[FFmpeg-cvslog] r12758 - in trunk/libavcodec: aac_ac3_parser.c?aac_ac3_parser.h aac_parser.c ac3_parser.c

Michael Niedermayer michaelni
Mon Apr 7 12:58:16 CEST 2008


On Mon, Apr 07, 2008 at 12:59:05PM +0200, Bartlomiej Wolowiec wrote:
> On niedziela, 6 kwietnia 2008, Reimar D?ffinger wrote:
> > On Sat, Apr 05, 2008 at 11:53:40PM -0700, Mike Melanson wrote:
> > > bwolowiec wrote:
> > > > Author: bwolowiec
> > > > Date: Sat Apr  5 22:09:36 2008
> > > > New Revision: 12758
> > > >
> > > > Log:
> > > > change of aac_ac3_parser, so it is able to send complete portion of
> > > > data to decoder
> > >
> > > This change seems to have broken regression tests pertaining to seeking.
> > > Is anyone else seeing this?
> >
> > Yes, now seektest is broken for me as well. Seems I tested a slightly
> > outdated version last time...
> 
> Ok, I enclose correct version, after which in my opinion FFmpeg works right 
> (but tests still aren't passed). 
> 
> Here are some explanations about tests failure:
> -> in the case of  size of frames: in this file frames have size 558/556 and 
> distance between their beginnings is 570/568. Because of it, the old parser 
> returned size of frame and new parser is returning  data until the next frame 
> comes. I don't know if it's right or wrong behaviour, but it makes work of 
> the parser easier when in eac3 few frames should be returned. 
> ->The position of found frame is also changed because after normal ac3 frames 
> there can occur eac3 dependent frames. Because of that, the frame is returned 
> only when it finds the next one.
> 
> In the case of http://roundup.mplayerhq.hu/roundup/ffmpeg/issue401 , I don't 
> know what may  caused  it, I don't have this bug.
> 
> What should I do?

Explain what exactly your patch corrects.
One bug i see is that your sync functions are off by one byte, loosing 1 bytes
and thus returing 1 byte to late. (that is header_size+1)

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Asymptotically faster algorithms should always be preferred if you have
asymptotical amounts of data
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-cvslog/attachments/20080407/bb4e3dd3/attachment.pgp>



More information about the ffmpeg-cvslog mailing list