On 8/24/06, <b class="gmail_sendername">Michael T. Dean</b> <<a href="mailto:mtdean@thirdcontact.com">mtdean@thirdcontact.com</a>> wrote:<div><span class="gmail_quote"></span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
On 08/24/2006 12:11 PM, Dylan Semler wrote:<br><br>> According to the Nvidia driver's readme, the GeForce4 Ti models' XvMC<br>> don't support "IDCT," only "motion-compensation." While I'm not sure
<br>> what either of those terms mean, I believe that mine is a Ti model and<br>> am hoping that IDCT is an integral part of processing HD streams. So<br>> I'm eyeing a 256MB 5200 for under $50.<br><br><br>Both Motion Compensation (MC) and inverse Discrete Cosine Transform
<br>(iDCT) must occur when decoding an MPEG-2 video stream. So, if you're<br>using XvMC and your card doesn't support iDCT, your CPU does it.</blockquote><div><br>Hey that's great to hear! <br></div><br><br>So I realize that the cpu consumption results that I previously gave were a bit biased, as mythtv was deinterlacing while mplayer was not. So I just tested it again with mythtv not deinterlacing. At 1x, processor usage is still peaked around 95%-99%, however if I slow it down to
0.5x, and only at 0.5x, it finally plays smoothly and cpu percentage is in the high 80s, low 90s. Going up to 0.55x will produce a prebuffering pause once every maybe 7 - 10 seconds.<br><br>I just tried running myth with XvMC and cpu is about 94% - 97% at
1.0x I can't do much testing because the computer usually hangs after a few seconds and requires a hard reboot.<br><br>Anyways thanks for the help<br><br><br>Cheers,<br>Dylan<br></div>