07-19-2004, 08:35 AM
Have a look at this link . its a power point presentation that talks about bsp tree depth which is causing you all the memory problems.
also try the following
<LI class=MsoNormal style="COLOR: navy">If you’re using displacement, reduce the “max displace” value (in the renderer tab) as much as possible. If you reduce it too much, then your displacement will be clipped. However, if the value is too high, then more memory may be used. <LI class=MsoNormal style="COLOR: navy">Use the “conserver memory” option (in the processing tab). To our experience, this option generally results in _less memory usage_ AND _faster rendering_. It will optimize geometry, and it will free a lot of resources used by 3ds max which are not needed when rendering with mental ray. <LI class=MsoNormal style="COLOR: navy">Use the “placeholders” option, IF you’re not doing GI or final gather. This option will allow mental ray to flush geometry out of memory when it’s no longer needed, and rebuild it when it needs it back. It’s not a good idea to use this option along with GI because mental ray would probably be constantly destroying/rebuilding geometry, resulting in very long rendering times. With Final Gather, only use placeholders if you also specify a “Final Gather Falloff”. The FG Falloff can help avoid the constant destroying/rebuilding of geometry. <LI class=MsoNormal style="COLOR: navy">Use the “Large BSP” raytrace algorithm, IF you’re also using placeholders. The large BSP can benefit from flushing, while the normal BSP can’t. <LI class=MsoNormal style="COLOR: navy">Use the “Grid raytrace algorithm” if you’re NOT using placeholders, and if you have a lot of instancing in your scene. The Grid algorithm will benefit from instancing, whereas BSP will not.
Use raytraced shadows – shadow maps eat up memory (and are often slower in mental ray)
07-19-2004, 04:23 PM
thanks for info ...
I'm going to try it today ...
01-18-2006, 06:00 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.