Hi, Just checked ffmpeg's libavcodec again, after Nick commited sync changes. Results are still bad :( VIDEO: [DIV3] 640x352 24bpp 23.98 fps 951.2 kbps (116.1 kbyte/s) Dshow: BENCHMARKs: V: 14.368s VO: 0.005s A: 0.000s Sys: 0.494s = 14.867s BENCHMARK%: V: 96.6455% VO: 0.0310% A: 0.0000% Sys: 3.3234% = 100.0000% libavcodec (mplayer cvs): BENCHMARKs: V: 12.834s VO: 0.007s A: 0.000s Sys: 0.507s = 13.347s BENCHMARK%: V: 96.1524% VO: 0.0524% A: 0.0000% Sys: 3.7952% = 100.0000% libavcodec (ffmpeg cvs): BENCHMARKs: V: 14.596s VO: 0.007s A: 0.000s Sys: 0.528s = 15.131s BENCHMARK%: V: 96.4649% VO: 0.0483% A: 0.0000% Sys: 3.4868% = 100.0000% isn't all optimizations synced yet, or it's a highlevel slowness problem of libavcodec (maybe the non-n*16 height movie fix?)? A'rpi / Astral & ESP-team -- mailto:arpi@thot.banki.hu http://esp-team.scene.hu _______________________________________________ Mplayer-dev-eng mailing list Mplayer-dev-eng@lists.sourceforge.net http://lists.sourceforge.net/lists/listinfo/mplayer-dev-eng
Hi,
Just checked ffmpeg's libavcodec again, after Nick commited sync changes.
Results are still bad :( found the reason (after some patching and checking different versions)
CPU test code is broken in ffmpeg CVS. just unrolled cpuid stuff changes, and it's fast again! before the cputest.c changes (-jun24): CPU flags: 0xB mmx mmxext sse after (jun25-) CPU flags: 0x0 so, this cpuid stuff should be fixed ASAP!!! A'rpi / Astral & ESP-team -- mailto:arpi@thot.banki.hu http://esp-team.scene.hu _______________________________________________ Mplayer-dev-eng mailing list Mplayer-dev-eng@lists.sourceforge.net http://lists.sourceforge.net/lists/listinfo/mplayer-dev-eng
participants (1)
-
Arpi