Piano World Home Page Forums Our Most Popular Forums Piano Tuner-Technicians Forum Imprecise pitch

What I noticed while tuning unisons with entropy piano tuner (and it is also audible), many strings are fluctuating. In other words, let's say, we muted two strings and have left only one string of A to tune. THe frequency reading on entropy reads it 440.4 thetn might be 440.5, then 440.2 and then it might arrive at 440.1 or 440.3. In other words its not the same frequency throughout the whole lenght from the start of string sounding to the end.
Is this normal? Are the strings faulty? can there be any other problem - let's say the bridge.

For A4 the values you give are within 1-1.5 cents which might happen on some strings.
You might try other strings of the unison to see whether they have less fluctuations.
Still 1 cent fluctuation IMO is not that bad. You should decide where to set the string;
Just after the attack or to the sustain etc.

It is probably normal so lets not get into esoteric and unlikely possibilities just yet.

Have you 'set' the string after turning the pin? Basically after turning the pin the speaking length of the wire will be at the tension that produces the note entropy is measuring, but the other parts of the wire are at different tensions. You now have to bang the note firmly a few times to equalise the tension throughout the wire at which point the sound should be stable but at a different value to how it first sounded. So you then have to adjust it again, and set it again, until it settles with the stable value at the right frequency. Until you set it the frequency will be continually changing, until it is stable.

The pitch goes sharp on the attack and flattens during the sustain. Normal.

I like to think of it as the string being at a slightly higher tension while it is vibrating at high amplitude. When it's at rest it's a straight line (shortest distance between two points) so when it's vibrating it must be stretched a bit longer. There are probably more accurate ways of describing what's going on in terms of non-linear wave dynamics (where the frequency is affected by amplitude) but this simple explanation satisfies me.

Edit: just realized I lied above, I think... The increase in tension when the string bends is the restoring force that gives you the regular harmonic ("linear") wave motion. So I guess it would have to be the stiffness of the string that adds an extra restoring force that makes it non-linear and dependent on amplitude. And the "small angle approximation" might have something to do with it too. Typically when you model harmonic motion you are assuming small angles (amplitudes) and zero stiffness. Which isn't always true in the real world.

TLDR: yes it's normal, and it's complicated enough that I can't confidently say why it's normal.

The universe as we know it is not as accurate as we might want it to be.

According to my observation all Instruments produce a higher tone on the attack.

Thank you everyone for your advice.
For more clarity, I took a little video.
Its not only that after striking the pitch gets lower gradually, it is just walking from left to right
https://drive.google.com/file/d/1F_GhI7Uww3q5aaQN6ntrNun7p-jqqGEm/view?usp=sharing

Exactly!

End your tuning precision at "musical" and be happy with it.

How wretchedly selfish of the universe!

I like this sentiment very much, David. Did you come up with this phrasing? Can I use this quote?

Wow. That is one ugly display.

Actually I quite like the app. For a free app, this is a marvelous tool to have for the people who can not tune aurally.

If that is a single string sounding, then it isn't very pure.... You can search for "false beats" to dive further down the tuning rabbit hole!

Ron Koval

Yes, this is a single string - and there are many many strings like that

I tried the screwdriver test, which I found here a few days ago, did not help.

I think that it is the imprecision of the a software that is causing the display to look unsteady. A strobe- type detection and displays would probably give a far greater steadiness. It looks like the software is displaying the inprecise processed results of contiguous sound samples each with a short sampling duration.

Looking at the video I think you are just seeing artifacts and noise in the FFT graph. Depending on how you window (slice up and process) the signal it can do weird things to the peaks. I agree you'd probably be better off using a strobe-like display, which I'm pretty sure Entropy has. (I remember it being colorful.) Try tapping or swiping on the graph maybe?

Entropy has a strobe display, but it is a fairly primitive one. In some ways I like that Entropy graph as I think it gives some representation of how the note dances around as it is struck and decays. In my mind of think of the note consisting of many partials that decay at different rates, so the perceived frequency of the note, and also probably what Entropy measures, also changes with the decay. Hence the peak in Entropy shifts and walks around. This I find interesting, but it does make tuning a frustrating experience trying to decide what value the tuned note actually has.

I find Pianometer much easier to use and interpret. Maybe this is because it has had more work balancing response with averaging values so that a steadier value results but I think the strobe display fundamentally helps. Because PM shows individual strobe bands for each partial those partials can individually get louder or quieter but the strobe bands remain stable and usable for tuning.

That's my current theory anyway. It might be rubbish, and there is probably more than just the decay affecting how the partials interact with each other but all understanding has to start somewhere.