Partials of Forks?

[Robert Scott]

Mark Schecter wrote:

> Well, that's very nice in theory. But I have RCT, and I have used its
> component Pianalyzer to measure the sound output from a fork, both held
> in the air next to the microphone, and held in physical contact with the
> computer case that contains the microphone. I assure you that in both
> cases, there is not only second, but third partial present, in varying
> degrees depending on exactly how the fork is presented to the mic. And
> in no case were the frequencies measured exactly harmonic - sometimes
> they were slightly sharp and sometimes slightly flat, but never exactly
> zero. The fact that they do vary suggests some possibility of anomalies
> in either the measuring or the calculating, and I have no way of
> determining what those may be, but there is not nothing there, even when
> the fork is held between two fingers in air.

Yes, you will will always get some harmonic distortion in the process of amplifying a microphone signal and converting it to a digital signal and subjecting the digital signal to an analysis program that uses round-off.  But the harmonics were added by the processing.  They were not there to begin with.  But why not trust your ears?  They have much less distortion than a computer sound card.  Play F3 softly and hold a ringing A-440 fork near your ear.  Do you hear a beat?


> Having said that, suppose we accept that the fork does produce 880.0. If
> you use the Yamaha method, you are tuning A3 such that its 4th partial
> is 880.0. If you then tune A4 to that, how do you know when you have
> A4's fundamental at 440? If you tune a 2:1, A4 will be flat, because
> there is more than an octave between A3's 2nd and 4th partials. If you
> tune a 4:2, A4 will be even flatter, amount depending on inharmonicity. No?

Absolutely right!  Sorry to have belabored a minor point while ignoring your major one.

> My point is, you have a fork, supposedly producing exactly 440.0. Why
> not use it directly, as in F2-A4=F2-Fork,...

Or, for more precision, how about this:  D3-A4=D3-Fork.  Instead of trying to compare two beats that are relatively fast you would be comparing two beats that are much slower.  Therefore a .5 Hz difference, for example, will be easier to distinguish.  The only risk is that by using the D3 you might be on the wrong side of beatless - something that could not happen using F2.  So maybe you could use F2-A4=F2-Fork to ensure that you are close and then use D3-A4=D3-Fork to fine tune A4.  If you were on the wrong side of beatless on the D3 check, then the F2 check would definitely show the error.

Robert Scott
Ypsilanti, Michigan