"Hot" Photons in an indoor scene

[Valtiel]

Hey everyone,

 

I'm just getting to know VRay's GI functions after switching from MR and am coming up with some questions I can't find answers for in the online documentation. Specifically, I'm trying to get a handle on using photons for indoor scenes with only partial or indirect fenestration from the exterior. Through the official tutorials, I've gotten familiar with tuning a photon map before switching to another GI method for primary bounces.

 

Being familiar with portals and Max's daylight system, I set up a daylight system with the skylight disabled, a vray sun as the sunlight, and the environment map conventions specified by the online documentation as well as a Vray light at each window to focus the daylight. The system units are in inches, and the distance from the window on the right to the corner near the centre of the image is ~10'. For the most part, it's turning out ok except I'm getting a few 'hot' photons, much brighter than the others, in each render which, as expected, translate to bright spots when a smoothing solution is applied to my primary bounces.

[ATTACH=CONFIG]46944[/ATTACH]

I can only assume it's coming from the sun light but I'm a little vexed as these hot photons remain when I disable the sunlight (in the posted image, it essentially looks the same as the 5 unit emit radius render). Moreover, I know in MR that having an inadequate photon emission and radius from your sunlight can result in too few photons entering the scene as it is not focused by the portals like the skylight. Lowering the emission radius seems to darken the scene but not really affect the behaviour of these hot photons any, suggesting that the sunlight's energy is entering the scene at higher radii in significant amounts but perhaps some photons are not adequately dissipating on their bounces? Increasing the diff subdivisions on the sunlight doesn't appear to affect these hot photons either, despite increasing render times and casting more photons into the scene.

 

I've set the search distance to less than the thickness of the walls so I'm fairly certain it's not sampling energy on the exterior walls. I do kind of miss the visible emission radius on the mr sun.

 

Any ideas on how to solve this? Am I even using the correct solution (maybe brute force instead?) for this kind of scene?

 

Thanks for any feedback provided, cheers!

[erickdt]

Hi Riley,

 

I think you'll find that most folks that use VRay will use Irradiance Map for the primary bouce and Light Cache for secondary. The are some instances where Brute Force for your primary bounce will yield better, more realistic results especially when it comes to fine detail like in with cornice mouldings. This comes at the cost of a higher render time though. That said, IR Map and LC are usually a pretty good a reliable solution.

 

I for one haven't used Photon Map ever inside of VRay so I can't speak to the settings in that regard. I would recommend you switch over to IR/LC and see how it turns out.

 

Hope this helps!

 

E

[Valtiel]

Thanks for the quick response, Erick.

 

I'll give IR/LC a try when I get back to my workstation (am in class atm). In my experience with MR, shooting methods like photon mapping were often necessary in indoor scenes where light needs to bounce a great deal to find the camera from the source; that gathering methods can't 'find' the light source if it has a small and indirect access to what the camera sees. Do VRay gathering solutions (both IR and LC I understand) work well enough to overcome this limitation?