Here is a paper by Miller Puckette on an efficient algorithm which makes use of the FFT

“An Efficient Algorithm for the Calculation of a Constant Q Transform”

http://www.wellesley.edu/Physics/brown/pubs/effalgV92P2698-P2701.pdf

Here’s another paper on an efficient algorithm:

“Constant-Q transform toolbox for music processing”

http://www.elec.qmul.ac.uk/people/anssik/cqt/

If there is already a Max object for this, or if this is a problem which has already been solved, please let me know. Otherwise, any help in pointing me in the right direction as to how this ought to be best implemented would be awesome.

]]>http://wwwmath.uni-muenster.de/logik/Personen/blankertz/constQ/constQ.pdf

I haven’t read it that closely, but I think it wouldn’t be rocket surgery to port that matlab implementation into a max object written in C. Or gen~, perhaps.

]]>I’ve been digging around to attempt a log spread fft freq output and much of the maths is quite over my head. After finding the “constant q transform” method, it seems like a great potential solution ( as opposed to simply reallocating linear bins into log bins as a post ftt process ).

C

]]>