[MPlayer-users] Some issues using MPlayer dev-SVN-r22485

[Reimar Döffinger]

Hello,
On Mon, Mar 19, 2007 at 01:38:28PM -0400, Victor Farias wrote:
> 1) -really-quiet does not work correctly under Windows when that command is 
> used in the "config" file (it will work as expected if the command is used 
> in CMD).
> 
> This is displayed at the console when the -really-quiet command is used in 
> the "config" file:
> 
> MPlayer dev-SVN-r22485-3.4.5 (C) 2000-2007 MPlayer Team
> CPU: AMD Athlon(tm) 64 Processor 3200+ (Family: 15, Model: 31, Stepping: 0)
> CPUflags:  MMX: 1 MMX2: 1 3DNow: 1 3DNowExt: 1 SSE: 0 SSE2: 0
> Compiled for x86 CPU with extensions: MMX MMX2 3DNow 3DNowExt

No surprise, the config file is read only after this is printed, so it's
not really fixable for now. The MPLAYER_VERBOSE environment variable
might provide a good enough workaround though (no idea if it works on
windows though).

> 2) There is an issue when playing WavPack files since ffmpeg-svn-r7518. 
> This problem is explained in the file "click-sound.txt" that I uploaded to 
> MPlayer's FTP some time ago. An updated report about the said problem is 
> attached to this message.

I probably have a fix. Try again in a week or so.

[...]
> a) I was checking the logs from ffmpeg-cvslog and I noticed that kostya 
> added support for DT$. I did some tests using MPlayer (I modified the 
> "codecs.conf" file) and the DCA support is working really nice.
> 
> Could be possible to add DCA support from ffmpeg to MPlayer?

Should be available in SVN. Haven't tested though.

> b) According to the documents, you cannot use -vo gl:scaled-osd and -ass at 
> the same time (I think this is because some text "corruption" may happen 
> under some circunstances). However, I discovered that the said text 
> "corruption" does not occur when using -vf ass.

Of course not. The corruption is due to OpenGL rendering of the
subtitles, so it of course won't happen with -vf ass. You can still use
-vo gl:scaled-osd -ass if the pixel errors don't bother you. I didn't
consider it worth fixing so far, since any fix will probably mean that
rendering via OpenGL will be slower than just using -vf ass and thus
quite pointless.

> c) It seems that after Reimar's "subtitle code cleanup" patches, the 
> negative delay when displaying subtitles while using -ao dsound was fixed. 
> I did some tests and the behavior is almost the same that when using -ao 
> win32. Thanks for the fix ;)

Good to know. I suspected it might fix some of these problems but nobody
had confirmed it so far.

Greetings,
Reimar Döffinger