Errata: In my post #20 I stated my readings from different positions varied by .02 cents, I meant .2 cents and that was referring to the movement or "jitter", actually the readings on the video are fairly calm. Thanks to Steve N. for pointing out the typo.
Paul, you make some good points. I have to say that due to all this discussion over the past several months, I've been paying more attention to what I'm doing and because of that my tuning has improved too. There's always more to learn.
Original Message:
Sent: 07-02-2024 12:47
From: Paul McCloud
Subject: How Spatial Position Affects Piano Tuning (YouTube Video)
With all due respect, when do you get your reading? I have your app, but the question is when are you supposed to take the measurement? The digital number in the center is fluctuating around, so which one is correct? I can see all the partials moving around, but the center indicator is hard to follow and know when I'm right on. It would be helpful if there was a time measurement window during which a consistent reading could be made.
What is not being acknowledged in these videos is that the partials are fluctuating and spreading out after the first second of impact. The best readings are done during the first second when the most partial information is present. The proof that this is true is that unisons can be tuned, each string by string, to within .1 cent, and there's no need to tune them together by ear. When there are false beats or mismatched strings of course, then you make your best judgment. I've been doing this for the last year, and my tunings are more clear than I have ever been able to do before. And my clients do notice the improvement.
This discussion is only aimed at the electronic tuning community. Tuning with an app has always had the caveat that there will be some corrections needed occasionally. Especially the single-partial devices. Since I jumped ship on my old device, which out of laziness I thought was "good enough", using one with multiple partials was a vast improvement because they take them into account. But this requires specific information from the input device. If one or more partials are weak or missing, it throws off the calculation and/or the pitch measurement. Therefore, we must look at where this information is coming from, whether it's as good as we assume it is. When the freeze function was invented, my tunings improved even more because the pitch measurements are taken at precisely the same narrow window of time.
These apps are not showing what is going on behind the scenes. How the calculations are being done, what is being averaged and by what method is proprietary. They are essentially "black boxes", to be truthful. The averaging which is being done might make a steadier indicator, but it's actually masking what is happening. We know that the pitch of the fundamental and all the partials are moving, even if we're not aware of it. When we dissect the sound with a spectrogram over time, then we can see what is happening behind the user-interface. Using this information, we can better understand which window of time is best to take the measurements. My experience is that tuning during the very early (attack) gives a better, more reliable and accurate result. If you wait for more than a second without considering when to set your pitch, you might get a steady display, as the videos show, but that will still not allow you to tune each string individually and have them in good unison when you're done.
As far as using mics, which every app uses, they introduce all kinds of anomalies for various reasons already stated. All of the acoustic variances coming from the piano interfere with the real information that the apps need to function optimally. The only information we need is the resonance of the strings, which means we need another way of getting this information besides a microphone or other acoustic pickup. Thus, using a magnetic coupling eliminates these problems and gives clearer information. The result of using a sensor is that there is less jitter and fluctuation in the indicator. Using a sensor during the first second gives the greatest benefit, though it does help somewhat during the later time after that. This is why some apps don't seem to improve much when using the sensor.
Admittedly, there is coupling through the soundboard and bridges even if they are muted, but that is insignificant. We aren't tuning while playing other notes (unless you're tuning by ear). The Weinrich effect of course is evidence that there is coupling going on when all strings are sounding.
As far as top or bottom octaves being off by a few cents, well that's your choice. I don't agree. The aural tuning methods that we've been using all these years do work, so why all these tuning apps? It's because they are more accurate and make the tuning process easier and faster. But now we are discovering that there are indeed errors and inaccuracies within these apps and from using mics to measure string frequencies. The question is, does it matter? Maybe some don't think so, but it does to me, and the feedback I get from clients confirms it. YMMV.
------------------------------
Paul McCloud, RPT
Accutone Piano Service
www.AccutonePianoService.com
pavadasa@gmail.com
Original Message:
Sent: 07-01-2024 23:44
From: Anthony Willey
Subject: How Spatial Position Affects Piano Tuning (YouTube Video)
@Steven Rosenthal
Nice video. Here's a similar video that Nathan Monteleone made a few months ago:
https://photos.app.goo.gl/uwvV45qSF6eavUWm6
showing TuneLab converging on about 0.0 to -0.2 cents for the note C4 when placed in several different positions, and another video I made as a sanity check after I watched Nathan's video, with PianoMeter doing basically the same.
https://youtu.be/TXsXJHcdpvw
It would be interesting to do similar tests using more notes outside the mid-range. Though from a practical standpoint, it's mostly the mid-range where I care about sub-cent accuracy. If the top or bottom octave is a couple cents off nobody cares.
------------------------------
Anthony Willey, RPT
http://willeypianotuning.com
http://pianometer.com
Original Message:
Sent: 07-01-2024 21:25
From: Steven Rosenthal
Subject: How Spatial Position Affects Piano Tuning (YouTube Video)
Okay, here is a short video checking for pitch variance due to proximity using Verituner. I used the note A4 and placed the device directly above, at the extreme end of the bass, and extreme treble and pretty much got the same reading. I tested single strings, not unisons which would introduce other variables. While the mic is of course subject to ambient noise, the readings are pretty smooth. I also tried this with A2,3,5 with similar results; I didn't record them as I think the file would be too big to post. I will continue to try this out on other pianos this week.
This isn't offered as a refutation to Steve N's tests which I'm sure are accurate, in a way it asks more questions than it answers. My interests are practical and specific to being able to get consistent results with regard to tuning using my device and app.
Thinking about this subject in general, I wonder if the phenomenon of pitch variance related to location is one that we perceive more as color as opposed to intonation as these variances are transient and non periodic. After all, mics aside, there is the performer, audience, and perhaps other musicians, all in different locales, one might think this phenomenon would make intonation a bigger issue than it seems to be. Imagine a piano playing with a string quartet, the players all in different positions, yet they are capable of playing in tune.
------------------------------
Steven Rosenthal RPT
Honolulu HI
(808) 521-7129
Original Message:
Sent: 07-01-2024 19:51
From: Steven Norsworthy
Subject: How Spatial Position Affects Piano Tuning (YouTube Video)
We got a reply from Rick Clark on the issue by email. Here it is:
------------------
There are many variables at play regarding this subject. While I have been a VT user for quite a long time, I had been running it on an iPaq ( a relatively ancient device) up to the time I got PianoSens, at which time I bought a new dedicated phone plus upgraded to the latest version of VT for Android. I at first was cradling the phone vertically for tuning, like the iPaq which did not have a horizontal mode and had a mic on the bottom edge. During the time I mounted the phone vertically I did get quite a bit of variance how well the tuning target display read the note from position to position and a lot of the readings were jittery. However, knowing the phone has a mic top & bottom (unlike the iPaq), I started using it in horizontal mode and got much more reliable readings that were more tolerant of position, and less jittery, though some jitter was still present on some notes.
When I added PianoSens into the array, The display got even steadier and more reliable and even less jitter. On a well-prepped good piano, I get solid readings all the way up to the top note. This created the ability to work more reliably to a 1/10 cent accuracy standard. To a higher degree than is possible using mic mode I mean. While I was quite happy and able to achieve beautiful tunings in mic mode, using PianoSens it feels like I am entering a realm of "super tunings" if you will. I also have to assume if I am achieving 1/10th cent accuracy on each string of the unison at the time of the tuning, as the tuning inevitably starts to drift over time, the unisons will hold together better than if the accuracy was less during the tuning.
I hope this has been of some value to the discussion.
Regards,
Rick Clark
------------------------------
Steven Norsworthy
CEO/President
RF2BITS, Inc.
Cardiff CA
619-964-0101
steven@rf2bits.com
Original Message:
Sent: 07-01-2024 15:46
From: Steven Norsworthy
Subject: How Spatial Position Affects Piano Tuning (YouTube Video)
I can put you in touch with Rick. He sees a large reduction in jitter with the sensor compared with the mic, using Veritune. Every piano and every note is different. Too many variables, but all I can do as a researcher is record the raw data, and perform spectrograms without an app doing it for me. The problem is that we don't know how Veritune or any other app processes the partial frequency drifting. For example, the single-partial programs may lock onto a partial that is fairly steady. A multi-partial program could be doing some averaging of which we don't know the details. I am confident that I have an app-independent methodology and that the eigenmodes are causing spatial variance in the results.
Steve N.
------------------------------
Steven Norsworthy
CEO/President
RF2BITS, Inc.
Cardiff CA
619-964-0101
steven@rf2bits.com
Original Message:
Sent: 07-01-2024 14:14
From: Steven Rosenthal
Subject: How Spatial Position Affects Piano Tuning (YouTube Video)
"Rick Clark Review on YouTube:"
I would have thought the same til I did the experiment. Perhaps Rick can give it a try next time he's near a piano.
------------------------------
Steven Rosenthal RPT
Honolulu HI
(808) 521-7129
Original Message:
Sent: 06-30-2024 18:54
From: Steven Norsworthy
Subject: How Spatial Position Affects Piano Tuning (YouTube Video)
Rick Clark Review on YouTube:
He is a Veritune user and what he gets is not what Steven R. is getting and here is his comment:
"Those are top grade mics placed in good positions. Imagine how much wilder the deviations would be where ETD mics end up getting positioned for a tuning- which is to say for good line-of-sight between the tuner and the screen, not good audio. Not to mention whatever weird characteristics cheap built-in device mics have."
Rick Clark
------------------------------
[Steven] [Norsworthy]
[CEO/President]
[RF2BITS, Inc.]
[Cardiff] [CA]
[619-964-0101]
[[steven@rf2bits.com]
Original Message:
Sent: 06-30-2024 17:33
From: Steven Norsworthy
Subject: How Spatial Position Affects Piano Tuning (YouTube Video)
The only scientific objective way to know what is happening is to capture the sound and perform spectrograms on all the partials. If you can propose from a signal processing professor or expert another way, I am willing to consider it. My partner, PhD MIT, is considered one of the top spectrum analysts in the world. He and I wrote the software to do the analysis I showed in my video. We stand behind the methodology. Let me say that objective measurements like this across many pianos is 'required' to engage in a dialog that moves the ball forward. Otherwise we are dealing with opinions and conjecture. --- Steve N.
------------------------------
[Steven] [Norsworthy]
[CEO/President]
[RF2BITS, Inc.]
[Cardiff] [CA]
[619-964-0101]
[[steven@rf2bits.com]
Original Message:
Sent: 06-30-2024 17:19
From: David Pinnegar
Subject: How Spatial Position Affects Piano Tuning (YouTube Video)
Steven N
The last 30 seconds of the video analysis
https://youtu.be/_yLt0GotsA8?t=454 will be of interest using your own samples from your own piano and support Steven Rosenthal's experience. It shows the 2nd partial which demonstrates rock solid pitch between strike and sustain, just a change of phase but not frequency, rock solidly the same between measurement with your device, with the left microphone further from the string and the right microphone nearer the string.
The machine used gives accurate and instant stroboscopic measurement of pitch and the graphs that you're producing do not accord with the measurements.
Best wishes
David P
-- - - - - - - - - - - - - - - - - - - - - - - - -
David Pinnegar, B.Sc., A.R.C.S.
- - - - - - - - - - - - - - - - - - - - - - - -
Original Message:
Sent: 6/30/2024 5:00:00 PM
From: Steven Norsworthy
Subject: RE: How Spatial Position Affects Piano Tuning (YouTube Video)
Steven R., since I don't have access to your piano nor do I use the same app, I can only say that Paul McCloud tunes 4 pianos per day 6 day per week and was the early beta tester, and he consistently got 1-2 cent variances on nearly every piano, some notes none and some notes worse. He was an early advocate of trying different types of mic patterns, and it did not make much improvement. He has tuned over 300 pianos now with this methodology. Your experience is the 'first' yet I have heard of where there is virtually no variance. Perhaps it is a combo of your piano and app? Here are a couple of things that could be going on: 1) the app could be doing less intra-bin interpolation and simply quantizing to a coarser level, not sure and you might want to ask the app maker. 2) your piano's soundboard has much less eigenmodes and more of an omni projection of the acoustic waves. --- Steve N.
------------------------------
Steven] [Norsworthy
CEO/President
RF2BITS, Inc.]
Cardiff] CA
619-964-0101
steven@rf2bits.com
Original Message:
Sent: 06-30-2024 16:42
From: Steven Rosenthal
Subject: How Spatial Position Affects Piano Tuning (YouTube Video)
Thanks for this presentation, very clear.
Since this issue came up a few months ago, I've tried on several pianos to duplicate it in a fashion by moving my device, using the Verituner app, to see if I would get different readings by putting it in different locations. In fact, I consistently got the same readings moving the device around within about a 4 foot radius. There was some slight variation above A5 from around 3 feet, about .02 cents but oddly enough that resolved after about one second to the original reading, the opposite of what one might expect given the drift over time shown in Steve's analysis. Note: I tested single strings to avoid blurring due to Weinrich influences.
Perhaps other's might try this using different apps. And, Steve, I know this isn't a substitute for your methodology but it does attempt to measure the effect in a practical way. If you can explain this, I'm all ears.
------------------------------
Steven Rosenthal RPT
Honolulu HI
(808) 521-7129
Original Message:
Sent: 06-30-2024 15:17
From: Paul McCloud
Subject: How Spatial Position Affects Piano Tuning (YouTube Video)
Thanks very much for posting this video. The explanation of how the graphs are created and what they mean is very helpful. Most of us are not engineers, so a lot of this information is a bit mysterious when presented with different colored and jagged lines. The bottom line is understood- that the upper partials and fundamental frequencies diverge over time.
Now that we have the capability to utilize this information, we can make better decisions in the tuning process. We've understood the difference between tuning to the attack vs tuning during the decay period, but it's now easier to quantify which is which.
Choosing a sensor vs a microphone is important because mic placement skews the results. This information is not specific to a particular product, as one could do the same experiment using any type of sensor (piezo, magnetic, etc.). The fact that a magnetic string sensor is picking up only the vibration of the string and not the soundboard/bridge acoustic system removes any extraneous sympathetic vibration or resonances from the signal to the ETD. That is in addition to the variations due to the Eigenmodes (I call them hotspots) caused by random placement of an acoustic sensor.
I hope we can agree that this informative video is useful for piano tuners, and not an implied advertisement for a product. The phenomenon of partial frequency spreading is important to understand, at least to the users of various ETD's, especially those which use multiple partials for calculation. It also explains the difference between tuning to the attack vs the decay. Which one we choose is up to us.
I wish to thank Steve for preparing these videos which are very helpful and enlightening to our tuning community. He has spent months and months exploring the actual science of tuning, utilizing his knowledge of signal processing to uncover what is really happening with piano strings and their behavior. It has certainly improved my own tunings immensely.
------------------------------
Paul McCloud, RPT
Accutone Piano Service
www.AccutonePianoService.com
pavadasa@gmail.com
Original Message:
Sent: 06-30-2024 05:06
From: Steven Norsworthy
Subject: How Spatial Position Affects Piano Tuning (YouTube Video)
I have completed a research study over the past 6 months on how spatial position (mic or ear position) affects piano tuning accuracy.
I first ran this past Patrick Draine and he approved it. Therefore, please keep it up. Thanks.
Link Below:
https://youtu.be/26zskT_GvkA
I would appreciate comments be kept within the thread rather than starting a new thread review of this thread topic.
Best,
Steve N.
------------------------------
Steven Norsworthy
619-964-0101
steven@rf2bits.com
Cardiff, CA
------------------------------