I don’t know if this is the right category…
because the work I do, many times I have to place large amount of persons (mesh) in order to validate a design.
Blender do a great Job on this. But Rhino gets slow and the files fire up till 1Giga.
There is any technique in Rhino to manipulate the meshes like blender does?
thanks in advance
Carlos.
Cheers
Jonas
ArchVision’s Rich Photorealistic Content (RPC) files may be smaller than the meshes.
Thank you for the tips Jonish, I will have a look.
Thank Andrew_nowicki.
best
C.
As @Jonishsay blocks will help a lot. It all depends a bit on how you have organised your meshes. If you have separate parts that repeat a lot you should create blocks out of them and instance them. That will decrease at least memory usage.
For what it’s worth, I can work with 2Gb files in Rhino 6 with no noticeable graphics lag. You should have no problem with files half that size.
Dan
Probably a question for @jeff - do blocks still require 4 Modelview matrix transformations for each object within the block? If so, blocks may not be a good way of increasing display speed.
In Addition to what others have suggested:
If the people meshes do not need separate materials: join them into one mesh instead of all separate mesh objects.
EDIT: That won’t have as big of a impact in V6 as it did in V5 with respect to display performance.
Maybe try to reduce the meshes and save as a separate version to have course representations of the people.
-Willem
Block are still a little slower than regular geometry in V6. There are more optimizations to do in that area, but I wouldn’t steer clear from blocks just based on this statement. They should still be pretty fast in V6.
That won’t have as big of a impact in V6 as it did in V5 with respect to display performance.
Thanks for the heads-up!
As Steve pointed out, V6 is a lot better at handling blocks so I would not purposely try staying away from them in V6. As for the number of transforms per object per block…I haven’t looked at that very closely…but many things in V6 are now cached, so getting any kind of “per object” statistics isn’t as straight forward as it is in V5. As such, blocks now use their own draw lists as well, so their overall impact on the display pipeline flow has been significantly reduced. Now, trying to draw a curve by “blocking” several thousand line segments (which some imports can do) is still not going to be very optimal…certainly nowhere near as optimal as just joining the segments into a single curve object. Use blocks, use them everywhere and anywhere you want…just use them wisely. 
-Jeff
That type of object is still on the list for optimization. If I can recognize this situation, the display code can be made smart enough to just draw at a similar speed as a polyline
I want to make an amendment to what I said above. As I begin installing V6 I’ve noticed that old computers with a 1Gb graphics card do not handle large files much better than V5. I’ve updated to the latest drivers, but the performance increase is still not as dramatic as what I experienced with my 4Gb graphics card.
We will be updating (or more likely, replacing) those old systems. I will be doing more installations today, so I’ll see how it works with 2Gb cards.
Dan
Edit: 2Gb cards work just fine. There is significant display improvements over V5 which are apparent at this level of card. With old 1Gb cards, not so much.
I would venture to guess that the memory space on the card may only be half the story. Older cards just don’t handle some of the newer OpenGL features that we implemented in V6.
I think this is a good excuse for some hardware updates!
Let me know if you get some Volta-based Nvidia GPUs. I’ll have to build the Cycles kernels for them still…
The new cards will most likely be Nvidia quadro P2000.