<html><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; "><br><div><div>On 24/08/2008, at 7:12 PM, Paul Kendall wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite"><div>On Sunday 24 August 2008 17:46:28 Steven Ellis wrote:<br><blockquote type="cite">On 24/08/2008, at 5:10 PM, Steven Ellis wrote:</blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite">My current patch to libs/libmythtv/avformatdecoder.cpp around line<br></blockquote><blockquote type="cite">1588 is as follows<br></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite"> if (CODEC_ID_H264 == enc->codec_id)<br></blockquote><blockquote type="cite"> {<br></blockquote><blockquote type="cite"> VERBOSE(VB_PLAYBACK, LOC + "Turning off loop<br></blockquote><blockquote type="cite">filter");<br></blockquote><blockquote type="cite"> enc->skip_loop_filter = AVDISCARD_ALL;<br></blockquote><blockquote type="cite"> VERBOSE(VB_PLAYBACK, LOC + "Enabling Fast decode<br></blockquote><blockquote type="cite">mode");<br></blockquote><blockquote type="cite"> enc->flags |= CODEC_FLAG2_FAST;<br></blockquote><blockquote type="cite"> }<br></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite">This appears to give me a 4-5 % boost on the SD channels, and possibly<br></blockquote><blockquote type="cite">15-30% on the HD channels. It is hard to tell on TV One HD at the<br></blockquote><blockquote type="cite">moment as the bitrate for the sports is all over the place. TV3 is<br></blockquote><blockquote type="cite">almost watch-able on my CPU so hopefully Graeme will get full support<br></blockquote><blockquote type="cite">on his 5600 X2. Watching TV2 now sits at an average of 70-80% rather<br></blockquote><blockquote type="cite">than close to 100%.<br></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite">Paul - I'd love to get any additional patches you have to stabilise<br></blockquote><blockquote type="cite">the audio. I haven't had time to play with debug builds to trace the<br></blockquote><blockquote type="cite">audio segfaults yet.<br></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite">Steve<br></blockquote><blockquote type="cite"><br></blockquote><br>The main problem with the audio is that it seems the sse2 version of <br>float_to_int16_interleave_sse2 causes the segfault. What I did to fix this was <br>to go the end of libavcodec/i386/dsputil_mmx.c change the assignment<br>c->float_to_int16_interleave = float_to_int16_interleave_sse2;<br>on line 2864 to<br>c->float_to_int16_interleave = float_to_int16_interleave_sse;<br><br>That fixed the audio for me. YMMV.<br><br><br></div></blockquote><br></div><div>Thanks Paul for that tip. It does appear to be a more stable if I use the SSE call.</div><div><br></div><div>Steve</div><div><br></div><br><div> <span class="Apple-style-span" style="border-collapse: separate; color: rgb(0, 0, 0); font-family: Helvetica; font-size: 12px; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; line-height: normal; orphans: 2; text-align: auto; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-border-horizontal-spacing: 0px; -webkit-border-vertical-spacing: 0px; -webkit-text-decorations-in-effect: none; -webkit-text-size-adjust: auto; -webkit-text-stroke-width: 0; "><div style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; "><div>Steven Ellis - Technical Director<br>OpenMedia Limited<br>email - <a href="mailto:steven@openmedia.co.nz">steven@openmedia.co.nz</a><br>website - <a href="http://www.openmedia.co.nz/">http://www.openmedia.co.nz</a><br></div></div></span> </div><br></body></html>