temposcal — Phase-locked vocoder processing with onset detection/processing, 'tempo-scaling'.
temposcal implements phase-locked vocoder processing using function tables containing sampled-sound sources, with GEN01, and temposcal will accept deferred allocation tables.
This opcode allows for time and frequency-independent scaling. Time is advanced internally, but controlled by a tempo scaling parameter; when an onset is detected, timescaling is momentarily stopped to avoid smearing of attacks. The quality of the effect is generally improved with phase locking switched on.
temposcal will also scale pitch, independently of frequency, using a transposition factor (k-rate).
ifftsize -- FFT size (power-of-two), defaults to 2048.
idecim -- decimation, defaults to 4 (meaning hopsize = fftsize/4)
idbthresh -- threshold based on dB power spectrum ratio between two successive windows. A detected ratio above it will cancel timescaling momentarily, to avoid smearing (defaults to 1)
ktimescal -- timescaling ratio, < 1 stretch, > 1 contract.
kamp -- amplitude scaling
kpitch -- grain pitch scaling (1=normal pitch, < 1 lower, > 1 higher; negative, backwards)
klock -- 0 or 1, to switch phase-locking on/off
ktab -- source signal function table. Deferred-allocation tables (see GEN01) are accepted, but the opcode expects a mono source. Tables can be switched at k-rate.