JesterN

Hi there, it's a common problem i'm facing: how to extract fft of lists in max.

Audio signals are ok but what about sensor data?

I attach a simple patch that trasforms the values in audio signals and then calculates the fft?

Sound a bit simplistic to me, but what is the correct way then to calculate the FFT in this case?

There is a windowing object that stores 256 samples, e.g. 1 second of osc signal from a general sample, what to do then?

Hi there, it's a common problem i'm facing: how to extract fft of lists in max.
Audio signals are ok but what about sensor data?
I attach a simple patch that trasforms the values in audio signals and then calculates the fft?
Sound a bit simplistic to me, but what is the correct way then to calculate the FFT in this case?
There is a windowing object that stores 256 samples, e.g. 1 second of osc signal from a general sample, what to do then?
Thanks in advance
A


fft-of-lists

One idea would be to store the data in a matrix, and to use [jit.fft].

Jean-Francois, thanks for the answer, I'm not really a Jitter guy, so I wouldn't know how to fill a matrix and how to extract the fft data.

But sounds interesting I'll start studying about it.

Jean-Francois, thanks for the answer, I'm not really a Jitter guy, so I wouldn't know how to fill a matrix and how to extract the fft data.
But sounds interesting I'll start studying about it.
Thanks


... or you can check out bach.fft — www.bachproject.net

... or you can check out bach.fft — www.bachproject.net
cheers
aa


Thanks andrea, I'll check it and in case write you back!

Thanks andrea, I'll check it and in case write you back!
A


HI Andrea thanks for bach, it's very helpful!!!! I still have couple of questions. The first concerns the bach.fft, is it calculated with a windowing? Cause When I use a window I introduce artifacts, so I'm thinking there is already a windowing inside the bach.fft itself. Second question is the folding over of the frequencies above the nyquist frequency. Is there a neat way to remove that? You can see in my attached patch (which duplicates my data) that when the sine is above the nyquist come back into the spectrum..

HI Andrea thanks for bach, it's very helpful!!!! I still have couple of questions. The first concerns the bach.fft, is it calculated with a windowing? Cause When I use a window I introduce artifacts, so I'm thinking there is already a windowing inside the bach.fft itself. Second question is the folding over of the frequencies above the nyquist frequency. Is there a neat way to remove that? You can see in my attached patch (which duplicates my data) that when the sine is above the nyquist come back into the spectrum..
Id like to clear that if possible...


fft of lists ???