[MPlayer-dev-eng] ADPCM Document

Mike Melanson melanson at pcisys.net
Sun Dec 30 18:42:22 CET 2001


On Sun, 30 Dec 2001, Arpi wrote:

> avi: there are lots of docs flying around on teh net, but most of them are
> inaccurate, miss important details or has bad information.
> 
> asf: m$ released specification of ASF v2.0 format. it'd detailed enough, and
> is nice, the only problem with it: nobody uses this file format.
> all products use asf v1.0, which is not documented at all. (and differs from 2.0)
> (except that rev.eng doc by avifile author i sent you yesterday)
> 
> vivo: not documented at all. i did the reverse engineering of file format.
> anyway, it was very easy. it's like rle-encoded stuff with interleaved
> audio-video chunks.

	I'm thinking I had better revise my assertion on the CRI that "The
internals of the most common multimedia file formats are well-documented."
Maybe it should be, "The internals of QT and RM are fairly
well-documented, but not entirely. Everything else, good luck!" Come to
think of it, I haven't seen a revision of the QT spec that mentions
compressed headers (though I haven't looked in awhile). If they still
don't mention that in the latest document, that would be a good candidate
for our QT-quirks document.

	As an aside, why on earth did Apple feel it necessary to employ
compressed headers? Honestly, I have some QT trailers that are 50
megabytes large, but they managed to save a few bytes by compressing a
header that, at worst, is a few 10s of kilobytes large. It just seems
negligible by comparison.

--
	-Mike Melanson




More information about the MPlayer-dev-eng mailing list