Investigating the object limit & poly limit
Posted: Tue Feb 21, 2012 7:28 pm
Greetings, everyone. I decided for the time being to turn my test map into a testing grounds for what the object count actually is, and what the poly count actually is, and whether and why this changes depending on the objects used, or the computer used. My first test is complete:
On my lower-end computer, I used a bunch of simple box props (six sided, so even if the collision makes the poly count double and it's triangles and not quads (not sure, I made the prop a while ago) the maximum number of polys for the prop should be 24) and tested various amounts to find the object limit as opposed to the poly limit. I found a limit, yes, and on my computer, anyway, it always made it lock up at the end of loading, rather than crashing to desktop. When I tested with roughly 1300 objects or more (in addition to some vehicles, 16 cps, and 32 troops) it froze up, but with slightly under 1200 it did not. If in fact that's 24 polys each, that would be only 28800 polys, which I believe is well under the poly limit. I will test this in the future by substituting both simpler and higher-poly models with the current box model.
On my lower-end computer, I used a bunch of simple box props (six sided, so even if the collision makes the poly count double and it's triangles and not quads (not sure, I made the prop a while ago) the maximum number of polys for the prop should be 24) and tested various amounts to find the object limit as opposed to the poly limit. I found a limit, yes, and on my computer, anyway, it always made it lock up at the end of loading, rather than crashing to desktop. When I tested with roughly 1300 objects or more (in addition to some vehicles, 16 cps, and 32 troops) it froze up, but with slightly under 1200 it did not. If in fact that's 24 polys each, that would be only 28800 polys, which I believe is well under the poly limit. I will test this in the future by substituting both simpler and higher-poly models with the current box model.