Pianotech

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

"The result was the following spread of errors with respect to placement. The data is below. Notice the sign..."

Could you clarify how you calculated the spreads and why some are negative? I would be interested in seeing a 9x6 table showing the measured frequency of each partial at each placement of the mic. 

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com
------------------------------

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

------------------------------
Don Mannino RPT

Original Message:
Sent: 02-14-2024 21:57
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

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

Original Message:
Sent: 02-14-2024 23:16
From: Don Mannino
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

------------------------------
Don Mannino RPT

Original Message:
Sent: 02-14-2024 21:57
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

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

Original Message:
Sent: 02-14-2024 23:16
From: Don Mannino
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

------------------------------
Don Mannino RPT

Original Message:
Sent: 02-14-2024 21:57
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

I have tried from 0.1 to 1. It is all scaled. The de-tuned never comes all the way back to the blue line of power. Impossible, violates the laws of physics.

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

Original Message:
Sent: 02-14-2024 23:16
From: Don Mannino
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

------------------------------
Don Mannino RPT

Original Message:
Sent: 02-14-2024 21:57
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

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

Original Message:
Sent: 02-14-2024 23:16
From: Don Mannino
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

------------------------------
Don Mannino RPT

Original Message:
Sent: 02-14-2024 21:57
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

------------------------------
Don Mannino RPT

Original Message:
Sent: 02-14-2024 21:57
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

Don, you also asked for a tighter spread. I showed 1 cent graph and now here is the 0.3 cent spread graph.

After 6 seconds, there is a 5 dB LESS SUSTAIN. Physics are Physics, but Psycho-Acoustics are also Psycho-Acoustics. 

See figure below.

Steve

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

I don't think it's fair to ignore psychoacoustics. An impure unison will sound louder than a pure unison of similar intensity because the impure unison activates more hair cells in the cochlea. I'm no expert on this, but it makes intuitive sense to me that we would hear a "bloom" in a slightly detuned unison when the amplitude decays enough that the strings stop phase locking. I don't know whether that's the bloom Don is talking about or if it corresponds with the bump in Steven's graph. (Probably not.) My main point is, the audience in a concert hall will be listening with human ears connected to a brain, not microphones connected to a computer. Psychoacoustics are important. 

Don, you also asked for a tighter spread. I showed 1 cent graph and now here is the 0.3 cent spread graph.

After 6 seconds, there is a 5 dB LESS SUSTAIN. Physics are Physics, but Psycho-Acoustics are also Psycho-Acoustics. 

See figure below.

Steve

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

I don't think it's fair to ignore psychoacoustics. An impure unison will sound louder than a pure unison of similar intensity because the impure unison activates more hair cells in the cochlea. I'm no expert on this, but it makes intuitive sense to me that we would hear a "bloom" in a slightly detuned unison when the amplitude decays enough that the strings stop phase locking. I don't know whether that's the bloom Don is talking about or if it corresponds with the bump in Steven's graph. (Probably not.) My main point is, the audience in a concert hall will be listening with human ears connected to a brain, not microphones connected to a computer. Psychoacoustics are important. 

Don, you also asked for a tighter spread. I showed 1 cent graph and now here is the 0.3 cent spread graph.

After 6 seconds, there is a 5 dB LESS SUSTAIN. Physics are Physics, but Psycho-Acoustics are also Psycho-Acoustics. 

See figure below.

Steve

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

I don't think it's fair to ignore psychoacoustics. An impure unison will sound louder than a pure unison of similar intensity because the impure unison activates more hair cells in the cochlea. I'm no expert on this, but it makes intuitive sense to me that we would hear a "bloom" in a slightly detuned unison when the amplitude decays enough that the strings stop phase locking. I don't know whether that's the bloom Don is talking about or if it corresponds with the bump in Steven's graph. (Probably not.) My main point is, the audience in a concert hall will be listening with human ears connected to a brain, not microphones connected to a computer. Psychoacoustics are important. 

------------------------------
Anthony Willey, RPT
http://willeypianotuning.com
http://pianometer.com

Original Message:
Sent: 02-15-2024 01:04
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Don, you also asked for a tighter spread. I showed 1 cent graph and now here is the 0.3 cent spread graph.

After 6 seconds, there is a 5 dB LESS SUSTAIN. Physics are Physics, but Psycho-Acoustics are also Psycho-Acoustics. 

See figure below.

Steve

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

Original Message:
Sent: 02-14-2024 23:16
From: Don Mannino
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

------------------------------
Don Mannino RPT

Original Message:
Sent: 02-14-2024 21:57
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

I don't think it's fair to ignore psychoacoustics. An impure unison will sound louder than a pure unison of similar intensity because the impure unison activates more hair cells in the cochlea. I'm no expert on this, but it makes intuitive sense to me that we would hear a "bloom" in a slightly detuned unison when the amplitude decays enough that the strings stop phase locking. I don't know whether that's the bloom Don is talking about or if it corresponds with the bump in Steven's graph. (Probably not.) My main point is, the audience in a concert hall will be listening with human ears connected to a brain, not microphones connected to a computer. Psychoacoustics are important. 

Don, you also asked for a tighter spread. I showed 1 cent graph and now here is the 0.3 cent spread graph.

After 6 seconds, there is a 5 dB LESS SUSTAIN. Physics are Physics, but Psycho-Acoustics are also Psycho-Acoustics. 

See figure below.

Steve

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

Original Message:
Sent: 2/15/2024 1:54:00 AM
From: Anthony Willey
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

I don't think it's fair to ignore psychoacoustics. An impure unison will sound louder than a pure unison of similar intensity because the impure unison activates more hair cells in the cochlea. I'm no expert on this, but it makes intuitive sense to me that we would hear a "bloom" in a slightly detuned unison when the amplitude decays enough that the strings stop phase locking. I don't know whether that's the bloom Don is talking about or if it corresponds with the bump in Steven's graph. (Probably not.) My main point is, the audience in a concert hall will be listening with human ears connected to a brain, not microphones connected to a computer. Psychoacoustics are important. 

------------------------------
Anthony Willey, RPT
http://willeypianotuning.com
http://pianometer.com

Original Message:
Sent: 02-15-2024 01:04
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Don, you also asked for a tighter spread. I showed 1 cent graph and now here is the 0.3 cent spread graph.

After 6 seconds, there is a 5 dB LESS SUSTAIN. Physics are Physics, but Psycho-Acoustics are also Psycho-Acoustics. 

See figure below.

Steve

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

Original Message:
Sent: 02-14-2024 23:16
From: Don Mannino
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

------------------------------
Don Mannino RPT

Original Message:
Sent: 02-14-2024 21:57
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

• Here is an actual recording I made from my Fazioli F308 with a small spread of unisons vs precisely matched unisons to 0.1 cents.
• There is 6 dB less power after careful calibration of the signals to make sure the comparison is correct, as I 'normalized' the instantaneous peaks.
• Everyone without exception I have played there recordings for have all agreed with the comments in the slide.
• Thinner sound, less stable sounding, less power. 
• We should be respectful of 'subjective opinions' but data is data. 

-- Steve

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

Original Message:
Sent: 02-15-2024 09:38
From: David Pinnegar
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Original Message:
Sent: 2/15/2024 1:54:00 AM
From: Anthony Willey
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

I don't think it's fair to ignore psychoacoustics. An impure unison will sound louder than a pure unison of similar intensity because the impure unison activates more hair cells in the cochlea. I'm no expert on this, but it makes intuitive sense to me that we would hear a "bloom" in a slightly detuned unison when the amplitude decays enough that the strings stop phase locking. I don't know whether that's the bloom Don is talking about or if it corresponds with the bump in Steven's graph. (Probably not.) My main point is, the audience in a concert hall will be listening with human ears connected to a brain, not microphones connected to a computer. Psychoacoustics are important. 

------------------------------
Anthony Willey, RPT
http://willeypianotuning.com
http://pianometer.com

Original Message:
Sent: 02-15-2024 01:04
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Don, you also asked for a tighter spread. I showed 1 cent graph and now here is the 0.3 cent spread graph.

After 6 seconds, there is a 5 dB LESS SUSTAIN. Physics are Physics, but Psycho-Acoustics are also Psycho-Acoustics. 

See figure below.

Steve

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

Original Message:
Sent: 02-14-2024 23:16
From: Don Mannino
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

------------------------------
Don Mannino RPT

Original Message:
Sent: 02-14-2024 21:57
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

I may be misreading the graph labels, but if that's three strings of a unison at -0.5c, -0.1c, and 0.2c, you have too wide of a spread (0.7c). No one is arguing that this sounds good.  Would also be helpful to specify if these readings are taken with a microphone or the new sensor -- I'm assuming the latter?

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-15-2024 13:23
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

• Here is an actual recording I made from my Fazioli F308 with a small spread of unisons vs precisely matched unisons to 0.1 cents.
• There is 6 dB less power after careful calibration of the signals to make sure the comparison is correct, as I 'normalized' the instantaneous peaks.
• Everyone without exception I have played there recordings for have all agreed with the comments in the slide.
• Thinner sound, less stable sounding, less power. 
• We should be respectful of 'subjective opinions' but data is data. 

-- Steve

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

Original Message:
Sent: 02-15-2024 09:38
From: David Pinnegar
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Original Message:
Sent: 2/15/2024 1:54:00 AM
From: Anthony Willey
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

I don't think it's fair to ignore psychoacoustics. An impure unison will sound louder than a pure unison of similar intensity because the impure unison activates more hair cells in the cochlea. I'm no expert on this, but it makes intuitive sense to me that we would hear a "bloom" in a slightly detuned unison when the amplitude decays enough that the strings stop phase locking. I don't know whether that's the bloom Don is talking about or if it corresponds with the bump in Steven's graph. (Probably not.) My main point is, the audience in a concert hall will be listening with human ears connected to a brain, not microphones connected to a computer. Psychoacoustics are important. 

------------------------------
Anthony Willey, RPT
http://willeypianotuning.com
http://pianometer.com

Original Message:
Sent: 02-15-2024 01:04
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Don, you also asked for a tighter spread. I showed 1 cent graph and now here is the 0.3 cent spread graph.

After 6 seconds, there is a 5 dB LESS SUSTAIN. Physics are Physics, but Psycho-Acoustics are also Psycho-Acoustics. 

See figure below.

Steve

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

Original Message:
Sent: 02-14-2024 23:16
From: Don Mannino
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

------------------------------
Don Mannino RPT

Original Message:
Sent: 02-14-2024 21:57
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

• 1) PTG examiners I have discussed with say that this spread would be well within the tolerance.
• 2) Taken with the sensor, measured in Matlab.
• 3) I already showed a prior graph of the power drop with 0.3
  
  Original Message:
  Sent: 2/15/2024 2:00:00 PM
  From: Anthony Willey
  Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...
  
  

I was also going to ask if these were taken with the sensor or a microphone. I'm trying to wrap my head around which measurement device would be better for measuring an effect like this. My intuition says microphone. (I worry that effects like the strings vibrating sideways in some modes or out of phase with each other could be measured differently by an electromagnetic pickup, or if the "bloom" effect has components that occur in the bridge and soundboard.) 

------------------------------
Anthony Willey, RPT
http://willeypianotuning.com
http://pianometer.com

Original Message:
Sent: 02-15-2024 13:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

I may be misreading the graph labels, but if that's three strings of a unison at -0.5c, -0.1c, and 0.2c, you have too wide of a spread (0.7c). No one is arguing that this sounds good.  Would also be helpful to specify if these readings are taken with a microphone or the new sensor -- I'm assuming the latter?

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-15-2024 13:23
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

• Here is an actual recording I made from my Fazioli F308 with a small spread of unisons vs precisely matched unisons to 0.1 cents.
• There is 6 dB less power after careful calibration of the signals to make sure the comparison is correct, as I 'normalized' the instantaneous peaks.
• Everyone without exception I have played there recordings for have all agreed with the comments in the slide.
• Thinner sound, less stable sounding, less power. 
• We should be respectful of 'subjective opinions' but data is data. 

-- Steve

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

Original Message:
Sent: 02-15-2024 09:38
From: David Pinnegar
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Original Message:
Sent: 2/15/2024 1:54:00 AM
From: Anthony Willey
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

I don't think it's fair to ignore psychoacoustics. An impure unison will sound louder than a pure unison of similar intensity because the impure unison activates more hair cells in the cochlea. I'm no expert on this, but it makes intuitive sense to me that we would hear a "bloom" in a slightly detuned unison when the amplitude decays enough that the strings stop phase locking. I don't know whether that's the bloom Don is talking about or if it corresponds with the bump in Steven's graph. (Probably not.) My main point is, the audience in a concert hall will be listening with human ears connected to a brain, not microphones connected to a computer. Psychoacoustics are important. 

------------------------------
Anthony Willey, RPT
http://willeypianotuning.com
http://pianometer.com

Original Message:
Sent: 02-15-2024 01:04
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Don, you also asked for a tighter spread. I showed 1 cent graph and now here is the 0.3 cent spread graph.

After 6 seconds, there is a 5 dB LESS SUSTAIN. Physics are Physics, but Psycho-Acoustics are also Psycho-Acoustics. 

See figure below.

Steve

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

Original Message:
Sent: 02-14-2024 23:16
From: Don Mannino
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

------------------------------
Don Mannino RPT

Original Message:
Sent: 02-14-2024 21:57
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

Don, you also asked for a tighter spread. I showed 1 cent graph and now here is the 0.3 cent spread graph.

After 6 seconds, there is a 5 dB LESS SUSTAIN. Physics are Physics, but Psycho-Acoustics are also Psycho-Acoustics. 

See figure below.

Steve

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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

Steven,

After your demonstration last night (while you were enjoying some pie 😋) I tuned a demonstration of the tonal difference between a solid unison (C4) and a unison (C#4) tuned for better tone by very slightly shifting one of the strings. Unfortunately another tuner decided to practice using the tuning device on the C# before I could go collect you and have you listen to the difference.  It was my hope to help you begin to understand the reason for slight <cracking> of the unison for better sustain.

Of course the attack is more solid with the unison super precisely tuned. However, I listened to all of the unisons which you had set precisely (and which had not drifted).  All of them indeed had a strong and solid sounding attack.  And all of them had worse sustain.  They exhibited what I call a "pinched" unison, where it is solidly in tune but the sound decays more quickly than one with a little life in it.

I tuned the C# with one of the strings offset (by ear) just enough to eliminate the pinching off effect in the sustain. I agree completely that the initial attack was less solid, but the tone swelled after the attack, and the sustain was noticeably longer.  The tone of the note had a much more beautiful sound.  And when done well in a concert tuning there is no question that the overall effect in the piano is really quite gorgeous.

Now, you love to measure and quantify everything in super detail, and if you would like to try to do that then consider this: If there is a 3 or 4db weaker initial attack in the tone, that energy is not lost!  I'm sure in your extensive reading and studying about piano tone you have learned about coupling between piano strings.  There are numerous articles available on this topic. I would like you to consider that the energy missing from the weaker attack is merely shifted later in time, as the energy shifts between the strings, and this explains the longer sustain. There is no beat, mind you (even though you rather insultingly suggested in an earlier email sent to my wife that we who tune these unisons are listening for ". . . a beautiful vibrato piece of music." ) There is only a clear but beautiful swell in the tone.  A truly artistic concert tuning is one in which all of the unisons are tuned with similar speed of swell.

I learned this concept in the 1980s from another very experienced concert tuner and practiced it when tuning for concerts, and have refined it and worked to become better at it in the many years since in concerts and studio recordings. 

Some years ago I was asked to tune a very beautiful Yamaha S4 grand for the evening recital at a PTG conference.  Afterwards I was approached by multiple piano technicians who asked me "What did you do to that piano!  It sang so beautifully!  How did you do that?"  I did nothing but tune the piano, but my style of tuning did make it sing - it was an excellent piano.

I have extensive experience working with the staff at Kawai's R&D facility in Japan, and learned many, many things about piano tone that are not at all straightforward and intuitive.  However, this does not make them mysterious or questionable.

Open your mind, Steven, to what the experienced people are telling you!  Your device is wonderful, I have no qualms about its effectiveness and excellent design. However, just because you have so much knowledge in signal analysis it is simply hubris for you to question things you are told by technicians with years of high level experience working with pianos in very demanding situations simply because you don't understand it!  Experiment with your undoubtedly very lovely Fazioli at home, and graph the decay envelope of the precise unison, then change 1 string just 0.2¢ - 0.3¢.  Record that into Matlab and chart the decay envelope again, see how it changes.  I would also suggest that you do the recording / charting with your sensor but also use one of the excellent microphones you mentioned that you own.  If the 'dirty' unison doesn't sustain longer, then try tuning it aurally until you hear the swell, but no repeating beat.

You should know that your clearly expressed superiority complex is what offends people, not specific statements.  I submit to you that this is seriously turning a lot of excellent technicians (and potential PianoSens customers) away from considering your excellent product. 

If the unisons are in tune at the short interval after the attack, say, the first 500 ms, there is significantly more power in the attack if the unisons are lined up in both phase and frequency. My measurements of this phenomenon are that if the unisons have a spread of 0.7 cents, there is a loss of 6 dB of power in the first 500 ms. I simply don't understand why there is a school of thought that spread unisons are somehow preferred, or that we can wait 2 seconds after the attack to dial in the pitch. Neither makes musical sense. Tuners are playing single notes by themselves out of musical context. It would be like a pianist walking on stage, playing a single note without the sustain pedal, holding it for 5 secs, and the audience saying, "Oh what a beautiful piece of music and what a great performance!"

Maybe a John Cage composition!

I still respect people's opinions who are on the other side of this argument but I don't yet have any of them making a musical argument in the context of real piano music.

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

Original Message:
Sent: 02-14-2024 21:49
From: Nathan Monteleone
Subject: Tune with a Single Partial or Multiple Partials? An Example...

Indeed, the pitch is known to drop as the note is held. But the shorter the notes the more forgiving the tuning -- at least with the exception of unisons.  So you be be getting a significantly better attack sound out of the unisons.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 21:35
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

The nonlinearity of the piano string and hammer cause pitch decay vs time. The power is in the first second, and THAT is where we are to measure pitch, i.e, the window of time with samples starting at the strike and ending in the first second. The speed of pitch decay is the most rapid right after the strike. After the first second, pitch decay is still going, and the higher we go into the upper register in particular. There is no such thing as steady state pitch in a piano. It is also a musically right thing to do to capture the pitch measurements in the first second. The piano musical literature moves quickly in most passages, i.e., quarter-note = 60 (one sec quarter note) is a slow passage and the sustain pedal is down usually at this slow tempo. The composers intuitively knew this from hundreds of years ago to present.

Steve N.


Original Message:
Sent: 2/14/2024 9:30:00 PM
From: Nathan Monteleone
Subject: RE: Tune with a Single Partial or Multiple Partials? An Example...

Just for grins I shot a video of C4 using TuneLab.  Please pardon the crappy videography.  I get that you can't really test an app against itself, I just wanted to see how much its own output changed from position to position.

I'll try uploading it to the forums here too but I shared it at:

https://photos.app.goo.gl/uwvV45qSF6eavUWm6

I can largely see the effect being described here in TuneLab's display; notice that in positions #3 and #6 the pitch reading goes flat by a little more than a cent in the first second or so, and some of the others positions actually go sharp.

I think it's worth nothing, however, that they all converge on readings between -0.2 and 0.0 after a couple seconds. (I know, I moved the camera too soon after position 2 and you can't really see it, sorry.)  In practice, this is what I'd actually tune to, if I was wanting to get max precision out of the ETD.

So it sounds like the data's showing us that an emf sensor stands to give us the right answer quicker, if nothing else.  On the other hand, perhaps we need not be concerned that our mic-based readings are actually leading us that _too_ far astray if we're waiting for them to settle.

------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com

Original Message:
Sent: 02-14-2024 18:47
From: Steven Norsworthy
Subject: Tune with a Single Partial or Multiple Partials? An Example...

We had a lively discussion last night at the South Bay PTG where I gave a class.

One of the issues is whether we tune with a single partial or with multiple partials. If an app choses a single partial for tuning, we could be lucky to see virtually no difference where the mic was placed. But if we are unlucky we could get more than a 1 cent error range with a small mic placement movement. It is not that the movement creates a different frequency, but it creates a different resolution of that frequency. 

Here is an example. Using the note C4 on a 7' grand piano, the mic was moved around to 9 different physical positions above the soundboard in reasonable locations that a tuner could place an iPhone. The .wav files were captured and analyzed in Matlab. The result was the following spread of errors with respect to placement. The data is below. Notice the sign of one relative to the other.

Partial 2:     -0.88 cents

Partial 3:     -0.10 cents

Partial 4:     +0.93 cents

Partial 5:     -0.47 cents

Partial 6:     +0.79 cents

Depending on the tuning app, if it tunes on one partial, it could either be lucky (-0.1 cents) or unlucky (+0.93 cents).

Alternatively, if the app tunes by trying to fit the data statistically, it could use a mean, median, or standard deviation.

mean             =   0.0540

median          =   -0.1000

std. dev.         =   0.7874

range              =  1.8100

A string sensor will not have these variances because small placement changes are inconsequential and produce one answer. Then it is up to the IH curve fitting in the app to make sure the measurements align with the IH initial calculations, thus ensuring we have extremely low variance and can trust the answer with more confidence.

Best,

Steven Norsworthy

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