>From: Eric Scheirer <eds@media.mit.edu>
>
>Juhana K Kouhia wrote:
>>
>> Since it is possible to load samples in to Saol as MPEG audio files
>> for saving bandwidth, I examined if MPEG layer 3 audio is able to store
>> samples properly. The answer seems to be *negative*!
>
>Note that, for compliant operation, only MPEG-4 audio coders
>may be used to compress samples. The current list does not
>include MPEG-1/2 Layer 3, as MPEG-AAC gives much better performance.
>
>I'm afraid I can't help with this, as it's much more a software
>problem based on MP3 tools than a Structured Audio problem.
The delay in the beginning of the decoded wave could be in the standard.
At least somebody owning the standard specs should look at it.
The AAC seems to be part of MPEG2 and not MPEG4 according to some
MPEG webpages.
Well, in my opinion, the problem is deep in MPEG M Layer N systems.
Perhaps they designed the standard for audio streams but didn't realize
that most people use the encoder for audiofiles.
With streams, it doesn't matter if there is a delay in the beginning, or a
fraction of the end is lost, or if an implemetor doesn't figure out that
the output stream continues even after the input stream is ended
(because specs doesn't say it, perhaps).
We can always do something about it now, or then -- as you seem to suggest
-- wait until MPEG 4 is finished and find out the problems afterwards.
Yours,
Juhana
This archive was generated by hypermail 2b29 : Wed May 10 2000 - 12:14:10 EDT