.  You can set this time window from 100 miliiseconds to 999 (1sec).  During this time, an average measurement is made, and then displays the pitch as a grey line.  Play the note again to reset the grey Freeze line.  Be sure to release the key so that the string stops vibrating for a half second so that the program knows you're resetting the Freeze.  Otherwise it gets confused and gives a false reading.  Very important point.

You can use the red line, and the "strobe" in the background, to get close to the pitch, but then use the grey Freeze line to land as close to zero as possible.  If you're in the high treble, this is more tricky especially with false beats and such.  Sometimes I'll watch the strobe in the background and try to get it to slow down and stop, or oscillate back and forth in one place.  I look at all the indicators and decide ultimately with my ears where I want to leave it.

As far as unisons go, if I use the Freeze line, I can tune beatless unisons consistently.  My tunings are much cleaner than before, and I can correct them later by simply isolating the strings and retuning them.  Bad unisons stand out like a sore thumb when the rest are quiet.  If you like smearing unisons, that's a choice too.  There is a reason that all of the ETD's suggest using your ears to tune unisons, and that is because they don't have the necessary consistency to make them perfect.  Of course, you can get them close, but by ear getting them closer than .4 cents is hit and miss.  Adding the Freeze to Pianoscope makes it easier and more accurate, down to .1 cent. Doing this requires a little extra time and finesse, but the result is worth it, IMO.

Using the Freeze feature makes it much easier to tune than watching the red line jumping all over the place.  I use them both while tuning, but the Freeze is where the pitch is most accurate.

I hope you'll give it another shot, and I hope I've explained things a little better.

------------------------------
Paul McCloud, RPT
Accutone Piano Service
www.AccutonePianoService.com
pavadasa@gmail.com

Original Message:
Sent: 03-20-2024 20:39
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 7:08:00 PM
From: Steven Norsworthy
Subject: RE: First tuning experience using PianoSens

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-20-2024 15:15
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 8:03:00 AM
From: Peter Grey
Subject: RE: First tuning experience using PianoSens

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

------------------------------
Peter Grey
Stratham NH
(603) 686-2395
pianodoctor57@gmail.com

Original Message:
Sent: 03-20-2024 02:24
From: Steven Norsworthy
Subject: First tuning experience using PianoSens

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

Fred and Jason and Paul,

The 'blows' should be 'medium' and not 'hard' for getting the freeze line to the right number. I 'invented' the 'concept', disclosed it to Frank at Pianoscope,  and Frank 'implemented' and did a good job, kudos to Frank. As Frank and I were 'experimenting' with it during the beta phase, we indeed recommended default settings faster for higher notes and slower for lower notes. That works well, but I like putting all the settings at 700ms because where there are false beats, there is more time to resolve them at the slower setting of 700ms. 

Again, 'meduim' blows. 

I had 5 very experienced pro tuners at my house yesterday for a demonstration. I got the unisons at zero cents just and they all told me that they have 'never' heard such a perfect unison. They all had collectively experience with all the commercial tools and apps.

I can take a call at 619-964-0101.

Steve

About the freeze line. There's a setting to change it. You can specify how many milliseconds it will average to set the freeze, and that number can be set for C0, C1, each C up to C8. On my iPad Pianoscope app, C0 is currently set to 500 ms, C1 to 400 ms, C2 through C8 all 300 ms. Nothing is set to 700. (I have not changed these settings.)
What difference would it make to change the milliseconds? I don't know. 
Test blows produce a sharper freeze line than normal tuning blows, so whom do you believe? Its erratic behavior makes the freeze line a nuisance distraction from the ear. 
Personally, at the moment, I ignore the freeze line. Instead I stop the strobe blocks. When the blocks lie still, I'm good. One of the pleasures of this is that I can gaze at something pleasant and keep the strobe blocks in my peripheral vision. Try it - it's profoundly different from focusing on a small screen, and promotes the practice of listening with your whole body. 
My pianosens is in the mail. I'll be tuning my own Kohler & Campbell studio in a few days, if I can locate the piano under the books and papers and boxes and bags.

No one out there seems to understand the Freeze frame concept. It's up to the inventor to explain.

No, the Freeze is not a snapshot, it is window of time from the hammer strike to about 700 milliseconds thereafter, where that 'average frequency' is measured over that duration of time window where the signal power is highest. There is no such thing as an instantaneous frequency measurement in zero time.

The real 'driving force' for the invention of the Freeze frame window is that the 'pitch' also decays as the amplitude decays. Can any human watch a continuous reading and remember were the pitch was consistently? No. The Freeze indicates that, and it is the ONLY consistent way that has ever been described or reported.

The piano decays in amplitude about 20 dB / sec in the mid range, and about 20 dB in 300 milliseconds in the upper register. Tune where the power is the highest, in the first second. The piano not only decays quickly but the human absolutely cannot perceive the pitch accurately in the first sec, nor can most any ETD, 'unless' it integrates it over that time using an FFT. 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-21-2024 14:41
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 11:52:00 PM
From: Paul McCloud
Subject: RE: First tuning experience using PianoSens

Hi Fred: 

I have been using the Accutuner since I began tuning in 1988, so I know what you mean about the slower display.  I can't imagine it being any jumpier.  

As far as using Pianoscope, I started using it about a year ago, and I've seen it make improvements along the way, especially the Freeze concept.  I wish all of the ETD's had this feature.  Anyway, the Freeze line helps you make better decisions because it only measures during a short duration of time after the initial blow.  The pitch is more accurately determined that way.  As you know, the pitch is constantly changing as the decay period goes on.  If you measure the pitch during the attack phase, in a limited time frame, you'll get a better reading.  Since PIanoscope (and other ETD's that use multiple partials) require measuring them to calculate the pitch, if they fade away, they can't be used.  So, the measurement needs to be done while the partials are sounding their loudest.  Waiting seconds later, the pitch is falling and the partials are fading.  All of the pitch information needed is in the first second of string vibration.

Here's the thing- this feature would not have been possible to create until recently because there wasn't enough processing power in our IOS devices to run it.  It is entirely a new concept.  Of course, we have always advised to "tune to the decay" except in the treble where we "tune to the attack".  Well, obviously, there is not other option up there.  But creating a "time window" where the pitch measurement is done, the reading is more consistent.  That's why it was invented.  The Freeze line appears after a specified time measured from the initial blow.  You can set this time window from 100 miliiseconds to 999 (1sec).  During this time, an average measurement is made, and then displays the pitch as a grey line.  Play the note again to reset the grey Freeze line.  Be sure to release the key so that the string stops vibrating for a half second so that the program knows you're resetting the Freeze.  Otherwise it gets confused and gives a false reading.  Very important point.

You can use the red line, and the "strobe" in the background, to get close to the pitch, but then use the grey Freeze line to land as close to zero as possible.  If you're in the high treble, this is more tricky especially with false beats and such.  Sometimes I'll watch the strobe in the background and try to get it to slow down and stop, or oscillate back and forth in one place.  I look at all the indicators and decide ultimately with my ears where I want to leave it.

As far as unisons go, if I use the Freeze line, I can tune beatless unisons consistently.  My tunings are much cleaner than before, and I can correct them later by simply isolating the strings and retuning them.  Bad unisons stand out like a sore thumb when the rest are quiet.  If you like smearing unisons, that's a choice too.  There is a reason that all of the ETD's suggest using your ears to tune unisons, and that is because they don't have the necessary consistency to make them perfect.  Of course, you can get them close, but by ear getting them closer than .4 cents is hit and miss.  Adding the Freeze to Pianoscope makes it easier and more accurate, down to .1 cent. Doing this requires a little extra time and finesse, but the result is worth it, IMO.

Using the Freeze feature makes it much easier to tune than watching the red line jumping all over the place.  I use them both while tuning, but the Freeze is where the pitch is most accurate.

I hope you'll give it another shot, and I hope I've explained things a little better.

------------------------------
Paul McCloud, RPT
Accutone Piano Service
www.AccutonePianoService.com
pavadasa@gmail.com

Original Message:
Sent: 03-20-2024 20:39
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 7:08:00 PM
From: Steven Norsworthy
Subject: RE: First tuning experience using PianoSens

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-20-2024 15:15
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 8:03:00 AM
From: Peter Grey
Subject: RE: First tuning experience using PianoSens

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

------------------------------
Peter Grey
Stratham NH
(603) 686-2395
pianodoctor57@gmail.com

Original Message:
Sent: 03-20-2024 02:24
From: Steven Norsworthy
Subject: First tuning experience using PianoSens

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

Fred and Jason and Paul,

The 'blows' should be 'medium' and not 'hard' for getting the freeze line to the right number. I 'invented' the 'concept', disclosed it to Frank at Pianoscope,  and Frank 'implemented' and did a good job, kudos to Frank. As Frank and I were 'experimenting' with it during the beta phase, we indeed recommended default settings faster for higher notes and slower for lower notes. That works well, but I like putting all the settings at 700ms because where there are false beats, there is more time to resolve them at the slower setting of 700ms. 

Again, 'meduim' blows. 

I had 5 very experienced pro tuners at my house yesterday for a demonstration. I got the unisons at zero cents just and they all told me that they have 'never' heard such a perfect unison. They all had collectively experience with all the commercial tools and apps.

I can take a call at 619-964-0101.

Steve

About the freeze line. There's a setting to change it. You can specify how many milliseconds it will average to set the freeze, and that number can be set for C0, C1, each C up to C8. On my iPad Pianoscope app, C0 is currently set to 500 ms, C1 to 400 ms, C2 through C8 all 300 ms. Nothing is set to 700. (I have not changed these settings.)
What difference would it make to change the milliseconds? I don't know. 
Test blows produce a sharper freeze line than normal tuning blows, so whom do you believe? Its erratic behavior makes the freeze line a nuisance distraction from the ear. 
Personally, at the moment, I ignore the freeze line. Instead I stop the strobe blocks. When the blocks lie still, I'm good. One of the pleasures of this is that I can gaze at something pleasant and keep the strobe blocks in my peripheral vision. Try it - it's profoundly different from focusing on a small screen, and promotes the practice of listening with your whole body. 
My pianosens is in the mail. I'll be tuning my own Kohler & Campbell studio in a few days, if I can locate the piano under the books and papers and boxes and bags.

No one out there seems to understand the Freeze frame concept. It's up to the inventor to explain.

No, the Freeze is not a snapshot, it is window of time from the hammer strike to about 700 milliseconds thereafter, where that 'average frequency' is measured over that duration of time window where the signal power is highest. There is no such thing as an instantaneous frequency measurement in zero time.

The real 'driving force' for the invention of the Freeze frame window is that the 'pitch' also decays as the amplitude decays. Can any human watch a continuous reading and remember were the pitch was consistently? No. The Freeze indicates that, and it is the ONLY consistent way that has ever been described or reported.

The piano decays in amplitude about 20 dB / sec in the mid range, and about 20 dB in 300 milliseconds in the upper register. Tune where the power is the highest, in the first second. The piano not only decays quickly but the human absolutely cannot perceive the pitch accurately in the first sec, nor can most any ETD, 'unless' it integrates it over that time using an FFT. 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-21-2024 14:41
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 11:52:00 PM
From: Paul McCloud
Subject: RE: First tuning experience using PianoSens

Hi Fred: 

I have been using the Accutuner since I began tuning in 1988, so I know what you mean about the slower display.  I can't imagine it being any jumpier.  

As far as using Pianoscope, I started using it about a year ago, and I've seen it make improvements along the way, especially the Freeze concept.  I wish all of the ETD's had this feature.  Anyway, the Freeze line helps you make better decisions because it only measures during a short duration of time after the initial blow.  The pitch is more accurately determined that way.  As you know, the pitch is constantly changing as the decay period goes on.  If you measure the pitch during the attack phase, in a limited time frame, you'll get a better reading.  Since PIanoscope (and other ETD's that use multiple partials) require measuring them to calculate the pitch, if they fade away, they can't be used.  So, the measurement needs to be done while the partials are sounding their loudest.  Waiting seconds later, the pitch is falling and the partials are fading.  All of the pitch information needed is in the first second of string vibration.

Here's the thing- this feature would not have been possible to create until recently because there wasn't enough processing power in our IOS devices to run it.  It is entirely a new concept.  Of course, we have always advised to "tune to the decay" except in the treble where we "tune to the attack".  Well, obviously, there is not other option up there.  But creating a "time window" where the pitch measurement is done, the reading is more consistent.  That's why it was invented.  The Freeze line appears after a specified time measured from the initial blow.  You can set this time window from 100 miliiseconds to 999 (1sec).  During this time, an average measurement is made, and then displays the pitch as a grey line.  Play the note again to reset the grey Freeze line.  Be sure to release the key so that the string stops vibrating for a half second so that the program knows you're resetting the Freeze.  Otherwise it gets confused and gives a false reading.  Very important point.

You can use the red line, and the "strobe" in the background, to get close to the pitch, but then use the grey Freeze line to land as close to zero as possible.  If you're in the high treble, this is more tricky especially with false beats and such.  Sometimes I'll watch the strobe in the background and try to get it to slow down and stop, or oscillate back and forth in one place.  I look at all the indicators and decide ultimately with my ears where I want to leave it.

As far as unisons go, if I use the Freeze line, I can tune beatless unisons consistently.  My tunings are much cleaner than before, and I can correct them later by simply isolating the strings and retuning them.  Bad unisons stand out like a sore thumb when the rest are quiet.  If you like smearing unisons, that's a choice too.  There is a reason that all of the ETD's suggest using your ears to tune unisons, and that is because they don't have the necessary consistency to make them perfect.  Of course, you can get them close, but by ear getting them closer than .4 cents is hit and miss.  Adding the Freeze to Pianoscope makes it easier and more accurate, down to .1 cent. Doing this requires a little extra time and finesse, but the result is worth it, IMO.

Using the Freeze feature makes it much easier to tune than watching the red line jumping all over the place.  I use them both while tuning, but the Freeze is where the pitch is most accurate.

I hope you'll give it another shot, and I hope I've explained things a little better.

------------------------------
Paul McCloud, RPT
Accutone Piano Service
www.AccutonePianoService.com
pavadasa@gmail.com

Original Message:
Sent: 03-20-2024 20:39
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 7:08:00 PM
From: Steven Norsworthy
Subject: RE: First tuning experience using PianoSens

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-20-2024 15:15
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 8:03:00 AM
From: Peter Grey
Subject: RE: First tuning experience using PianoSens

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

------------------------------
Peter Grey
Stratham NH
(603) 686-2395
pianodoctor57@gmail.com

Original Message:
Sent: 03-20-2024 02:24
From: Steven Norsworthy
Subject: First tuning experience using PianoSens

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

Fred and Jason and Paul,

The 'blows' should be 'medium' and not 'hard' for getting the freeze line to the right number. I 'invented' the 'concept', disclosed it to Frank at Pianoscope,  and Frank 'implemented' and did a good job, kudos to Frank. As Frank and I were 'experimenting' with it during the beta phase, we indeed recommended default settings faster for higher notes and slower for lower notes. That works well, but I like putting all the settings at 700ms because where there are false beats, there is more time to resolve them at the slower setting of 700ms. 

Again, 'meduim' blows. 

I had 5 very experienced pro tuners at my house yesterday for a demonstration. I got the unisons at zero cents just and they all told me that they have 'never' heard such a perfect unison. They all had collectively experience with all the commercial tools and apps.

I can take a call at 619-964-0101.

Steve

About the freeze line. There's a setting to change it. You can specify how many milliseconds it will average to set the freeze, and that number can be set for C0, C1, each C up to C8. On my iPad Pianoscope app, C0 is currently set to 500 ms, C1 to 400 ms, C2 through C8 all 300 ms. Nothing is set to 700. (I have not changed these settings.)
What difference would it make to change the milliseconds? I don't know. 
Test blows produce a sharper freeze line than normal tuning blows, so whom do you believe? Its erratic behavior makes the freeze line a nuisance distraction from the ear. 
Personally, at the moment, I ignore the freeze line. Instead I stop the strobe blocks. When the blocks lie still, I'm good. One of the pleasures of this is that I can gaze at something pleasant and keep the strobe blocks in my peripheral vision. Try it - it's profoundly different from focusing on a small screen, and promotes the practice of listening with your whole body. 
My pianosens is in the mail. I'll be tuning my own Kohler & Campbell studio in a few days, if I can locate the piano under the books and papers and boxes and bags.

No one out there seems to understand the Freeze frame concept. It's up to the inventor to explain.

No, the Freeze is not a snapshot, it is window of time from the hammer strike to about 700 milliseconds thereafter, where that 'average frequency' is measured over that duration of time window where the signal power is highest. There is no such thing as an instantaneous frequency measurement in zero time.

The real 'driving force' for the invention of the Freeze frame window is that the 'pitch' also decays as the amplitude decays. Can any human watch a continuous reading and remember were the pitch was consistently? No. The Freeze indicates that, and it is the ONLY consistent way that has ever been described or reported.

The piano decays in amplitude about 20 dB / sec in the mid range, and about 20 dB in 300 milliseconds in the upper register. Tune where the power is the highest, in the first second. The piano not only decays quickly but the human absolutely cannot perceive the pitch accurately in the first sec, nor can most any ETD, 'unless' it integrates it over that time using an FFT. 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-21-2024 14:41
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 11:52:00 PM
From: Paul McCloud
Subject: RE: First tuning experience using PianoSens

Hi Fred: 

I have been using the Accutuner since I began tuning in 1988, so I know what you mean about the slower display.  I can't imagine it being any jumpier.  

As far as using Pianoscope, I started using it about a year ago, and I've seen it make improvements along the way, especially the Freeze concept.  I wish all of the ETD's had this feature.  Anyway, the Freeze line helps you make better decisions because it only measures during a short duration of time after the initial blow.  The pitch is more accurately determined that way.  As you know, the pitch is constantly changing as the decay period goes on.  If you measure the pitch during the attack phase, in a limited time frame, you'll get a better reading.  Since PIanoscope (and other ETD's that use multiple partials) require measuring them to calculate the pitch, if they fade away, they can't be used.  So, the measurement needs to be done while the partials are sounding their loudest.  Waiting seconds later, the pitch is falling and the partials are fading.  All of the pitch information needed is in the first second of string vibration.

Here's the thing- this feature would not have been possible to create until recently because there wasn't enough processing power in our IOS devices to run it.  It is entirely a new concept.  Of course, we have always advised to "tune to the decay" except in the treble where we "tune to the attack".  Well, obviously, there is not other option up there.  But creating a "time window" where the pitch measurement is done, the reading is more consistent.  That's why it was invented.  The Freeze line appears after a specified time measured from the initial blow.  You can set this time window from 100 miliiseconds to 999 (1sec).  During this time, an average measurement is made, and then displays the pitch as a grey line.  Play the note again to reset the grey Freeze line.  Be sure to release the key so that the string stops vibrating for a half second so that the program knows you're resetting the Freeze.  Otherwise it gets confused and gives a false reading.  Very important point.

You can use the red line, and the "strobe" in the background, to get close to the pitch, but then use the grey Freeze line to land as close to zero as possible.  If you're in the high treble, this is more tricky especially with false beats and such.  Sometimes I'll watch the strobe in the background and try to get it to slow down and stop, or oscillate back and forth in one place.  I look at all the indicators and decide ultimately with my ears where I want to leave it.

As far as unisons go, if I use the Freeze line, I can tune beatless unisons consistently.  My tunings are much cleaner than before, and I can correct them later by simply isolating the strings and retuning them.  Bad unisons stand out like a sore thumb when the rest are quiet.  If you like smearing unisons, that's a choice too.  There is a reason that all of the ETD's suggest using your ears to tune unisons, and that is because they don't have the necessary consistency to make them perfect.  Of course, you can get them close, but by ear getting them closer than .4 cents is hit and miss.  Adding the Freeze to Pianoscope makes it easier and more accurate, down to .1 cent. Doing this requires a little extra time and finesse, but the result is worth it, IMO.

Using the Freeze feature makes it much easier to tune than watching the red line jumping all over the place.  I use them both while tuning, but the Freeze is where the pitch is most accurate.

I hope you'll give it another shot, and I hope I've explained things a little better.

------------------------------
Paul McCloud, RPT
Accutone Piano Service
www.AccutonePianoService.com
pavadasa@gmail.com

Original Message:
Sent: 03-20-2024 20:39
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 7:08:00 PM
From: Steven Norsworthy
Subject: RE: First tuning experience using PianoSens

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-20-2024 15:15
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 8:03:00 AM
From: Peter Grey
Subject: RE: First tuning experience using PianoSens

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

------------------------------
Peter Grey
Stratham NH
(603) 686-2395
pianodoctor57@gmail.com

Original Message:
Sent: 03-20-2024 02:24
From: Steven Norsworthy
Subject: First tuning experience using PianoSens

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

Any blow force below a certain level will 'not' deviate the freeze line biow to blow. That is what Paul and I and others see. You should see the same. It has to do with the nature of the nonlinearity of spring physics. The hard blows are not the way to tune.

Steve

Fred and Jason and Paul,

The 'blows' should be 'medium' and not 'hard' for getting the freeze line to the right number. I 'invented' the 'concept', disclosed it to Frank at Pianoscope,  and Frank 'implemented' and did a good job, kudos to Frank. As Frank and I were 'experimenting' with it during the beta phase, we indeed recommended default settings faster for higher notes and slower for lower notes. That works well, but I like putting all the settings at 700ms because where there are false beats, there is more time to resolve them at the slower setting of 700ms. 

Again, 'meduim' blows. 

I had 5 very experienced pro tuners at my house yesterday for a demonstration. I got the unisons at zero cents just and they all told me that they have 'never' heard such a perfect unison. They all had collectively experience with all the commercial tools and apps.

I can take a call at 619-964-0101.

Steve

About the freeze line. There's a setting to change it. You can specify how many milliseconds it will average to set the freeze, and that number can be set for C0, C1, each C up to C8. On my iPad Pianoscope app, C0 is currently set to 500 ms, C1 to 400 ms, C2 through C8 all 300 ms. Nothing is set to 700. (I have not changed these settings.)
What difference would it make to change the milliseconds? I don't know. 
Test blows produce a sharper freeze line than normal tuning blows, so whom do you believe? Its erratic behavior makes the freeze line a nuisance distraction from the ear. 
Personally, at the moment, I ignore the freeze line. Instead I stop the strobe blocks. When the blocks lie still, I'm good. One of the pleasures of this is that I can gaze at something pleasant and keep the strobe blocks in my peripheral vision. Try it - it's profoundly different from focusing on a small screen, and promotes the practice of listening with your whole body. 
My pianosens is in the mail. I'll be tuning my own Kohler & Campbell studio in a few days, if I can locate the piano under the books and papers and boxes and bags.

No one out there seems to understand the Freeze frame concept. It's up to the inventor to explain.

No, the Freeze is not a snapshot, it is window of time from the hammer strike to about 700 milliseconds thereafter, where that 'average frequency' is measured over that duration of time window where the signal power is highest. There is no such thing as an instantaneous frequency measurement in zero time.

The real 'driving force' for the invention of the Freeze frame window is that the 'pitch' also decays as the amplitude decays. Can any human watch a continuous reading and remember were the pitch was consistently? No. The Freeze indicates that, and it is the ONLY consistent way that has ever been described or reported.

The piano decays in amplitude about 20 dB / sec in the mid range, and about 20 dB in 300 milliseconds in the upper register. Tune where the power is the highest, in the first second. The piano not only decays quickly but the human absolutely cannot perceive the pitch accurately in the first sec, nor can most any ETD, 'unless' it integrates it over that time using an FFT. 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-21-2024 14:41
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 11:52:00 PM
From: Paul McCloud
Subject: RE: First tuning experience using PianoSens

Hi Fred: 

I have been using the Accutuner since I began tuning in 1988, so I know what you mean about the slower display.  I can't imagine it being any jumpier.  

As far as using Pianoscope, I started using it about a year ago, and I've seen it make improvements along the way, especially the Freeze concept.  I wish all of the ETD's had this feature.  Anyway, the Freeze line helps you make better decisions because it only measures during a short duration of time after the initial blow.  The pitch is more accurately determined that way.  As you know, the pitch is constantly changing as the decay period goes on.  If you measure the pitch during the attack phase, in a limited time frame, you'll get a better reading.  Since PIanoscope (and other ETD's that use multiple partials) require measuring them to calculate the pitch, if they fade away, they can't be used.  So, the measurement needs to be done while the partials are sounding their loudest.  Waiting seconds later, the pitch is falling and the partials are fading.  All of the pitch information needed is in the first second of string vibration.

Here's the thing- this feature would not have been possible to create until recently because there wasn't enough processing power in our IOS devices to run it.  It is entirely a new concept.  Of course, we have always advised to "tune to the decay" except in the treble where we "tune to the attack".  Well, obviously, there is not other option up there.  But creating a "time window" where the pitch measurement is done, the reading is more consistent.  That's why it was invented.  The Freeze line appears after a specified time measured from the initial blow.  You can set this time window from 100 miliiseconds to 999 (1sec).  During this time, an average measurement is made, and then displays the pitch as a grey line.  Play the note again to reset the grey Freeze line.  Be sure to release the key so that the string stops vibrating for a half second so that the program knows you're resetting the Freeze.  Otherwise it gets confused and gives a false reading.  Very important point.

You can use the red line, and the "strobe" in the background, to get close to the pitch, but then use the grey Freeze line to land as close to zero as possible.  If you're in the high treble, this is more tricky especially with false beats and such.  Sometimes I'll watch the strobe in the background and try to get it to slow down and stop, or oscillate back and forth in one place.  I look at all the indicators and decide ultimately with my ears where I want to leave it.

As far as unisons go, if I use the Freeze line, I can tune beatless unisons consistently.  My tunings are much cleaner than before, and I can correct them later by simply isolating the strings and retuning them.  Bad unisons stand out like a sore thumb when the rest are quiet.  If you like smearing unisons, that's a choice too.  There is a reason that all of the ETD's suggest using your ears to tune unisons, and that is because they don't have the necessary consistency to make them perfect.  Of course, you can get them close, but by ear getting them closer than .4 cents is hit and miss.  Adding the Freeze to Pianoscope makes it easier and more accurate, down to .1 cent. Doing this requires a little extra time and finesse, but the result is worth it, IMO.

Using the Freeze feature makes it much easier to tune than watching the red line jumping all over the place.  I use them both while tuning, but the Freeze is where the pitch is most accurate.

I hope you'll give it another shot, and I hope I've explained things a little better.

------------------------------
Paul McCloud, RPT
Accutone Piano Service
www.AccutonePianoService.com
pavadasa@gmail.com

Original Message:
Sent: 03-20-2024 20:39
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 7:08:00 PM
From: Steven Norsworthy
Subject: RE: First tuning experience using PianoSens

Fred, make sure the IH is 'different' with the mic and sensor. It matters. Did you use separately acquired IH's? 

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-20-2024 15:15
From: Fred Sturm
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/20/2024 8:03:00 AM
From: Peter Grey
Subject: RE: First tuning experience using PianoSens

And the sensor's total deafness to outside interference (as many of us including Fred) have noticed means that we can carry on our work in total defiance of even the most obnoxious and persistent noises going on around us. No more: "Quiet please!, I'm tuning the piano!" 

I recall trying to tune on a tourist ship going out to the Isles of Shoals here out of Portsmouth years ago. The piano was right above the engine compartment (or at least it seemed like it). I thought I could accomplish it...forget it. I had to wait till the ship was docked (and it was NOT a planned trip). Those risks are now gone. 

Peter Grey Piano Doctor 

------------------------------
Peter Grey
Stratham NH
(603) 686-2395
pianodoctor57@gmail.com

Original Message:
Sent: 03-20-2024 02:24
From: Steven Norsworthy
Subject: First tuning experience using PianoSens

Fred, there is a very rational reason why older apps that go back 30+ years will not show the instantaneous differences as compared with newer apps. The newer apps are based on using the FFT (Fast Fourier Transform) whereas older apps may use either bandpass filtering or zero crossing counts to measure frequency differences. The zero crossing technique goes like this: let's say that you want to measure 440 Hz. There are two zero crossings per sine wave, so in one second you have 880 zero crossings. If the 440 Hz tone (assuming all other partials filtered out infinitely so) is now 1 cent off, which would be 440.25 Hz. There are 1761 zero crossings in 2 seconds at 440.25 Hz. That means that you have to wait 2 seconds to detect 1761 zero crossings to get to a 1 cent-accurate measurement. If you want a 0.1 cent accurate measurement, you would need to detect 20 seconds of zero crossings. So the zero crossing technique simply estimates the 0.1 cent and cannot directly measure it. However, using the FFT like the newer apps, lets say the sample rate is 32 kHz, and so in one second you use a 32k-point FFT which gives you 16K frequency bins per second, making that spacing 2 Hz per bin. Then you use an interpolation technique such as is well known to FFT engineers and that gives you 100's of points between 2 Hz bins, or milli-Hz resolution in one second. It also gives you continuous overlapping readings every couple hundred milliseconds, giving the user visibility on an ongoing basis to see things like jitter and pitch decay. This kind of visibility is NOT possible with the older zero-crossing apps. Hope that helps explain what you vs others are observing with other apps.

Kind regards,

Steve

------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101

Original Message:
Sent: 03-19-2024 13:13
From: Fred Sturm
Subject: First tuning experience using PianoSens

I have had Pianosens for a couple weeks, but it was a light tuning time (and I am not doing that much tuning these days) so have only tuned about 4 - 5 pianos with it. I will make a few initial comments (I may do a more detailed review later):

It maybe adds five more minutes to the tuning time all told in setting up, moving the sensor around and whatnot. This time is largely made up by not needing to move around the phone to get the mic placement better. Logistics aren't as bad as I imagined, and easy to adapt to (eg, the hand is already moving the mute, so shifting the sensor isn't that much more). It would be nice if the mic wire from the sensor (hard wired) that goes to the preamp wasn't so long. It has to be kept coiled up, and takes up that much more space.

I thought I'd compare display results for some individual strings by toggling between plugged in and unplugged, but that isn't practical. It takes quite a while for the phone to figure out it needs to change input, so there is a big time delay. I have an iPad with RCT on it, so will do them in parallel next time I have an opportunity. But my sense is that the display is comparable in being clear (not jumpy) for most notes, but significantly better than mic for the more problematic notes. (If the sensor placed away from the termination, the display can actually be more jumpy than with a mic - interesting. But the vibrations at the termination are what get conveyed to the bridge. Looking at high speed video, you can see the shiver of the upper partials at the inner parts of the string, like the strike point, but they have a clearer and more defined pattern at the bridge pin). 

Some mention the top octave. Using RCT, I don't find that dramatic a difference.

I have tuned unisons to the RCT for many years, so that I could clearly establish the stability of each string. No difference here except when there is a string or two that are displaying less ambiguously with the sensor. (Of course, I use my ear to verify, but I use the ETD to correct when I decide it needs refining).

IMO, while there is at least a noticeable improvement overall in precision of display and ability to refine a tuning, the main and quite useful benefit is the fact that it isn't listening: it doesn't care if there is a leaf blower, a furnace fan, a vacuum cleaner, stage crew setting up, even musicians warming up while you are trying to finish the concert tuning. It is essentially a high quality pick-up, and is only responding to the movement of the string.

I had hoped it might be helpful for harpsichord and clavichord, but was disappointed at least in initial trials. I knew it wouldn't pick up the brass strings, but it wasn't that great for the steel ones (if there was more resolution, it was subtle). 

------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda

Original Message:
Sent: 03-17-2024 12:21
From: James Busby
Subject: First tuning experience using PianoSens

Original Message:
Sent: 3/17/2024 11:35:00 AM
From: Scott Kerns
Subject: RE: First tuning experience using PianoSens

Thanks for sharing your experience Rick. It's very helpful. I'd be interested in how it works on a spinet. 

------------------------------
"That Tuning Guy"
Scott Kerns
Lincoln, Nebraska
www.thattuningguy.com

Original Message:
Sent: 03-16-2024 19:22
From: Rick Clark
Subject: First tuning experience using PianoSens

Here is my first piano tuning experience using PianoSens. I am using Verituner for Android on a 2023 Moto G Play running Android 12.

Having reviewed the instructions in advance, I connected the devices and wires in the order given in the written instructions and confirming the gain setting on the preamp as 8. I then opened Verituner and everything was running as it should be, with the Verituner receiving the signal from PianoSens.

The piano was a Kawai GE-1. So it is an older, and a very small scale grand. Since I needed to create a new tuning file, as usual I had to allow for some extra time for Verituner to acquire the inharmonicity data. I did notice that on average, the time needed for the inharmonicty "I" indicator to fill up was less than when using the microphone. A couple notes did hang up a bit, but fewer than usual. On average, without a doubt it took Verituner less time to get all the inharmonicty data into memory. Reviewing the inharmonicity data afterward, none of the notes had any spurious values they occasionally get using the mic indicating they need to be remeasured. So some time was saved there.

Since I work the tuning strictly with mutes rather than a temperament strip, in this short piano it was sometimes a bit crowded because due to the plate shape I wanted to sometimes put the mutes in the same place the sensor needed to be, but I can see I will be able to figure out ways of adjusting my technique,using different mutes, and getting it into my body memory so it all works. I am not worried about it. And on a different piano the mute positions will be different.

This particular piano had its share of pitch oscillations caused by less-than-perfect terminations or other issues. Throughout most of the middle I felt I was getting pretty similar pitch readings with the sensor as I get using the mic. There were sometimes oscillations or jittery readings, but you could hear they were really happening at the string, and the sensor was just transmitting what the string was doing. If it were in mic mode it would have been worse, because it would have also been picking up various "in the air" acoustic effects also. For instance my Android device has microphones top & bottom, while the bottom mic is aimed at the soundboard, the top mic is hearing reflections off the underside of the lid. That has to be creating acoustic confusion with all kinds of phase cancellation or comb filtering effects. So overall, the PianoSens pickup with its focus on only the string just has to work better in that regard.

It also became clear what a pleasure it was to place my ETD wherever I wanted to based on visibility rather than having to take into account whether it is also well positioned for the microphone to pick up the sound.

The biggest "aahhhh moment" however was in the high treble. Up where your ear is hearing all kinds of beating you can't tune out of it, and the indicators on the tuning device are having the hardest time locking in. With the PianoSens, despite all the aural confusion, I was getting nice steady readings on the ETD. Really a pleasure. Then the capper was the top few notes where my device is just all over the place and unable to lock on anything in mic input mode. And aurally there is so much false beats confusion. With the sensor working I was getting not only a clear, locked-on pitch signal, but it was actually sustaining a bit and remaining locked onto the pitch.

So overall, based on my first tuning using PianoSens, I feel kind of like it is "worth it for the top octave alone" but with so many other improvements over aural or mic input ETD, it's kind of a "no-brainer". Can I get a great sounding tuning without it? Yes. But it feels like I will be preferring it, for sure.

------------------------------
Rick Clark
Carlsbad
(760) 436-3322
------------------------------

An update to Fred.. of course even a medium blow force will integrate as a sharp note relative to 2 seconds later. The power is in the first second. That is what we tune to acoustically and musically.

Steve

Any blow force below a certain level will 'not' deviate the freeze line biow to blow. That is what Paul and I and others see. You should see the same. It has to do with the nature of the nonlinearity of spring physics. The hard blows are not the way to tune.

Steve