[FFmpeg-devel] [RFC] special "broken DV" demuxer

Baptiste Coudurier baptiste.coudurier
Wed Mar 11 21:13:18 CET 2009


Reimar D?ffinger wrote:
> On Wed, Mar 11, 2009 at 12:19:32PM -0700, Baptiste Coudurier wrote:
>> Then I suggest we revert the support of spacejam*.dv until we find an
>> adequate solution. "pond.dv" has IMHO preference since it was working
>> before.
> 
> Maybe someone could start saying what an adequate solution would be! I
> proposed one I consider completely adequate (extra demuxer) and one I
> consider mostly adequate (except it needs to be re-checked against the
> spec).
> Also I can reason just as much that spacejam*.dv is spec-compliant while
> pond.dv is not, thus it should have preference.

I believe the adequate solution would be to indentify what exactly _is_
needed to be able to demux/decode and check for this information.

Definitely it seems the header is not mandatory to be able to decode, so
we should find another method.

The goal of FFmpeg has always to support as many files as we could, even
broken ones (we wouldn't play 50% of avi files otherwise), and be
resiliant against errors. I don't think we should stop supporting pond.dv.

>> Furthermore, I believe if we add a new demuxer for this case, we should
>> add a new demuxer for every format having a startcode and can be "in the
>> middle of a cd image". I find this definitely non reasonable.
> 
> There is no _need_ to add a new demuxer, it just is a really simple way
> to do it. And such hacks are not necessary for any other formats, DV is
> the only format where it is customary to support files that follow
> almost no part of the specs.

Who knows :)

[...]

-- 
Baptiste COUDURIER                              GnuPG Key Id: 0x5C1ABAAA
Key fingerprint                 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA
checking for life_signs in -lkenny... no
FFmpeg maintainer                                  http://www.ffmpeg.org




More information about the ffmpeg-devel mailing list