<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Verdana
}
--></style>
</head>
<body class='hmmessage'>
<br><br>> Date: Sun, 21 Mar 2010 18:21:03 -0400<br>> From: tterribe@email.unc.edu<br>> <br>> Id Kong wrote:<br>> > The problem with starting with a parallel decoder is that decoding is<br>> > much easier. Correct me if I'm wrong but decoding requires much less<br>> <br>> That depends on resolution and bitrate, of course. 400 Mbps 2160p files<br>> are quite difficult to decode in real time (with current hardware). But<br>> the point was that a good bit of the work needed to make a parallel<br>> decoder would also be required to make a parallel encoder anyway, so<br>> it's a good first step. Especially as the code is much simpler.<div><br></div><div><span class="Apple-tab-span" style="white-space:pre"> </span>...but of course, decoding 400 Mbps 2160p video is cheaper than encoding that same quality of video!</div><div><span class="Apple-tab-span" style="white-space:pre"> </span>The point is that encoding is more processor intensive than decoding and thus a better test. It could be that decoding such high resolution video is memory bound rather than processor bound...</div><div><br><br></div><div>> > not an outrageous idea but I think it will be better to leave threading<br>> > to the client and support parallelism through the API much like Vorbis.<br>> <br>> This is non-trivial to export, and there is a cost to doing it that way.<br>> There are too many complex data dependencies to make exporting<br>> within-frame parallelism a good idea. However, once you go to<br>> multi-frame parallelism, you have to start changing the way rate control<br>> works (and worse, _exposing_ how rate control works to the client,<br>> because you may now have to dump frames that are in the process of being<br>> encoded and start over, e.g., when you decide to insert a keyframe).<br>> This could maybe be done for unconstrained two-pass with reasonably<br>> minimal pain, but as I've already said, that's not the interesting case.<br></div><div><br></div><div><span class="Apple-tab-span" style="white-space:pre"> </span>I just wanted to check whether it were feasible to export blocks of work like Vorbis. I'll take your word for it that it is not possible since I don't currently understand the source code for the encoder...</div><div><br></div> <br /><hr />Take your contacts everywhere. <a href='http://go.microsoft.com/?linkid=9712958' target='_new'>Try Messenger for mobile</a></body>
</html>