DSPRelated.com
Forums

statistical characteristics ACELP

Started by mehdi_m_5442 February 25, 2008
hi
can somebody please tell me about statistical characteristics a frame
speech when that coded by ACELP?
many thanks
Hi mehdi,

You should be more specific about your question, what do you mean by
"statistical characteristics" ? The ACELP differs from other CELP-like
algorithms by the way the pulses are partitioned, searched and
indexed, but has many things in common with other CELP algorithms.
If you provide more details about your concern you should probably
receive better answers.

Cheers

--- In s..., "mehdi_m_5442"
wrote:
>
> hi
> can somebody please tell me about statistical characteristics a frame
> speech when that coded by ACELP?
> many thanks
>
Hi Mehdi,
First of all, you should reply/comment to the messages through the
Group's forum, to let every one share the information, and also to get
different views/advices from group's members.

now, back to your questions. If the use of ACELP is not a requirement,
you could use simpler CELP codbook such Mutlipulse or Vector sum
codebooks.
Although, the ACELP is relatively more complex, and will take more
time to implement

Even if you plan to use the ACELP, you should start with simpler
codebooks such those mentionned above.

hope this helps,
----------------------------------

Hello , Mr Nacer

I have been started my final project for three month. My project is
about a vocoder scheme basis ACELP (similar G.723.1)(bit rate 5.67, lenghe fram0ms,Lp filter coefficient & bit). But I have
several problem . I hope you can help me. My problem are :

1. Can I use algebraic codebook G.723.1 or G.729 toward my scheme?

2. If I have document related to structure and search (algebraic
codebook), I can generat algebraic codebook?

Thank you in advance.

Best regards.
--- In s..., "nacer_ghoul"
wrote:
>
> Hi mehdi,
>
> You should be more specific about your question, what do you mean by
> "statistical characteristics" ? The ACELP differs from other CELP-like
> algorithms by the way the pulses are partitioned, searched and
> indexed, but has many things in common with other CELP algorithms.
> If you provide more details about your concern you should probably
> receive better answers.
>
> Cheers
>
> --- In s..., "mehdi_m_5442"
> wrote:
> >
> > hi
> > can somebody please tell me about statistical characteristics a frame
> > speech when that coded by ACELP?
> > many thanks
>