PDA

View Full Version : PFlow+MRay Render Errors!!


ExP
10-30-2008, 05:10 AM
Hi everybody,

I'm working on a scene with a character that emits particles as he walks (simple tiny spheres). For some reason I get this random fluttering little black boxes in the render over the particles area around the character. I'm using Max9, Particle Flow, and Mental Ray. At first I though it was an issue of the cache op in PFlow, but I've tried any possible combination and the problem's still there.

Any suggestion?? This is driving me crazy...

Here's a little sequence of frames. I've rendered the alpha channel so it shows the problem more clearly.

Thanks!

http://experimentalusa.com/PFlowErrors.jpg

3DMadness
10-30-2008, 08:19 PM
This not a particle flow problem, this is the famous "black bucket" that appears in mental ray when you hit a memory limit. The best way to get rid of the problem is upgrading to xp and max 64-bit. If this is not an option, you have to try some way to optimize your scene to reduce the memory its using.

depleteD
10-31-2008, 01:07 AM
In the processing tab, you can increase the memory limit, but as 3dmadness said switch to 64 bit. Blur switched this year, and its amazing, you can push a lot more through, especially with particles , I did some limit testing and I got up well over 10 gigs of memory usage and no crash.... ****ing awesome shit dude.

Also you could switch to mitchell filtering and/or lower your bucket size. That might help....

ExP
10-31-2008, 02:18 AM
Hey guys thanks for the reply,

I'm actually already on XP64 and Max64. We HAD to switch a year and a half ago since we were doing a lot of huge crowd work... that saved our ass!!
But that's why I though it was really strange I was getting those errors. I have the MR memory limit set pretty high and I'm already using Mitchell filtering... plus it's not that many particles...
I'm net rendering on our farm so I used the PFlow cache op to save the entire particle anim with the file... could that be the problem?
Also I'm using some pretty heavy raytraced area lights to cast shadows on the particles with 3D motion blur on... you guys think that might be an overkill even for Max64?

depleteD, I'll definitely try the smaller bucket size and maybe even the MR conserve memory thing. That might do the trick.

Cheers

depleteD
10-31-2008, 02:22 AM
Hmmmm, my best advice is thinking particles. heh.

Uh I know that conserve memory thing should never be checked, thats hwo ive been advised. Do you have cache test resutls unchecked?

ExP
10-31-2008, 02:29 AM
yeah that's what I've heard too on the conserve memory... it tends to do more harm than good. That's why I didn't have it on in the beginning. I'll run a couple of tests and double check the "cache test results."

TP huh? I know, I should have switched already hehe... although we're implementing XSI in our pipeline and were thinking of handling particle work with ICE... but that's another story... especially now...

depleteD
10-31-2008, 02:41 AM
The last thing I can recomend is run diagnostic renders and push your bsp settings around maybe, good luck man, tech issues like this are the worst.

We have xsi for animation and rigging, steven caron has been showing me some stuff in ICE and ive been really impressed with some of the stuff ive seen, guys are codeing in fluid solvers straight from the node graph. The only thing thats holding me back from switching is that xsi doesnt do to well with building geometry on the fly, which is a problem when it comes to fragmentation and metaballs, steven would know more about that.

I can recommend TP Fume Final Render and Realflow, I can get everything and anything done with those tools with super fast turn around time. Shit works pretty smooth.

But everyone works differently, I like it, hopefully it gives you some ideas.

Good luck man, let me know if you find a solution to that droped bucket.

Strang
10-31-2008, 03:02 AM
@ExP

ICE in XSI is great, but keep in mind it is version 1.0 so there is a bit of pioneering right now. fortunately the dev team is very open to suggestions and really want to push ICE to the entire application, lets hope this doesn't change under autodesk. ICE will work well for many fx tasks except for creating/destorying geometry. no native fluids yet but we got our first 3rd party fluid solver over here...

http://www.mootzoid.com/html/XsiCorner/emFluid2.html

so if you got some brianiacs in your crew the SDK can take you guys farther. i myself have written quite a few plugins for ICE for supporting .bin files and .prt files.

now when you gotta use TP point caching between the two applications is quite easy. there is native support for .pc2 and there is third party support through the blurbeta tools and kai wolter's pc2 plugin.

good luck with ICE in XSI...

steven

PsychoSilence
10-31-2008, 03:35 AM
@ExP

ICE in XSI is great, but keep in mind it is version 1.0 so there is a bit of pioneering right now. fortunately the dev team is very open to suggestions and really want to push ICE to the entire application, lets hope this doesn't change under autodesk. ICE will work well for many fx tasks except for creating/destorying geometry. no native fluids yet but we got our first 3rd party fluid solver over here...

http://www.mootzoid.com/html/XsiCorner/emFluid2.html




looks freakin awesome! wish xsi would have nice volumetrics. thats still fumes marketing advantage. it looks awesome in renders :)

CGTalk Moderation
10-31-2008, 03:35 AM
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.