corkscrew and speed of sound.

[solid]

I have increased the lengths of all buffers by 4, it goes in good direction but maybe it is not optimal and the program has to be rectified..
The results are for the the same my 250 mL cylinder.
       

[julien]

I tried this :

   

One other problem is that the value  of the peak frequency doesn't freeze once calculated...
I tried few things to change this but i failed...

[solid]

Two changes and it works better:

1. In initial if for the "false" "value:0" changed to "empty" (I do not think that it is really important..)
   
2. Initial values set to some values, not 0 (Why it could be important ???)
   

One should still optimize the buffer lengths.

[julien]

Great, those modifications improve the program indeed !

[solid]

A year passed away and new Fête de la Science approaches. For this occasion here is a new version of the program. In addition to the resonance frequency determination from the Fourier transformation I tried the alternative technique using the autocorrelation function.  Two results are similar but a systematic underestimation of the resonance frequency and the corresponding sound speed is observed. Maybe it happens due to the presence of the cork near to the bottleneck when the sound is emitted (the 1/4 wavelength tube transforms to a Helmholtz resonator).

               

With special greetings to Julien.
Yours Mikhail (solid).

[Jens Noriʇzsɔɥ]

Dear Mikhail,

your «tire bourbon» experiment is my favourite end-of-the-workshop slide, so thank you for the update

I have attached a screenshot on iOS that shows some difference to yours (these have already been present in previous versions). The horizontal rulers look sort of weird and I have not yet found time to check how to improve the vertical spacing.

Furthermore, the text is handled differently and does not wrap. I have once briefly discussed with Sebastian about it. I check if I find those old notes for an alternative.

[solid]

Dear Jens,

The program is not perfect yet and we have to progress in doing it better. Unfortunately I do not have an iPhone yet ?.

A+
Mikhail

[Jens Noriʇzsɔɥ]

It is already highly appreciated by teachers in Germany as it is.

I forgot to hit the “post reply” button on this part:
I observe an opposite effect for an air column of 66 mm: 1101.56 Hz by FFT and 1111.11 Hz by Autocorrelation. For 79 mm: 984.38 Hz vs. 984.62 Hz…

There is some oddity: I can do the measurement just once being different to previous versions. Furthermore, leaving the experiment and returning does not help, even starting Audio Spectrum and returning then does not. Hm… perhaps a phyphox bug.

[solid]

I do not have the last problem. It is sufficient to stop the experiment, clear recorded data and restart again. Nearly always... Then repeat stop - start twice..

[julien]

Hello !

Excellent Mikhail ! I am happy to see that this experiment is still alive !
These improvements look great ! I will have to open a bottle this evening to try it !!!

Julien