Russian version

MELP vocoder 600 bps

Common description

Low bit rate MELP (Mixed Excitation Linear Prediction) speech coder with rate 600 bps use classical parametric model of description of speech signals and vector quantization (VQ) of transmitted parameters. MELP model is implemented on the base of linear prediction filter with mixed excitation in several frequency bands.

Independent estimation of voiced/unvoiced decision in different frequency bands is providing more natural synthetic speech sounding then quality of classical LPC vocoders with binary voiced/unvoiced decision for single band.

Vocoder MELP 600 bps is recommended for very low bit rate digital channels and with scheme of error coding may be used for channels 800-1200 bps with great level of bit errors. (digital speech over HF channels)

[top] [main]

Implementation

MELP 600 bps vocoder is implemented as fixed point ANSI-C model for PC. Real-time implementation on the base of DSP TMS320C54x/C55x may be done under contract.

[top] [main]

Speech samples

Here you can test speech quality of MELP 600 bps algorithm by listening following speech samples:

Original speech

fspeech.wav

MELP vocoder 600 bps

melp1200.wav

[top] [main]

Navigation

Common description

Implementation

Speech samples


All technical and commercial questions about products, prices, terms of delivery, software libraries, source code etc. please send us by e-mail:

e-mail в графическом файле

Copyright © 1999-2009, DSP center SPb SUT