Final Gather shuts down when Camera Nears Object



(Maya 2012)

I’m having an issue with a scene that uses incandescent particles in conjunction with final gather and GI for lighting:

FG/GI simply shuts down at a certain point in the animation.

[ul]I have 15 objects nested within each other. [/ul]
[ul]They are eaten away, outer to inner, by an animated texture map. [/ul]
[ul]The particles emit throughout as a camera moves toward it.[/ul]
[ul]The particles are the only source of light in the scene.[/ul]
[ul]I have FG mode set to Automatic, (tried others without success).[/ul]
[ul]FG Tracing and Raytracing Refraction limit are set to 20, Trace depth 20 [/ul]

Everything works as expected until the camera very gets close to the outermost object, at which point, the FG solution quits. The outermost object is fully ‘eaten away’ by the texture by now but there are still particles visible in-camera.

Now, if I render the same frame from further back using the perspective view, the FG solution is in full effect. If I move the perspective view close to the animated cam, FG shuts off.

So it would seem to be something to do with proximity to the object, but I can’t for the life of me figure out what the issue is. I can’t use pre-computed FG maps as the animated texture map changes on every frame.

Any help or advice would be greatly appreciated.



FG is currently based on camera projection (although you can bake maps) but hard to tell, can you post examples or a link to an animation?



I didn’t realise it was based on camera projection, but I would have thought / hoped that as long as light receiving geometry was on-camera, the lighting would be solved.

Unfortunately I can’t post any images due to the work being un-released, but I can say that the geometry changes on every frame, so baking a single solution isn’t possible unless there’s a way to sequence lighting solutions.

As I write this, I’m thinking I could give this a go manually, as it’s only on the last 16 frames of the shot where the problem is occurring. I could output a separate FG solution from a perspective camera, that may be the last resort.



There’s always the fgshooter seen on the blog in my signature. It will use other cameras to project the points.

Still not quite sure what is going on or if that will help.


Hey David,

I just checked out your blog-post and it looks like it could solve my problem. I’d never have found that - been trawling the web and this is the first mention I’ve heard of it.

it’s a weird problem and I can only surmise that it’s a combination of the camera being too close the to geometry and the particles are mostly off-screen. There are still a few emitting light in-camera though, but the shut off is sudden, from almost fully lighting the geo in one frame to complete dark in the next.

Your post has given me hope though. Many thanks in advance. If it’s successful, I’ll post back!



Hi David,

Finally got a chance to try your fg_shooter script and the production shader. It has completely resolved the problem!

I had hacked a part-solution across 2 files and it was taking a lot of time to resolve the disparities between the two lighting approaches.

Thank you so much for pointing me in the right direction, I’ve got a single file with consistent lighting now!



Glad it worked!

You should find this workflow will solve quite a bit of your possible variance artifacts and it’s fast to setup. So you may end up using it by default rather than do more renders to test.



I was able test a few approaches with it and opted for static cams only (removed the 0 cam from the fgshooter node) as it was producing an artefact at the edge of the surface.

I can’t imagine rendering anything FG without this technique tbh, it’s amazing that it’s not available front and centre in the render-globals.

Thanks again, the pass finished rendering this morning and looks exactly as I hoped it would! You’ve also opened my eyes to the MR production shaders.



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.