I need to figure out this FFT stuff. I've read a pretty good wave theory text book, and understand the concept (i think?) Now, how does this apply to PD?

1. The FFT is realtime fourier analysis right? What limitations does this have and how is it done?

2. What is the specific convention used by FFT objects within PD, and what are common ways to work with that data?

Any outside resources that you can point me at would be great, especially if they're pd specific. I want to get deeper into DSP, and the time-domain has begun to bore me...

Thanks,
--N--

I hear you there, later on I'd like to experiment with building a shaper/granulator that would work by sequencing frames of spectra (way down the road for me at this point).

I don't know if you've checked Miller Puckette's book already but he talks about the differences between the different FFT objects available:
http://www-crca.ucsd.edu/~msp/techniques/latest/book-html/node179.html#12898

I don't understand all of the math myself; I didn't have any real DSP coursework in college other than a primer course that touched briefly on Fourier theory

dspguide.com might not have your answers, but it should sufficiently boggle you up to know what your answers might taste like.

Frank Barknecht wrote a good tutorial for FTT objects in PD:

http://footils.org/cms/show/60

Marius Schebella did it too:

http://www.parasitaere-kapazitaeten.net/Pd/fft_und_pd_en.htm

cheers,

Alberto

Ah! here I am going through the same trouble. I get the gist what the FFT does, but I don't understand how it's represented as magnitude in the pd window. I'm reading the tutorial in loadbang, but of course he assumes he knows enough math the understand what he's going on about.

It would be awesome to directly extract the information from a complex sound with this....

actually I've been wondering to myself weather you could use the FFT to extract the notes from a strummed guitar chord sent directly into the adac.

This might not be too useful for actually using PD and learning all of the applications of FFT, but if you do want to know more of the math behind FFT you can read through this book: https://ccrma.stanford.edu/~jos/mdft/mdft.html

It only assumes that you have a strong knowledge of algebra and maybe a little calculus. I'm reading through the book now, and although I do have a mathematical background up to first year calculus, I do find it extremely confusing at times. Still, it's worth checking out.