MPLAYER poorformance (FloX's issue moved from other forum) (DOSX)
FloX wrote:
> juhu! Now it works? but bad things:
> * it is very slow
> * doesn?t support sound blaster live
> * doesn?t switch back to text mode
> So QV is still first choice?
I wrote:
> > slow
> Known issue A native DOS player would be faster ?
> > doesn?t support sound blaster live
> Maybe with HX 2.20 or 3.0
> > doesn?t switch back to text
> Works for me. Bugged VESA ?
> > QV is still first choice
> No OGG Theora
Japheth wrote:
> > doesn?t support sound blaster live
> I also have a SB Live. Since this card is SB16 compatible, it works with mplayer.
> > Known issue A native DOS player would be faster ?
> Yes, may be 1% faster. I doubt this will change the impression from "slow" to
> "wow, super fast". Things which REALLY fasten things: 1. use VESAMTRR, 2. don?t
> load the mouse driver (reading video memory is VERY slow), 3. don?t run
> mplayer fullscreen.
-------------------------------------------------------------
> Since this card is SB16 compatible, it works with mplayer
SB Live (PCI IIRC ? ) compatible (by hardware ? ) with SB16 (ISA ? ) ? I always considered this as a lie coming from "Creative"
> may be 1% faster
Is the "loss" caused by OGL emulation via VESA that negligible ?
> 2. don?t load the mouse driver
Doesn't work ... no effect or up to cca 2% (!!!) speedup at best
> 3. don?t run mplayer fullscreen.
Very big effect ... where (MP or HX ?) and how ( NN / linear / bilinear / Lancosz / ... ) is the scaling done ?
> 1. use VESAMTRR
Almost dummy ... up to 5% speedup ...
From my recent tests MP wastes cca:
- 60% of the time in the scaling of "-fs"
- 30% in the codecs
- 10% by writing into LFB
So, very "easy" solution: don't use "-fs" or speed it up / port the/some codecs to ASM and optimize them "correctly"
But there is still some mystery here ... transcoding to a less CPU hoggy codec does not help much either ... and nor reducing bpp ... maybe an odd timer bug ?
---
This is a LOGITECH mouse driver, but some software expect here
the following string:*** This is Copyright 1983 Microsoft ***
Complete thread:
- MPLAYER poorformance (FloX's issue moved from other forum) - DOS386, 13.08.2007, 06:52 (DOSX)
- MPLAYER poorformance (FloX's issue moved from other forum) - Rugxulo, 13.08.2007, 13:34
- [ testers needed ] MPLAYER poorformance | Timing BUG :-( - DOS386, 11.09.2007, 06:43
- MPLAYER compiler with D**PP ... but - DOS386, 25.10.2007, 07:29
- MPLAYER compiler with D**PP ... but - rr, 25.10.2007, 10:39
- MPLAYER compiler with D**PP ... but - Khusraw, 25.10.2007, 14:43
- MPLAYER compiler with D**PP ... but - DOS386, 26.10.2007, 08:39
- MPLAYER compiler with D**PP ... but - rr, 26.10.2007, 11:14
- MPLAYER compiled ....... emulating shots - DOS386, 28.10.2007, 02:49
- MPLAYER compiled ....... emulating shots - rr, 28.10.2007, 17:49
- MPLAYER compiled ....... emulating shots - DOS386, 15.12.2007, 18:48
- MPLAYER compiled ....... emulating shots - rr, 15.12.2007, 21:18
- MPLAYER compiled ....... emulating shots - DOS386, 15.12.2007, 18:48
- MPLAYER compiled ....... emulating shots - rr, 28.10.2007, 17:49
- MPLAYER compiled ....... emulating shots - DOS386, 28.10.2007, 02:49
- MPLAYER compiler with D**PP ... but - Laaca_, 26.10.2007, 16:47
- MPLAYER compiler with D**PP ... but - Deniska, 27.10.2007, 10:27
- MPLAYER compiler with D**PP ... but - rr, 26.10.2007, 11:14
- MPLAYER compiler with D**PP ... but - DOS386, 26.10.2007, 08:39
- MPLAYER poorformance (FloX's issue moved from other forum) - Rugxulo, 13.08.2007, 13:34