John Villar
2005-Mar-07 14:09 UTC
[Speex-dev] Differences between ACELP and CELP.... Reloaded
And the main point of interest is the following: how much difference in SPEEX terms would it be if it were developed with ACELP in mind? (not that i would do an aSpeex codec implementing ACELP, just to know and point my efforts) -- John Villar Gerente de Proyectos Computadores Flor Hard Soft 2058 C.A. www.florhard.com -------------- next part -------------- A non-text attachment was scrubbed... Name: john.villar.vcf Type: text/x-vcard Size: 396 bytes Desc: not available Url : http://lists.xiph.org/pipermail/speex-dev/attachments/20050307/2e513fc3/john.villar.vcf
Jean-Marc Valin
2005-Mar-11 12:02 UTC
[Speex-dev] Differences between ACELP and CELP.... Reloaded
Hi, The A in ACELP stands for algebraic. It means that the fixed codebook (cb_search.c) is not "real" codebook (as contained in the *table.c files), but rather an algebraic (implicit) one, where each entry can be computed directly from the codebook index. The main advantage of ACELP lies in the fact that the search is much faster, which allows larger codebook for the same complexity. In terms of Speex, implementing ACELP would mean replacing cb_search.c and most likely changing a bit how the fixed codebook gain is implemented too. Not that I would recommend it anyway, but since you asked... Jean-Marc Le lundi 07 mars 2005 ? 18:10 +0000, John Villar a ?crit :> And the main point of interest is the following: how much difference in > SPEEX terms would it be if it were developed with ACELP in mind? (not > that i would do an aSpeex codec implementing ACELP, just to know and > point my efforts) > > _______________________________________________ > Speex-dev mailing list > Speex-dev@xiph.org > http://lists.xiph.org/mailman/listinfo/speex-dev-- Jean-Marc Valin <Jean-Marc.Valin@USherbrooke.ca> Universit? de Sherbrooke