topic Crashing when rendering, help! in 3ds Max Forum

Crashing when rendering, help!

Anonymous — Mon, 16 May 2016 22:04:22 GMT

I have a huge project I'm trying to render (save file is 1Gb), the render will start but always comes back with errors (running out of memory) about 30mins in. I have over 20mil polygons in my file & 1250 objects. The models were imported from autodesk inventor.

I can't upload any images unfortunately as the project is confidential but any suggestions would be appreciated!

My computer specs are:
Intel i7 3.4ghz
30Gb Ram
Nvidia quadro K2100M

Re: Crashing when rendering, help!

silentboof — Mon, 16 May 2016 22:16:29 GMT

there is a magic called "instance"
if you have mutiple objects similar to each other; let's say some trees for example don't leave them just as they are! delete them and try to instance them n place'em again in the scene. the same story is about materials. believe me when it comes to thousands of objects, instancing them could be a huge help in reducing proccess and load times.
hope that helps

Re: Crashing when rendering, help!

Anonymous — Tue, 17 May 2016 12:50:03 GMT

Is this render an animation or a hi-res still frame?

Which renderer are you using?

Re: Crashing when rendering, help!

silentboof — Tue, 17 May 2016 13:34:40 GMT

I'm usually into animations and stuff.Still frames not much.
about the engine, well I use more then 2 engines!
flat diffuse channel (color n texture) = scanline
lighting = mentalray ( just a simple white material)
zdepth ( to get DOF) = again scanline
AO = vray or mental....mostly vray cuz it's **** fast in AO comparing to other engines.
and then boom .... I composite them all in afterFX 🙂

Re: Crashing when rendering, help!

Anonymous — Tue, 17 May 2016 23:29:18 GMT

Thank for the input

I used instancing as much as i could.

Im rendering with mental ray

I got it to work by making a new scene and xrefing the scene the original scene with cameras and lighting. Used batch render and sent it to backburner to render frames out.

Its working successfully so far using 14gb of ram instead of 24gb