Return-Path: Received: from mail-vc0-f180.google.com ([209.85.220.180] verified) by media-motion.tv (CommuniGate Pro SMTP 4.2.10) with ESMTP-TLS id 5425785 for AE-List@media-motion.tv; Sun, 06 Apr 2014 21:05:59 +0200 Received: by mail-vc0-f180.google.com with SMTP id lf12so4707567vcb.11 for ; Sun, 06 Apr 2014 12:07:32 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; bh=cCffcc1RPdggjP44IOE9bzGjp7Cl10fA3Xk1gBL5m3Q=; b=UBF41UGl68OpreDaIdLrPJ35x73K12o6ooHGPq+1iWPAg3VhQTUV9cRTEkF++ztmxf Y2ym1CK+BWajC0ihhDqiH3NVDCvumbGK+BBAVrI2kWrKAAtSpaW7mFTFQGYXGELApcPY Ak0eMXCezdygkKP7hMlz1KSLOCprr6B3wiqbHiEqhk7nwiYOQFebCj3HFvAYJcWIc6od 6EWPMrk57Gnw4j+9fvsWQVZFVdB8knChFsj/B9bpE8xKJ5UvkR2hEfWFIbjyH9eLxsJr UxrW8AMgktM7R/NC13EGbApJ6CUifCFf1A0TKhmrvQ9pFcN+90jZ2D44kBPtO/m/JXse W1Ig== MIME-Version: 1.0 X-Received: by 10.53.1.69 with SMTP id be5mr1609188vdd.27.1396811252610; Sun, 06 Apr 2014 12:07:32 -0700 (PDT) Received: by 10.52.116.3 with HTTP; Sun, 6 Apr 2014 12:07:32 -0700 (PDT) In-Reply-To: References: Date: Sun, 6 Apr 2014 12:07:32 -0700 Message-ID: Subject: Re: [AE] Which output codecs are full 64bit? From: Darby Edelen To: After Effects Mail List Content-Type: multipart/alternative; boundary=001a1136015487c10a04f6647658 --001a1136015487c10a04f6647658 Content-Type: text/plain; charset=ISO-8859-1 I haven't fully considered this but I would wager that in general the part of the render where the frame is actually written to its output format would be miniscule relative to the actual rendering. I'd expect that any difference would become more noticeable the less actual rendering needs to be done. For example, if using After Effects to transcode then I'd expect a higher percentage of time per frame is spent writing the output format than rendering so any difference in the time it takes to write that output would be more noticeable. I've heard from at least one AE compadre that image sequences are generally read/written slower than movie files by AE (something about accessing one file vs. multiple files), but that could be out of date information. If you want to just test the speed of the output module then I'd recommend rendering unaffected footage. -Darby On Thu, Apr 3, 2014 at 3:59 PM, Teddy Gage wrote: > Looks like .tif sequences actually rendered slower on the benchmark than > the raw QT 422 exports, so it doesn't look like a bottleneck after all. > Something to consider if you render a lot of sequences. > -TG > > > On Thu, Apr 3, 2014 at 11:32 AM, Teddy Gage wrote: > >> So Stephen Van Vuuren ran the AfterBenchCCbenchmark and came up with a good question, one I had not considered. He >> wasn't sure whether the 8 bit 422 raw quicktime export I have set as >> default output in the benchmark was actually 64 bit enabled, as AE has had >> issues in the past with quicktime libraries. I thought those had been >> resolved but maybe not. >> >> So I have two questions: >> >> - Is there a better default codec to render to within AE? I wanted to >> avoid image sequences because the "render and replace usage" feature is >> used several times, and also because it's a bit more complicated for the >> end user. I was trying to keep things as idiot proof as possible. >> >> - What render codecs provide the least bottleneck in terms of output? >> i.e. What outputs are fully threaded and 64 bit enabled? Is the QT 8bit 422 >> sufficient? One would think that because the renders do seem to use 100% of >> available resources when exporting that it is, in fact, 64 bit. >> >> I am going to be doing some tests with image sequences and DNX export to >> see what happens >> -TG >> >> -- >> _____________________________ >> VFX & Motion Graphic Artist >> teddygage dot com >> > > > > -- > _____________________________ > VFX & Motion Graphic Artist > teddygage dot com > --001a1136015487c10a04f6647658 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable
I haven't fully considered this but I would wager= that in general the part of the render where the frame is actually written= to its output format would be miniscule relative to the actual rendering.= =A0 I'd expect that any difference would become more noticeable the les= s actual rendering needs to be done.=A0 For example, if using After Effects= to transcode then I'd expect a higher percentage of time per frame is = spent writing the output format than rendering so any difference in the tim= e it takes to write that output would be more noticeable.

I've heard from at least one AE compadre that image sequences are g= enerally read/written slower than movie files by AE (something about access= ing one file vs. multiple files), but that could be out of date information= .

If you want to just test the speed of the output module then= I'd recommend rendering unaffected footage.

-D= arby


On Thu, Apr 3, 2014 at 3:59 PM, Teddy Gage <teddygage@gmail.com><= /span> wrote:
Looks like .tif sequences actually rendered slower on the = benchmark than the raw QT 422 exports, so it doesn't look like a bottle= neck after all. Something to consider if you render a lot of sequences.=A0<= span class=3D"HOEnZb">
-TG


On Thu, Apr 3, 2014= at 11:32 AM, Teddy Gage <teddygage@gmail.com> wrote:
=A0 =A0 =A0So Ste= phen Van Vuuren ran the AfterBenchCC benchmark and came up with a good questi= on, one I had not considered. He wasn't sure whether the 8 bit 422 raw = quicktime export I have set as default output in the benchmark was actually= 64 bit enabled, as AE has had issues in the past with quicktime libraries.= I thought those had been resolved but maybe not.

So I have two questions:=A0

- Is th= ere a better default codec to render to within AE? I wanted to avoid image = sequences because the "render and replace usage" feature is used = several times, and also because it's a bit more complicated for the end= user. I was trying to keep things as idiot proof as possible.

- What render codecs provide the least bottleneck in te= rms of output? i.e. What outputs are fully threaded and 64 bit enabled? Is = the QT 8bit 422 sufficient? One would think that because the renders do see= m to use 100% of available resources when exporting that it is, in fact, 64= bit.

I am going to be doing some tests with image sequences = and DNX export to see what happens
-TG

=
--
_______________________= ______
VFX & M= otion Graphic Artist
teddygage dot com



--
=
_____________________________=
VFX & Motion Graphic Artist
teddygage dot com

--001a1136015487c10a04f6647658--