Maya 2008 Caustics/GI/FG Failure

[KamiNoNamida]

I love the comments on the screens :applause:
More seriousely, mmight just be a dumb question but, you do have a photon shader applied to your blinnSG, right ?

[jpiette]

Is that a new step for caustics rendering with Maya 2008? Because I've never put a Photon shader on my previous-Maya-version scenes and they've always been caustics friendly. I know most of the materials in Maya used to double as photon shaders, or has that changed?

[KamiNoNamida]

Well, on the contrary i've never used (or to be honest, managed to use) caustics before maya 2008, so I couldn't tell you for sure, but i think photon shader have always been necessary for a shader to manage caustics.
Some (or maybe all ?) native mental ray shader have their own built-in photon shader, as the mi-material/x, dielectric or misss_physical, but basic maya materials do not (well, anyway, I don't think so). But since i've only rendered caustics with the mia, i can't be 100% sure either.
Doesn't cost much to test it though, so just check out in the mental ray tab of your shader SG if a photon shader is pluged in (most likely not), and if not, well plug one (basic/dielectric should be fine).
Could fix your problem (but i'm still wondering how you did render caustics without this step before...)

edit: Well, i checked things out and it seems like it is in fact not necessary, photons shaders only seem to give you more controle over it. So maybe a trace depth problem, depending on how your bottles are constructed ?

edit²: damn, i'm sorry i'm really tired... but i don't think a photon trace depth set loo low would give you this error, caustics just wouldn't render i think. So in the end, it really is strange... it is not obvious on the images you posted, but your scene seems to be closed/wraped into inward-facing geometry ?

Another edit btw while i'm on it, i see there's a point light, in the side view of your 2nd and 3rd screenshots, which seems to be in the bottle. If it is indeed in the bottle and intersects with the pipe inside it for some reason, there should be no photons emitting for this light. Don't know if it is the case, but just in case, that (for sure this time) will definintely lead to this kind of error message.

[jpiette]

There is a point light inside the bottle, yes. I put that in there temporarily while I tried to figure out what the heck was going on with the caustics since the liquid material's light absorbtion wasn't showing up. Most of that lighting setup you see is a rather temporary cheat while I try to get my caustics, GI, and FG to work, because I know the second these things decide the render my scene will be really bright.

Yeah, you are correct about photon shaders, they're there to direct the photons, but this doesn't help me since it seems there aren't any photons to direct. x_x;;;

What's more, this isn't just happening on my computer. I tried to render this scene out on several different machines when I was at Georgia Tech, and they all produced the same result. In this case it seems to be either mental ray or Maya 2008 that's farking up.

edit: earlier this morning I tried to render caustics on a simple ramp-transparent and refractive sphere and that failed as well. I had one narrow spotlight trained directly on top of it and emitting about 8000 photons but the only thing I got out of that render was a dark shadow. On the scene above, the photons should be pretty scattered due to the circumference of the spotlights, but again, I'm not getting jack-doodle in the form of a photon map.

[slipknot66]

are all the objects visible to reflection/refraction?

[KamiNoNamida]

The fact that testing the scene on another computer gave you the same results should imply that the problem comes from the scene itself... have you tried caustics in a basic scene on those machines ?
Because if this comes from your scene, there's no reason you couldn't get caustics in a basic scene, even on your computer.

I'll make up a simple scene whith caustics working well after tests, and upload it.
You can then feel free to try it out, so you can know for sure if the problem is deeper than the scene setup itself.

[jpiette]

I tested basic caustics on the other machines as well and they wouldn't work in Maya 2008. They would work in Maya 7 (which the Georgia Tech machines also had) but everything mentioned above wouldn't work in Maya 2008 at all. I agree that at first I thought it was the scene itself, but that appears to not be the case.

I'm sure it's just an issue of some random new checkbox not being checked somewhere. I dunno...

I believe I have all the objects visible for reflections/refractions. Has that changed in Maya 2008? The ground plane isn't reflective, nor is it refractive, but should I turn these on for it anyways?

[KamiNoNamida]

I tested basic caustics on the other machines as well and they wouldn't work in Maya 2008. They would work in Maya 7 (which the Georgia Tech machines also had) but everything mentioned above wouldn't work in Maya 2008 at all. I agree that at first I thought it was the scene itself, but that appears to not be the case.

I'm sure it's just an issue of some random new checkbox not being checked somewhere. I dunno...

I believe I have all the objects visible for reflections/refractions. Has that changed in Maya 2008? The ground plane isn't reflective, nor is it refractive, but should I turn these on for it anyways?
I don't think there's anything more to do in 2008, but i can be mistaking.
It is very strange that none of the machines you'ves tested would redner caustics... i'm thinking there's something wrong in the setup, because if it is some bug problem it would really be freaky that it is occurs on every machine.

About the visible in reflection/refraction stuff, i just ran some tests and it doesn't prevert caustics from working apparently.
( Tests done with "visible in reflection/refraction" in the render stats options as well as "visible in transparency/trasmit transparency/trace reflections/transmit refractions" in the mr tab unchecked on both ground and wrap sphere).
Moreover, your ground plane do not need to be reflective nor refractive (i'm of course talking about its material). That would be pretty stupid since real-world matt materials do receive caustics as well.

I'm uploading the scene right away if you want to test it, that is all I can offer you right now :shrug:

Here for the .ma file (http://kaminoftp.free.fr/Caustics_base_scene.ma)

edit: oh btw, the material for the glass is a dielectric in this scene, i'm testing with a blinn just now.

[KamiNoNamida]

Uploaded the scene with a blinn shader here (http://kaminoftp.free.fr/Caustics_blinn_scene.ma).

[jpiette]

Thanks a lot Kami! I'm at work right now, but once I get out I'll be sure to race home and give those scenes a whirl. Did you create them in 2008 or 8.5? Either way I will post the renders. *crosses fingers*

[KamiNoNamida]

Thanks a lot Kami! I'm at work right now, but once I get out I'll be sure to race home and give those scenes a whirl. Did you create them in 2008 or 8.5? Either way I will post the renders. *crosses fingers*
I'm glad to help you as far as I can :)
The scenes are made in maya 2008 since you seemd to use this version.

[slipknot66]

I believe I have all the objects visible for reflections/refractions. Has that changed in Maya 2008? The ground plane isn't reflective, nor is it refractive, but should I turn these on for it anyways?

The reason i asked you this, its because when you import some objects into Maya, they come with the visible to reflection/refraction flag turned off, and that makes all the diference when using GI and FG.
To make sure, select all your objects, and open the spread sheet editor, check the Render tab to see if all the objects have the visible to reflection/refraction flags turned to on.
Hope it helps.

[jpiette]

Hey guys, thanks so much for the suggestions. Between you guys and the Maya specialist here at work, I was able to figure out some bugs and other issues worth resolving. It turns out to be a funky scaling issue. If the scene is tiny, the caustics work. If it's big, they fail, no matter how high you crank them up. x_x;;;

Thanks again for ALL your help! Oh, and Kami, your scenes rendered perfectly.

[KamiNoNamida]

Hey guys, thanks so much for the suggestions. Between you guys and the Maya specialist here at work, I was able to figure out some bugs and other issues worth resolving. It turns out to be a funky scaling issue. If the scene is tiny, the caustics work. If it's big, they fail, no matter how high you crank them up. x_x;;;

Thanks again for ALL your help! Oh, and Kami, your scenes rendered perfectly.

Well, that's good to know :curious: (http://forums.cgsociety.org/misc.php?do=getsmilies&wysiwyg=1&forumid=0#)
Thanks for the feedback, happy it worked out well for you :)

[jpiette]

Yeah, the Maya guy here at work managed to tinker it down to one of two things it could be: either an xdiff-related issue or a scale-related photon dispertion issue. I'm moving some items over to a new apartment this weekend, but when I have time I will go through both cases and tweak it down until I can figure out exactly why scaling is an issue. If it's fixable by hand then I'll post the fix, otherwise I may submit a ticket to Autodesk on the matter.

[slipknot66]

Well, thats true, one thing i noticed, always keep your scene center with the Maya grid and things work.
Thats a Maya problem not a mental ray one, just check in 3d max.
In max (ah i dont like this software..personal opinion :) )mental ray dont have this problem.
But thats an integration problem.. Autodesk! :/

[jpiette]

Hehe, yeah, it did seem to pretty much scream "I'M MAYA BEING A DUMBASS!!!" I have no doubts it works well in 3D Studio Max, I've never once had a mental ray problem in that piece of software. Too bad I work in Maya now. You're saying it's also an issue of how centered the scene is to the grid? Hm... when I go to putter around in it tonight, I'll be sure to take a look at that as well.

EDIT - Slipknot66, I just saw your award-winning piece. That is AMAZING. :applause:

[CGTalk Moderation]

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.