[MEncoder-users] segfault when using "probesize"

Nico Sabbi nicola_sabbi at fastwebnet.it
Sun Dec 10 12:02:30 CET 2006


Benjamin Adler wrote:

>Hello!
>
>I'm using
>
>mencoder dvb://"NDR FS HH" -o out.mpg -of mpeg -mpegopts format=dvd -demuxer
>lavf -lavfdopts probesize=256 -oac copy -ovc copy -endpos 00:30:00
>
>to record DVB-C to mpeg files on linux. I was told by Nico to use
>
>	-of mpeg -mpegopts format=dvd -demuxer lavf -lavfopts probesize=128
>
>because I had problems with a/v-sync when using 
>
>-of mpeg -mpegopts 'format=dvd:tsaf:drop:vaspect=16/9' -vf harddup.
>
>The sync became better, but not perfect. I tried increasing the probesize
>and it seems when using a probesize of at least 256, the sync becomes
>perfect.
>  
>
this sounds nonsense to me: probesize only affects the duration of
the probing phase in demux_lavf, not the syncronization

>The problem is, the higher the probesize, the more likely mencoder is to
>crash. As I schedule my recordings, this often means that a show is not
>recorded at all.
>
>This is an example output of the above (first) command:
>
>MEncoder dev-SVN-r20171-4.1.1 (C) 2000-2006 MPlayer Team
>CPU: AMD Athlon(tm) processor (Family: 6, Model: 4, Stepping: 4)
>CPUflags: Type: 6 MMX: 1 MMX2: 1 3DNow: 1 3DNow2: 1 SSE: 0 SSE2: 0
>Compiled with runtime CPU detection.
>dvb_tune Freq: 386000000
>success: format: 29  data: 0x0 - 0x0
>libavformat file format detected.
>Segmentation fault
>
>- Why does the probesize parameter affect a/v-sync so much. What exactly
>does this parameter do? The manpage is not very helpful here ("This value
>identifies the maximum number of TS packets to scan during the detection
>phase")
>  
>
as above

>- Why does mencoder crash, how can I (help to) fix it? Should I recompile
>with debug information and post a backtrace?
>  
>

yes




More information about the MEncoder-users mailing list