reconstructVocal
reconstructVocal(
mixtureStft,eqCurves,sr,nFft?,hopLength?):Float32Array<ArrayBufferLike>
Defined in: packages/pleco-xa/src/scripts/xa-vocal-separation.js:566
Reconstruct a vocal ESTIMATE from EQ curves learned against a reference stem (supervised — not blind separation; for that use decompose.hpss / softmask / nn_filter)
Parameters
Section titled “Parameters”mixtureStft
Section titled “mixtureStft”any[]
Mixture STFT in (freq x time) format
eqCurves
Section titled “eqCurves”number[][]
EQ curves for each window
number
Sample rate
number = 2048
FFT size
hopLength?
Section titled “hopLength?”number = 1024
Hop length
Returns
Section titled “Returns”Float32Array<ArrayBufferLike>
Reconstructed audio