Forum Replies Created

Page 4 of 8
  • Aaron Pozzer

    March 25, 2015 at 11:36 pm in reply to: Tracer ONLY on Mograph Selection

    yup, that seems to do the trick. had to fiddle with it a bit, but i got it working. thanks for the suggestion.

  • Aaron Pozzer

    March 25, 2015 at 11:07 pm in reply to: Tracer ONLY on Mograph Selection

    hmmm, interesting. the problem is the random part of it. i need some degree of control over which clones get the tracer/hair, hence why i was wondering if it could be done with a specific selection. unless i can use the selection tag to somehow limit the random effector.

  • Aaron Pozzer

    March 17, 2015 at 7:50 pm in reply to: Texture from random video frame

    thanks for the response. yup, this is an option as i said in the original post, but i was wondering more from an ‘is it possible’ standpoint. being able to do it without loading 200 images into multishader would be alot cleaner.

  • Aaron Pozzer

    February 4, 2015 at 11:31 pm in reply to: New Build, Slow Multithread render

    this particular project is realy complicated. plenty of fx, layers, pre comps, etc. There is no footage, only image layers. a few of the precomps are 4k, which im sure isnt helping. it also seemed to get alot worse when i stacked up 2 copies of lens distortion.

    its running off SSD drives all the way, no HDDs. for normal stuff its usually fine, usage wise, but i noticed this performance issue with this large project in question and i wondered why it was so. some googling showed me its a common issue with AE. its annoying. hopefully the rewrite DOES fix it.

  • Aaron Pozzer

    February 1, 2015 at 2:37 pm in reply to: New Build, Slow Multithread render

    thanks Walter, i was going to try that after posting yesterday, but had to leave, so i did this morning. settings ended up giving me 3 cpus w/6gb each. that seems to have helped. render went from 12:20 on the old machine, to 7:30 yeterday, to 3:30. so, yesterday was definitely a lack of ram problem. overall render speed has definitely increased, but…

    now here is another annoying problem, and i dont think this one is solvable. this is a very heavy scene, so when i load it up, i drop my playhead over a very complex part of the timeline to see how fast it loads up all the layers. again, watching task manager, it sits there with 1 thread working, bouncing around 10% cpu, slowly filling up the ram to about 18gb. clearly this is a heavy scene, but why is it only using 1 cpu at 10% !?!? it took like 45 seconds to draw 1 frame at half res. like, ok, use 1 cpu, but at least use all its power! whats up this this?

  • Aaron Pozzer

    January 19, 2015 at 4:54 pm in reply to: New System Build. i7 vs Xeon: need avdice

    no prob walter, i appreciate the info either way. going back thru that video posted above for some ideas on parts. seems it was sponsored by nzxt, haha. probably not what i would buy, but i can find equivalents.

  • Aaron Pozzer

    January 19, 2015 at 2:53 pm in reply to: New System Build. i7 vs Xeon: need avdice

    thanks Steve, i saw this a couple weeks ago. initially my intent was just to update my existing amd box, but when i saw this, i realized i could get a much better build for not too much cash, parring back a bit from what dave had built. thanks for reminding me of it though! ill have another watch thru.

  • Aaron Pozzer

    January 19, 2015 at 2:26 pm in reply to: New System Build. i7 vs Xeon: need avdice

    thanks Walter, i was hoping id get a response from you.

    good points, and thanks for clearing up some of my questions. could you just clarify your one point about “but will probably have periods, especially during interactions, where they are less multithreaded“.

    i would not imaging them to be fully MT’d at all times. like during use, navigating viewport, or setting keyframes, etc, would all be normal use, and they would only start using all cores at render time.

    current build plan was for 32gb of ram. i suppose i could step it up to 48 or 64 by adding more sticks. i have a few SSD’s, but they’re probably out of date at this point (ocz vertex 3) and was looking at getting 1 newer one as a boot/apps drive. there was some M.2 samsung drive everyone seemed to be talking about a while ago, but i cant seem to find it on sale at my local stores. XP941? they have other M.2 drives, but they aren’t as “name brand” as samsung.

    so after some consideration on the whole thing, i think the best dollar/performance option is to get the i7 5930k and OC it a bit passed 4ghz. i read some people are getting stably up to 4.6, and while i wouldn’t want to push it that far and iv never been one for OCing, it seems pretty easy these days to squeeze a bit more juice out of these chips.

    i would love to hear any parts recommendations you (or anyone else) have overall. like i said above, current plans is for an Asus x99 XPower. if there are better boards, im all ears! i also havent nailed down my choice of ram yet. everyone seems to love G SKillz, though my initial pricing lead me to corsair or crucial. cooling, air or liquid? iv always used air though i know liquid is very popular these days.

    thanks again! the insight is much appreciated.

  • Aaron Pozzer

    June 20, 2014 at 2:07 pm in reply to: keyframe path, change/update all keys

    this would do it, though i was hoping there was just some native AE option i wasnt aware of.

  • Aaron Pozzer

    June 19, 2014 at 11:10 pm in reply to: keyframe path, change/update all keys

    thanks for the reply dave, but its not quite what i was after. maybe my wording on the problem wasnt precise enough. i meant a single point on a mask, not the entire mask. so if 1 point is wrong, but the 10 other points making up the mask are correct over time, i could somehow select that 1 point and adjust it… whatever…. 10pixels to the left… for all keyframes.

Page 4 of 8

We use anonymous cookies to give you the best experience we can.
Our Privacy policy | GDPR Policy