ogg is similiar to twinvq? Options

[jianxin yan]

hi all
just now, i scan over a brief specification about ogg rc1.0;
i find it is very similiar to Twinvq: for example,
1. use lpc to get spectral envelope and two size mdct;
2. use the envelope to flatten mdct coef. ;
3. use VQ for lsp and mdct residue;


am i right?

could anyone give me a correct answer. thanks.


regards.
yan.

[Garf]

Originally posted by jianxin yan
hi all
just now, i scan over a brief specification about ogg rc1.0;
i find it is very similiar to Twinvq: for example,
1. use lpc to get spectral envelope and two size mdct;
2. use the envelope to flatten mdct coef. ;
3. use VQ for lsp and mdct residue;


am i right?

could anyone give me a correct answer. thanks.


regards.
yan.

Since I don't know the exact workings of VQF (have a link?) I can't comment much, but Monty said the similarity between Ogg and VQF pretty much only consisted of the fact that they both use Vector Quantization. I guess there might be other similarities too, but then again MP3 and Ogg also have things in common.

Note that Ogg no longer uses LSP/LPC, but piecewise linear approximation.

--
GCP

[jianxin yan]

hi
because the ogg document is not whole which i had seen, maybe i can't get through ogg principle;

could you give me more SITE for ogg coding main idea?
i will look into the encoder later.

thanks.
YAN

[Garf]

Originally posted by jianxin yan
hi
because the ogg document is not whole which i had seen, maybe i can't get through ogg principle;

could you give me more SITE for ogg coding main idea?
i will look into the encoder later.

thanks.
YAN

I don't know of any that are fully up to date.

--
GCP

[PatchWorKs]

Vorbis Illuminated by Mike Coleman:

http://www.mathdogs.com/vorbis-illuminated/

[Garf]

Originally posted by PatchWorKs
Vorbis Illuminated by Mike Coleman:

http://www.mathdogs.com/vorbis-illuminated/

That one is outdated.

(It's probably enough to get you started though)

--
GCP