How's the so called "spectrum matching" done in plug-ins?
Is it frequency- or time-domain processing?
Spectrum matching techniques
-
- KVRian
- Topic Starter
- 1097 posts since 28 May, 2010 from Finland
-
- KVRAF
- 15275 posts since 8 Mar, 2005 from Utrecht, Holland
My naieve implementation would be to analyse source & target spectra with FFT, then apply a 64-band graphical EQ to the target with settings derived from the analysis.
We are the KVR collective. Resistance is futile. You will be assimilated. 
My MusicCalc is served over https!!
My MusicCalc is served over https!!
-
- KVRian
- 1050 posts since 6 May, 2008 from Poland
If I'm not mistaken I think it's based on FFTs of both sources, probably from several FFTs being summed together for a single source, some smoothing done, division of the spectrum of the reference with the target and then a finite impulse response designed from that curve and of course convolution with that FIR. At least that's how I'd do it, not every developer enjoys convolving FIRs as much as I do.
