Return-Path: Received: from mail-ie0-f169.google.com ([209.85.223.169] verified) by media-motion.tv (CommuniGate Pro SMTP 4.2.10) with ESMTP-TLS id 4923818 for AE-List@media-motion.tv; Thu, 13 Dec 2012 01:10:14 +0100 Received: by mail-ie0-f169.google.com with SMTP id c14so3012525ieb.28 for ; Wed, 12 Dec 2012 16:14:23 -0800 (PST) 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=b3YigKmimCBbbuMYH2dQmXNQtoZ8XqwHtDzQXkIyWlM=; b=tdWEUqHuzCXO4YVio8I2Y3BzGLMPIPn8ecC8lWMqbumkkbIQnYjIfjk1CRvnh18Z4l wga66C/mEYNg2y6J1ODPfCmT6gkJF1g1Hs9wzsmFkGgqXWNAMI1ToJbIq4SJJHxyYDya 6nH840DEzbt13AdAWPeOF5g5EkiqhKKqxuZvfET5LgBsXzxDaVb0ZqowyCBiwSf09nYx KL8gDM0Kq0vHbohNCFsEoEoMibjUyn2N+hUAujkSFcxOMC3HapuHCb3c1auZd0CPR8zg 8h8kS3bP1t+EbMLXGPzC6b57dkz8X19MjD1dI2dsUQT2H5XxCM08wxd4s3BBza9Y80Qi OVmg== MIME-Version: 1.0 Received: by 10.50.42.169 with SMTP id p9mr2654640igl.17.1355357663168; Wed, 12 Dec 2012 16:14:23 -0800 (PST) Received: by 10.42.150.69 with HTTP; Wed, 12 Dec 2012 16:14:23 -0800 (PST) In-Reply-To: References: Date: Wed, 12 Dec 2012 16:14:23 -0800 Message-ID: Subject: Re: [AE] best system optimization for rotobrush? From: Todd Kopriva To: After Effects Mail List Content-Type: multipart/alternative; boundary=14dae93410b90e88eb04d0b0cc4d --14dae93410b90e88eb04d0b0cc4d Content-Type: text/plain; charset=ISO-8859-1 I had some folks look at this thread, and the response was that it seems that you have a pretty good handle on what's going on. On Tuesday, December 4, 2012, Teddy Gage wrote: > I am definitely aware of that, however is it possible to freeze a single > segment while you work on a different segment with a new base frame on the > same layer? When I tried that it simply froze the whole layer. I suppose I > could split the layer and keep going but that gets complicated. In my > experience on a hex core machine (with 12 ht cpus) and MP on, I havent run > into any issues with bugs or crashing; however it seems to propagate a bit > faster with MP off. But I can't prove it, it just feels faster. Maybe I'll > run a quick test if I have time. Would appreciate if anyone had extra info > on how roto data is processed... > > On Tue, Dec 4, 2012 at 5:08 PM, Chris Meyer > > wrote: > >> I think it would run afoul of MP in some situations, as previous >> information needs to be propagated frame by frame from the Base Frame (or >> last-calculated frame) to the current frame in order to process it. This >> can be sped up by finishing your Segmentation Boundary work and then >> clicking the Freeze button in the Layer panel to render and lock those >> calculations. >> >> - Chris >> >> >> On Dec 4, 2012, at 1:57 PM, Teddy Gage wrote: >> >> Very quickly; I'm wondering whether having MP on or off is better for >> working in rotobrush, and whether or not having HT turned off would help >> speed up propagation, so that I can allocate more RAM per core. Any other >> settings that may help (ie hardware accelerating? does it make any >> difference?) would be great to know. One thing I've realized recently is >> that it's very important to have several base frames. I was using a single >> segment for a whole sequence and it would take forever to propagate on file >> load or unfreezing. Thanks, >> TG >> >> -- >> Animator & Editor >> www.teddygage.com >> Brooklyn >> >> >> > > > -- > Animator & Editor > www.teddygage.com > Brooklyn > > -- -------------------------------- Todd Kopriva 1510 E Alder St Seattle, WA 98122 (206)437-2296 -------------------------------- "Ultimately, all knowledge is self-knowledge." - Bruce Lee --14dae93410b90e88eb04d0b0cc4d Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable I had some folks look at this thread, and the response was that it seems th= at you have a pretty good handle on what's going on.
On Tuesday, December 4, 2012, Teddy Gage wrote:
I am definitely aware of that, however is it possible to freeze a single se= gment while you work on a different segment with a new base frame on the sa= me layer? When I tried that it simply froze the whole layer. I suppose I co= uld split the layer and keep going but that gets complicated. In my experie= nce on a hex core machine (with 12 ht cpus) and MP on, I havent run into an= y issues with bugs or crashing; however it seems to propagate a bit faster = with MP off. But I can't prove it, it just feels faster. Maybe I'll= run a quick test if I have time. Would appreciate if anyone had extra info= on how roto data is processed...

On Tue, Dec 4, 2012 at 5:08 PM, Chris Meyer = <chris@crishdesign.com><= /span> wrote:
I think it would run afoul of MP in som= e situations, as previous information needs to be propagated frame by frame= from the Base Frame (or last-calculated frame) to the current frame in ord= er to process it. This can be sped up by finishing your Segmentation Bounda= ry work and then clicking the Freeze button in the Layer panel to render an= d lock those calculations.

=A0- Chris

=
On Dec 4, 2012, at 1:57 PM, Teddy Gage wrote:

Very quickly; I'm wondering whether having MP on or= off is=20 better for working in rotobrush, and whether or not having HT=20 turned off would help speed up propagation, so that I can allocate more=20 RAM per core. Any other settings that may help (ie hardware=20 accelerating? does it make any difference?) would be great to know. One thi= ng I've realized recently is that it's very important to have sever= al base frames. I was using a single segment for a whole sequence and it wo= uld take forever to propagate on file load or unfreezing. Thanks,
TG

--
Animator & Editor
www.tedd= ygage.com
Brooklyn





--
Animator & Editor
<= span style>www.teddy= gage.com
Brooklyn



--
--------------------------------
Todd Kopriv= a
1510 E Alder St
Seattle, WA 98122
(206)437-2296
-------------= -------------------
"Ultimately, all knowledge is self-knowledge.&q= uot;
=A0=A0 - Bruce Lee
--14dae93410b90e88eb04d0b0cc4d--