First person view (bigger pic) Top view (bigger pic)
SCALABLE PERCEPTUAL MIXING AND FILTERING OF AUDIO SIGNALSMixing and equalization application:
USING AN AUGMENTED SPECTRAL REPRESENTATIONN. Tsingos
REVES - INRIA Sophia AntipolisAll example files are .wav (PCM int 16 bits or IEEE 32-bit floating point format)
Reference mix
Mix calculated with masking evaluation and all 4096 FFT coefficients/frame
Mix calculated with masking evaluation and up to 2000 FFT coefficients
Mix calculated with masking evaluation and up to 500 FFT coefficients
Mix calculated with masking evaluation and up to 500 FFT coefficients
using an optimized encoding of the input data.
Block FIR reverberation
application:
Rotor example (original sound)
Song example (original sound)
Reference processingProcessing with masking turned on and all FFT coeffficients
Processing with masking turned on and 50000 coeffficients
Reference processingProcessing with masking turned on and all FFT coeffficients
Processing with masking turned on and 50000 coeffficients
Voice example
(original sound)
Reference processingProcessing with masking turned on and all FFT coeffficients
Processing with masking turned on and 50000 coeffficients
3D audio application:
1000 sound sources (the red spheres in the pictures) spatialized in real-time !
click here for an example movie file (.avi divx 5.2.1 format ~12 MBytes)
First person view (bigger
pic)
Top view (bigger pic)
Reference processingProcessing with masking turned on and 80000 coeffficients