How resample (gamma) works?

hi guys! I’m exploring the FFT stuff inside the Girlpower/audio, and I don’t get how the Resample (Spreads Gamma) works…
In the description I found

“Defines a value for a gamma function used to create bin sizes. Slices of the incoming spread will be packed into these bins and an average of every bin is calculated.”

but seems that when the Bin Count is close to the number of original champions, the result is strange, even if the Spreads Count seems correct (see the red line)

ResampleWhat.v4p (12.9 kB)

Hello Luper,

this module is specifically useful for dealing with FFT calculations on sound.

It takes a spread and resamples it to the smaller count, but it does so non-linearly. At the lower end the values are close to the ones from the Input spread, but toward the high frequencies, more and more values are packed together into bins and an average of those bins is returned.

Humans can pretty good distinguish different low-pitched sounds, but as the pitch goes higher we are not so good at isolating different pitches. (that is why these frequencies are packed into groups (bins) and an average of those bins is returned).

http://en.wikipedia.org/wiki/Mel_scale

I’ve updated the Resample (Spreads Gamma) module, now it returns a better curve. Thank you very much for pointing that out!

Note, the sizes of the bins are not strictly the values coming from the Gamma function. But they are lying on a curve that resembles it.

Was that useful?

Best,
Anton

Resample (Spreads Gamma).v4p (11.0 kB)

Tnx Robotanton