Return-Path: Received: from mout.perfora.net ([74.208.4.194] verified) by media-motion.tv (CommuniGate Pro SMTP 4.2.10) with ESMTP-TLS id 5345027 for AE-List@media-motion.tv; Mon, 13 Jan 2014 06:19:11 +0100 Received: from [192.168.0.5] (CPE-72-135-117-181.new.res.rr.com [72.135.117.181]) by mrelay.perfora.net (node=mrus2) with ESMTP (Nemesis) id 0LeeIQ-1VacDZ1xWQ-00prK8; Mon, 13 Jan 2014 00:22:56 -0500 Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 (Mac OS X Mail 7.1 \(1827\)) Subject: Re: [AE] a very important question for every After Effects user From: Steve Oakley In-Reply-To: Date: Sun, 12 Jan 2014 23:22:54 -0600 Content-Transfer-Encoding: quoted-printable Message-Id: References: To: After Effects Mail List X-Mailer: Apple Mail (2.1827) X-Provags-ID: V02:K0:dUHC8bb7m/8oooZlMNGwkJ0K88k0S4MwK92PAwmw8dH 5yUgN2duKow2JJyF2VjctrRVcf+eSRpeRahRQH8fxCqujUQx2s X6PPPzh6WM9+lesbnPoe9XU4zIelx+ml+PcUjpR6H6fiPv4HW8 M5FQftWDaC2dDJ4qGov0EIurZlckHopRVer7Vn6A2D8q0qJemZ pbbuvTydtjaeBe08SxKVu9F8zIDqhdM7kAdbP6xJUkebjI2LL+ Qm006YrZUTIFxCaSv82WLkZQZVY/jpEmWAqTrhLptsubzbo53F Jo5L7G4TKAVn8BhOK9BMhEG/wuRTWhm+bGJxpVY/SrVJyszMwX 3v+sHYU/q/8px3AF2ZPNIgkLKZHE3BGv3/JSZXezn5Nsm1DPnW ko/RaKJuugRNg== lets be specific then. there are certain jobs that are simply labor intense like roto. = Rotobrush can sometimes do the job, but far more often its mocha pro for = me. thats fine. however there are other places I feel the lack of speed : starting up AE = self renders, not being able to simply copy and paste clips from PP to = CC - yes you can do it if you get the right mode. I can't just select a = clip in PP and paste into AE project window. simple thing but time = saver.=20 rendering : the big evil. why can't we have more seamless network = rendering. Ok, I get bandwidth issues on the network, but when my laptop = has 4+4 it should just show up as available network cores to render. AE = should evaluate the network speed, files, time per frame based on = previous render where it tracked per frame render times, and just figure = out what the best config is. if it figures spreading things out onto = network cores is good because its long per frame CPU times, great, do = it. if not, TELL ME THAT, and just render on the local machine because = that will be faster.... but why are we rendering at all ? there are a lot of things I used to do in AE I now do in PP because they = are RT thanks to CUDA. I'm saving a HUGE amount of time, never mind hit = play and see the results for a lot of stuff with 5-10 layers or more. = thats what I need in AE. I need hit play and see it go. not render, = render, render, render, render, ok, coffee cold yet ? render, render, = ok, now I'll play. No CUDA fast RT results. all the basic filters and = transforms should render on the GPU and the CPU should be last resort. the AE raytracer is a great start... but is it going to get faster and = better ? speed just so much matters that I'd say if you could deliver a CUDA GPU = rendering ver of AE and use 100% of resources to do that in the next = cycle, I'd be all in. AE has great features already and there have been = a lot of small good ones added along the way. what we desperately need = is RT feedback when working in 3D scenes including the raytracer. = waiting 30-60 secs to see a frame in 1/2 or 1/4 res in HD just doesn't = fly anymore. instead of thinking small, think big ! how about mostly ditching CPU = rendering period ! all GPU all the time. I get sick when I see what = video games do and how amazingly great they are on the same hardware. = think video game performance levels. S On Jan 12, 2014, at 10:58 PM, Todd Kopriva wrote: >=20 > Making things faster than they currently are takes research and = development and testing. Those tasks would be done by the same people = who would be doing research and development and testing on new creative = features. >=20 > The question is how much of our finite resources you want us to spend = on making things faster versus adding new creative features, new = formats, etc. >=20 > For After Effects CS6 and CC, we spent the lesser fraction = (considerably less than half) of our resources on making things faster. = For the near future, the question on the table is whether we crank that = up to the greater fraction (more than half) or even to nearly the entire = team.