JesterN

Hi there i'm thinking of doing an FFT of incoming sensor data, anyone has done something about that already?

What would you use it? I'm thiking of accumulating 1024 points in a lista nd then fluish them regularily.. but i have no clue how to make them in a fft with control rate...

Hi there i'm thinking of doing an FFT of incoming sensor data, anyone has done something about that already?
What would you use it? I'm thiking of accumulating 1024 points in a lista nd then fluish them regularily.. but i have no clue how to make them in a fft with control rate...
any idea?
A


fft-from-osc-data

The bach library ( www.bachproject.net ) contains a tool for performing the fft of a list

- and sorry for the shameless self-promotion ;)

Hi.
The bach library ( www.bachproject.net ) contains a tool for performing the fft of a list
- and sorry for the shameless self-promotion ;)
aa


can you tell me what the use of this would be? I'm having a hard time visualising what would happen with an fftransform of list data.

@wttenberg i think it's possible to extract some information of the frequency content of a signal for further processing... like in every audio signal..

HI Andrea i'll check that.
@wttenberg i think it's possible to extract some information of the frequency content of a signal for further processing... like in every audio signal..


So to detect something like "overtones" in the signal? I'd reckon there are easier ways to detect frequency in a simple moving sensor input.

- and can't you just use normal fft for the more complex stuff?

The list you mention - and what makes Andrea mention list processing is the accumulation that would normally happen in fft, right? The fft window, yea?

- and can't you just use normal fft for the more complex stuff? 

The list you mention - and what makes Andrea mention list processing is the accumulation that would normally happen in fft, right? The fft window, yea?


Well not only overtones, just checking in which frequency you have energy and what kind of centroid you have .. stuff like that that allows you to collect frequency statistics of your signal instead than in the time domain. Can be useful as an extra tool. The normal FFT is audio rate, i need as control rate FFT that's what is different from the normal case.

that's my point, why *not* just perform audio-rate fft? Convert your control signal into audio and process it normally?

it doesn't work that way i think. You need data in format of audio frames with precise number of samples to know which frequency bins you're calculating.. i don't think you can just use sig...

it doesn't work that way i think. You need data in format of audio frames with precise number of samples to know which frequency bins you're calculating..  i don't think you can just use sig...


FFT from OSC data?