[Mplayer-users] Benchmarks

Felix Bьnemann atmosfear at users.sourceforge.net
Tue Jul 10 11:29:43 CEST 2001


Hello Nick,

Tuesday, July 10, 2001, 12:07:50 PM, you wrote:

NK> Hello, Arpi!

NK> Probably I have a stupid questions but IMHO mplayer needs RTFM benchmark.
NK> Yestoday I've done some minor optimization of native ffmpeg-divx codec.
NK> These results I get:

NK> Before:
NK> BENCHMARKs: V:   9.406s VO:   1.421s A:   0.600s Sys:  34.003s =   45.429s
NK> BENCHMARK%: V: 20.7046% VO:  3.1270% A:  1.3208% Sys: 74.8477% = 100.0000%

NK> After:
NK> BENCHMARKs: V:   9.280s VO:   1.397s A:   0.593s Sys:  34.153s =   45.424s
NK> BENCHMARK%: V: 20.4299% VO:  3.0761% A:  1.3063% Sys: 75.1877% = 100.0000%

NK> Could you explain me: what's difference between first string and second?
NK> Why Systime always 100%? What means 's' suffix in the first string.
it's total percentage and always 100% of course, systime includes also
idle calls of course so if other percentages go down, systime will go
up (at least if ya play with sound). s suffic imo means seconds.
NK> Why VO is changed too? Where AO time? Does 'V' include 'VO'?
VO time is time dneeded by vo plugin to display image (there can also
be extra systime spend for it not shown in there) and V doesn't
include VO, think AO isn't calculated cause audio device cpu time is
very minimal.

NK> Thanks!

NK> Best regards! Nick


-- 
Best regards,
 Felix                            mailto:atmosfear at users.sourceforge.net



_______________________________________________
Mplayer-users mailing list
Mplayer-users at lists.sourceforge.net
http://lists.sourceforge.net/lists/listinfo/mplayer-users



More information about the MPlayer-users mailing list