- Thu Sep 09, 2010 6:32 am
#330229
short story: have particle system within max emitting 'smoke'.
when doing initial simulation tests (by itself), rendering performance had no problem.
when doing animation test with object (to emit smoke), had no problem.
when the two were combined in scene... Problem. (each involving 'default sky dome') *(granted the base object-whale contains no key frames for this test)
i've tried just about everything i could think of, considering displacement has been turned off with and without motion blur. separating particle system intersection from object. hiding object(s) from global illumination, reflection, cast shadows, receive shadows, etc.
considering i exported the (tested) frame to .mxs and it still seemed to have performance issues when rendering directly from MXS, i figured i'd post it here opposed to the max plugin section.... hell, i even tried drinking a beer and that didn't help.
below are visual examples to clarify problem... would greatly appreciate any help, and or, feedback to this described problem. hopefully this make sense... thanks in advance!
here is particle setup. you'll see that i added "negative" frames... tried with frame count starting at zero and still slow.

object mesh with modifiers...

object eyes have mesh that use position script to attach to base object... (just trying to give as much info as possible here in case of error considering there was an internal max error earlier when position constraint could not locate vertices due to "re-smoothing" and completely botched a previously saved file that required re-applying a new skin wrap modifier to address the issue)

with everything as presumably planned. (the big problem seems to be during the .mxi writing process. even though i tried setting the .mxi write process other than "automatic" within the console... voxelization happens within a second)

with particle system emitter completely separated from object (granted, there is an emitter within the cigarette material, but even without emitter layer, results are same):

with just "smoke" emitting particle system:

with just mesh:

for all instances posted, here are render settings (although i've tried multiple variations and all containing both particle system and object seem to have the same performance issue):

here is the same frame exported to studio:

here is the result from rendering from studio. note that MXI had yet to be written, and voxelization happened at same rate.

hopefully i'm overlooking something simple.... either that, or i'm off my trolley.... btw, the materials are not that complex at all.
thanks in advance for any help. hopefully this makes sense............
when doing initial simulation tests (by itself), rendering performance had no problem.
when doing animation test with object (to emit smoke), had no problem.
when the two were combined in scene... Problem. (each involving 'default sky dome') *(granted the base object-whale contains no key frames for this test)
i've tried just about everything i could think of, considering displacement has been turned off with and without motion blur. separating particle system intersection from object. hiding object(s) from global illumination, reflection, cast shadows, receive shadows, etc.
considering i exported the (tested) frame to .mxs and it still seemed to have performance issues when rendering directly from MXS, i figured i'd post it here opposed to the max plugin section.... hell, i even tried drinking a beer and that didn't help.
below are visual examples to clarify problem... would greatly appreciate any help, and or, feedback to this described problem. hopefully this make sense... thanks in advance!
here is particle setup. you'll see that i added "negative" frames... tried with frame count starting at zero and still slow.

object mesh with modifiers...

object eyes have mesh that use position script to attach to base object... (just trying to give as much info as possible here in case of error considering there was an internal max error earlier when position constraint could not locate vertices due to "re-smoothing" and completely botched a previously saved file that required re-applying a new skin wrap modifier to address the issue)

with everything as presumably planned. (the big problem seems to be during the .mxi writing process. even though i tried setting the .mxi write process other than "automatic" within the console... voxelization happens within a second)
Code: Select all
Voxelization
[08/September/2010 17:56:24] Start Rendering
Start writing MXI file...
End writing MXI file...
[08/September/2010 18:22:14] Benchmark of 0.595. Time: 25m49s. SL of 1.00
[08/September/2010 18:22:14] Time left: 14h34m11s
Start writing MXI file...
End writing MXI file...
[08/September/2010 18:25:51] Benchmark of 0.603. Time: 29m27s. SL of 1.13

with particle system emitter completely separated from object (granted, there is an emitter within the cigarette material, but even without emitter layer, results are same):
Code: Select all
[08/September/2010 19:53:26] Start Voxelization
[08/September/2010 19:53:27] End Voxelization
[08/September/2010 19:53:27] Start Rendering
Start writing MXI file...
End writing MXI file...
[08/September/2010 20:16:45] Benchmark of 0.659. Time: 23m18s. SL of 1.00

with just "smoke" emitting particle system:
Code: Select all
[08/September/2010 21:25:58] Start Rendering
Start writing MXI file...
End writing MXI file...
[08/September/2010 21:26:17] Benchmark of 47.262. Time: 19s. SL of 1.00

with just mesh:
Code: Select all
[08/September/2010 21:40:44] Start Voxelization
[08/September/2010 21:40:44] End Voxelization
[08/September/2010 21:40:45] Start Rendering
Start writing MXI file...
End writing MXI file...
[08/September/2010 21:40:47] Benchmark of 368.640. Time: 2s. SL of 1.00

for all instances posted, here are render settings (although i've tried multiple variations and all containing both particle system and object seem to have the same performance issue):

here is the same frame exported to studio:

here is the result from rendering from studio. note that MXI had yet to be written, and voxelization happened at same rate.

hopefully i'm overlooking something simple.... either that, or i'm off my trolley.... btw, the materials are not that complex at all.
thanks in advance for any help. hopefully this makes sense............