Quote:
Originally Posted by IXix
It seems some guys have just solved a fifty year old problem. I only vaguely understand what this is about but it sounds kind of exciting. I figure since it's related to FFT/IFFT it might have some relevance for our audio mangling activities.
https://www.nature.com/articles/s41598-019-50234-9 |
Interesting stuff. I just very quickly read parts of that. I probably got it wrong but it looks like they figured out a fast way to make the inverse transform of a transform that could work more closely to how humans hear frequencies. The (discrete) Fourier transform's pain point has always been that so much of the bandwidth of the calculations is dedicated to the uppermost frequencies which are not really so important musically and all kinds of tricks have been needed to get things working better in the lower frequencies. (You could always increase the FFT size to get a better low frequency resolution, but that would decrease the time resolution and increase latency for real time applications.)
__________________
I am no longer part of the REAPER community. Please don't contact me with any REAPER-related issues.
Last edited by Xenakios; 10-13-2019 at 12:10 PM.
|
Linear Mode
