Jump to content
Sign in to follow this  
undream

AE Rendertimes and Myths

Recommended Posts

Ok. Here is a topic about getting faster render times in AE.

I will start this topic with ym question:

 

I'm working on a project right now with 50+ layers (around 20 adjuıstment) and as you can guess that takes hell long to render. (3+ hours for a 30 sec. clip, HD)

 

How can I get faster rendertimes without changing the "general outlook" of my movie ?

 

And yes, I have a lot of motion blurs and frame blends.... Pain.

 

 

 

Suggestions ?

 

 

P.S. Also I heard about Gridiron Nucleo Pro, does it really help ? Or any better alternatives to what it does ?

Share this post


Link to post
Share on other sites

I'd say 3 hours wasn't that long to be honest with a complex setup with motion blur. Are you on CS4 cos as far as I'm aware nucleo pro hasn't been sorted for that yet.

 

One thing you could do is play with the motion blur settings and take them down a bit but you're going to get into way longer render times than a few hours. What I'd do is leave it til evening, set it going and watch a movie, or leave it going overnight (and hope the sheep doesn't wake you if it fails)

Share this post


Link to post
Share on other sites

How can I get faster rendertimes without changing the "general outlook" of my movie ?

 

And yes, I have a lot of motion blurs and frame blends.... Pain.

 

In that case most likely not. Temporal sampling operations are quite linear and thus hard to parallelize. Even if MP rendering works, the individual threads/ processing steps within the instances may execute in strict order. 3 hours doesn't sound too bad, though. Without actually knowing the project it's hard to advise, anyway.

 

Mylenium

Share this post


Link to post
Share on other sites

As of yesterday, that is no longer true. ;-)

 

Mylenium

 

Oh really? Amazing news. Going to check that out immediately. I just got paid too, this will make an upcoming project much less stressful I think

Share this post


Link to post
Share on other sites

I'd like to know how one can speed up working in the view port? I've got an 8 core mac (2009) with the standard graphics card (nvidia i think). I've been working in C4D since I got it and its been a dream but now I'm doing my first AE project on it and it feels very very laggy just doing basic 3D stuff :S

Share this post


Link to post
Share on other sites

I'd like to know how one can speed up working in the view port? I've got an 8 core mac (2009) with the standard graphics card (nvidia i think). I've been working in C4D since I got it and its been a dream but now I'm doing my first AE project on it and it feels very very laggy just doing basic 3D stuff :S

 

Well, 3D applications like C4D default to draft shading modes so it's no wonder they are very speedy at that. Consider what it takes to render a frame at high quality, and then imagine trying to interact with C4D in that way :)

 

For speeding AE interactions/Comp panel updates, the best trick IMO (which can be automatic in CS4) is syncing view magnification and resolution (set resolution to "Auto"), in such a way that when view magnification is 50 per cent, resolution is half (and thus AE only cooks half of the pixels). Consider that a 1080 HD comp, when set to 50 per cent/half, will give you a pixel perfect 960x540 frame. This is larger than a widescreen SD comp. Should be enough for doing most of the job. Memory usage will be half as much, and updates/renders/previews will often take half as much (some tasks are computationally intensive, independently from frame size).

 

The OpenGL - Interactive mode can also give speedy, draft-like interactions. But the problem is, a couple of non-OpenGL accelerated operations (nearly all plug-ins and a handful of AE's own effects) is going to make OpenGL a moot point.

 

Render Multiple Frames Simulataneously can drasitcally speed up renders and RAM previews, but not Comp panel updates.

Share this post


Link to post
Share on other sites

Well, 3D applications like C4D default to draft shading modes so it's no wonder they are very speedy at that. Consider what it takes to render a frame at high quality, and then imagine trying to interact with C4D in that way :)

 

For speeding AE interactions/Comp panel updates, the best trick IMO (which can be automatic in CS4) is syncing view magnification and resolution (set resolution to "Auto"), in such a way that when view magnification is 50 per cent, resolution is half (and thus AE only cooks half of the pixels). Consider that a 1080 HD comp, when set to 50 per cent/half, will give you a pixel perfect 960x540 frame. This is larger than a widescreen SD comp. Should be enough for doing most of the job. Memory usage will be half as much, and updates/renders/previews will often take half as much (some tasks are computationally intensive, independently from frame size).

 

The OpenGL - Interactive mode can also give speedy, draft-like interactions. But the problem is, a couple of non-OpenGL accelerated operations (nearly all plug-ins and a handful of AE's own effects) is going to make OpenGL a moot point.

 

Render Multiple Frames Simulataneously can drasitcally speed up renders and RAM previews, but not Comp panel updates.

 

Thanks Adolfo. I already run most things at 50% with the CS4 auto function which helps but my top of the range mac still feels way slower in the view port compared to my dell workstation at the office. Will upgrading the 3D card make a difference or should I go for more RAM?

Share this post


Link to post
Share on other sites

Thanks Adolfo. I already run most things at 50% with the CS4 auto function which helps but my top of the range mac still feels way slower in the view port compared to my dell workstation at the office. Will upgrading the 3D card make a difference or should I go for more RAM?

 

There is a performance gap, mostly related to many, many years of Intel-specific optimizations in the OS, compilers, etc. compared to just a few years since the Mac OS runs on the Intel architecture.

But I don't think it's (or should be) as significant as you are experiencing. I myself am Mac-based, and in fact do most of my production work on a MBP and it feels very responsive.

The whole OpenGL/graphics card subject is a delicate one. It really depends on the nature of your comps. If you use third party plug-ins all the time, or you are a fan of certain non-accelerated (or "non-accelerable") built-in effects, then the best graphics card in the world won't help. For many of the built-in, accelerated features, a good card can provide smoother/more responsive interactions, with draft-quality rendering. A good graphics card should keep interacting nicely when adjusting things like depth of field, light/shadows, blend modes, motion blur, etc. And all this, even for nested Comps. But again, if there are non-accelerated plug-ins or effects in the processing chain, it won't be as rewarding (if at all). As of CS4, OpenGL interactions also can use dynamic resolution, so even for lowlier cards (or complex Comps), it can still do something usable in many cases.

 

At the same time, I can't blame people who decide to turn it off for good. Driver conflicts, especially on the Windows side, are really painful. If it affect stability or reliability, turning it off is good advice.

 

RAM is always extremely important.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Sign in to follow this  

×
×
  • Create New...