View Full Version : Does shape instance cause a massive viewport memory hit when off?

05 May 2008, 08:26 AM
Hi folks,

i've a pf setup with some find targets and it's using shape instances on animated meshes - I know that this is nasty in terms of rendering overheads but for viewport cacheing using display as dots it sees to race ahead memory wise - is there still a major hit purely for using the op as in does it still calculate the mesh per frame even if it isn't showing it?



05 May 2008, 07:39 PM
Point cache your mesh if you're not. Get rid of all keyframes possible.

05 May 2008, 08:09 PM
Yes it is still eating RAM. I believe that Cache will record any Ops that are active. That is why the Caching tools in Box3 (also available by themselves now) were released. They have the ability to exclude data from the cache and re-apply it after the fact.

05 May 2008, 09:11 PM
Cheers supreme - I'll give it a go alright, the mesh itself only has 6 keyframes set to loop (bird wings flapping) but I'll check if the cached version is lighter.

Charley - cheers for that, I bought box #3 a while back and I never found the cache tools really worked as well as I'd like - it'd still update or need to calculate operators even though it was baked - I'm not sure if it's better at this stage but I never found it as solid as top for example. That said you're far better on pflow than I am and I'd trust your word on it.


05 May 2008, 09:43 PM
After you cache the flow, you can turn off the events that are cached. That should resolve the issue of it re-calculating.

05 May 2008, 11:12 PM
Cheers charley - I'll give it another look.

05 May 2008, 01:27 PM
Yup, just got the latest ts builds of box #3 and it's working as advertised now - maybe a bit laggy in the viewports but a hell of a lot quicker than calculating the particles each frame I want to render!

CGTalk Moderation
05 May 2008, 01:27 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.