RE: [MPlayer-users] mplayer and low end machines
This is not hardware acceleration, it's hardware support for YUV colorspace, plus drivers that allow you to write directly to video memory rather than letting the slow broken xf86 drivers copy the image around and convert it multiple times before actually displaying it.
I thought this is what -vo sdl:dga was for. Isn't it faster than vidix? On my Geforce 256 card, for full-screen playback @ 640x480 using libav (default) for mpeg4 decoding, -vo xv seems to be the smoothest and sdl:dga is second, being just a little jerky. I'm using an AMD Athlon 1.2 Ghz. Shaun
On Thu, Nov 07, 2002 at 08:47:55AM -0600, Green, Shaun wrote:
[Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
This is not hardware acceleration, it's hardware support for YUV colorspace, plus drivers that allow you to write directly to video memory rather than letting the slow broken xf86 drivers copy the image around and convert it multiple times before actually displaying it.
I thought this is what -vo sdl:dga was for. Isn't it faster than vidix?
No, MUCH slower. Anything RGB-based is slow (non-xv SDL, x11, svgalib, dga, fbdev, ...). Anything that involves multiple copying or conversion steps is slow (x11, xv, X-based SDL). Those listed in both categories are doubly slow. The fast vo's are the direct access YUV-space ones (mga, xmga, vesa:vidix, xvidix, tdfxfb).
I'm using an AMD Athlon 1.2 Ghz.
Then it doesn't matter what vo you use. Rich
participants (2)
-
D Richard Felker III -
Green, Shaun