[vorbis] Compression Artifacts at -q 5. Help!

[Greg Wooledge]

Tyler Knott (vault-responce at charter.net) wrote:

> But wouldn't oggenc encode the exact same file the exact same way with 
> the exact same parameters every time? Couldn't we include a few test 
> files(10 to15 seconds, uncompressed, various sample rates/bits) with 
> some scripts to encode them and them compare the MD5sums to known good 
> copies?  Wouldn't this catch the vast majority of broken oggencs?

There is some permissible error due to floating point rounding.  This
means the files aren't guaranteed to be bit-for-bit identical between
platforms, compilers, etc.


-- 
Greg Wooledge                  |   "Truth belongs to everybody."
greg at wooledge.org              |    - The Red Hot Chili Peppers
http://wooledge.org/~greg/     |


-------------- next part --------------
A non-text attachment was scrubbed...
Name: part
Type: application/pgp-signature
Size: 188 bytes
Desc: not available
Url : http://lists.xiph.org/pipermail/vorbis/attachments/20030616/901d28f6/part.pgp