Sign in
Search speechcoding
Subscribe to speechcoding
speechcoding by Keywords
ACELP | ADPCM | AMBE | AMR | AMR-NB | CELP | Codebook | DTMF | G.723 | G.726 | G.729 | GSM | Interpolation | LPC | LSF | LSP | MELP | PCM | Perceptual | Pitch | PSOL | QCELP | Quantization | SMV | VAD | Vocoder
Discussion Groups
Discussion Groups | Speech Coding | G729 generation of fixed codebook
Technical discussions related to Speech Coding (all itu and other vocoders, ACELP, CELP, AMR, etc)
G729 generation of fixed codebook - Author Unknown - Nov 29 12:59:00 2001
|
Hi all, How can we find the appropriate positions(m0,m1,m2,m3) for placing the unit pulses and its sign(+/-) in the fixed codebook vector? I mean " c(n) = s 0d(n - m 0 ) + s 1d(n - m 1 ) + s 2d(n - m 2 ) + s 3d(n - m 3 ) n = 0,...,39 The codebook vector c(n) is constructed by taking a zero vector of dimension 40, and putting the four unit pulses at the found locations, multiplied with their corresponding sign:" in Section 3.8 (G729.pdf, Eq45,Pg 24) thanks in advance John |
|
|
(You need to be a member of speechcoding -- send a blank email to speechcoding-subscribe@yahoogroups.com )
Re: G729 generation of fixed codebook - karthik kannan - Nov 30 8:52:00 2001
|
hi John, The first of all codebook structure is algebraic structure and is based on ISSP(interleaved single pulse permutation). Basic priciple is to find the combination of 4 positions and signs from those 5 tracks and to pick out the one which yields the maximum criteria. The criteria can be found out in the itu doc. the algo used to find out is that of data structure algo. It is depth first tree search algo. if you are looking out for this algo, i can explain to you in the next mail. Hope i answered what you are looking for? Regards karthik kannan. -- On Thu, 29 Nov 2001 12:59:40 code_maniac2001 wrote: >Hi all, >How can we find the appropriate positions(m0,m1,m2,m3) for placing >the unit pulses and its sign(+/-) in the fixed codebook vector? >I mean > " c(n) = s 0d(n - m 0 ) + s 1d(n - m 1 ) + s 2d(n - m 2 ) + s 3d(n - > m 3 ) n = 0,...,39 >The codebook vector c(n) is constructed by taking a zero vector of >dimension 40, and putting the four unit pulses at the >found locations, multiplied with their corresponding sign:" > >in Section 3.8 (G729.pdf, Eq45,Pg 24) > >thanks in advance >John |
(You need to be a member of speechcoding -- send a blank email to speechcoding-subscribe@yahoogroups.com )