Matt McDonald Posted July 20, 2011 Share Posted July 20, 2011 I've recently noticed that when I place a photometric light near a piece of geometry I get random hotspots throughout my image. [ATTACH=CONFIG]43874[/ATTACH] What seems to be happening is that the light from the photometric light bounces off of the very near object and hits something else. I know this physically correct and all but the problem that I am having is that it happens randomly and on scenes of all types. I can exclude objects from FG as a work around but that seems a bit hackish as I want as many object as possible to included in the FG solution. Sooooo, does anyone have any suggestions or tips or something like this? I'd certainly appreciate the help. Matt Link to comment Share on other sites More sharing options...
Justin Hunt Posted July 21, 2011 Share Posted July 21, 2011 does increasing the fg density help? it will take longer to render though. also try using the interpolation radius method. Not too sure if these will help, they aer just thoughts. jhv Link to comment Share on other sites More sharing options...
Matt McDonald Posted July 22, 2011 Author Share Posted July 22, 2011 Thanks Justin On my interior scene I have a very low initial FG point density setting of .05. Over the weekend, I'll submit a render with something a bit higher and report back. My other thought is to set the near component of falloff a few feet from the light and see what happens. My render time on the interior scene is currently 10 hours so I'm hoping I can fix it without much added render time. I have an exterior scene lit primarily by parking lot lights and it does the same thing. A stray ray seems to bounce off the light pole and hit my building. Initial FG in that scene is .3. Link to comment Share on other sites More sharing options...
Justin Hunt Posted July 22, 2011 Share Posted July 22, 2011 it could also be an over-bright material. With photometric lights you used to be able to set a new and Far attenuation, that (sadly) is no longer the case. Near attenuation was perfect for this situation. I dont know why they got rid of it. Are you using far attenuation? If not , do, it helps heaps with rendertimes. jhv Link to comment Share on other sites More sharing options...
Matt McDonald Posted July 26, 2011 Author Share Posted July 26, 2011 I do have far attenuation on. I haven't had a chance to re-render. Some other things came up on one of my projects that prevented me from doing so but I really wanted to ask you about over-bright materials...I'm trying to read up on them but there isn't too much info. so basically what I understand is that a material becomes over-bright when diffuse level and reflection level are great than 1.0. Is that correct? Mental Ray is supposed to correct for that automatically. Right? So when you create a new mental ray material do you typicall setup your material to be diffuse value plus reflection value =1? Link to comment Share on other sites More sharing options...
Justin Hunt Posted July 26, 2011 Share Posted July 26, 2011 Yes it is supposed to automatically correct it , although I have my doubts that it does it correctly or it may reduce to to 1 but it still could be too bright. Generally I do lower the diffuse level as the reflections increase. It does result in a darker surface, but to my eye it has more depth, by reducing the "milkyness" or washed out reflections. This is especially true for glass. I came across another hint yesterday, try switching off the Specular component in the lights. It might not relate to your problem but it does help with speckles in glossie reflections. jhv Link to comment Share on other sites More sharing options...
Matt McDonald Posted July 27, 2011 Author Share Posted July 27, 2011 I've cut the specular lighting out of lights that are farther away from my near field objects in order to reduce render time. I can't say for sure if that has had an impact on the glowing spots. I'll keep testing as best I can (10 to 12 hour render times, with little time to test makes this difficult). Back to energy conservation, so do you look at the diffuse level * the value of the color + reflection level * the peak reflectivity in the BRDF curve? As I understand it that is the equation to preserve energy. Thanks so much for taking the time to help me out. As you know, this kind of stuff isn't documented very well so its tough to determine what one should do based on research alone. Matt Link to comment Share on other sites More sharing options...
Justin Hunt Posted July 27, 2011 Share Posted July 27, 2011 I dont go into things to scientifically to get all the numbers right, rather do it be look and feel. If it looks and feels right then it is right. Your render times concern me, there must be a way to bring them down. If you can or want to can you shoot me the file (stripped out if nessesary), maybe there is something else I'm missing. PM me if your interested jhv Link to comment Share on other sites More sharing options...
Matt McDonald Posted July 27, 2011 Author Share Posted July 27, 2011 Oh, I wish I could take you up on the offer but we've got all kinds of confidentiality clauses and such. I think I might be on the trail of my problem though. I've mentioned that this problem rears up in both interior and exterior scenes. In one scene I was able to isolate the offending problem down to one glass material. In it, we were using BRDF curve to control reflections but (obviously) the refractions are controlled by IOR. I checked the problematic interior scene and it has a different glass material which is setup the same way. I think this is breaking the energy conservation model. I'll have to keep testing but I hope I'm on the right track. Thanks for all your help Justin! Matt Link to comment Share on other sites More sharing options...
Justin Hunt Posted July 28, 2011 Share Posted July 28, 2011 Could be, I like to set glass to "Pass through" for FG, and "Not Generate" GI for photons. What happens when you do a material overide of a matte grey, are there still random hot spots? jhv Link to comment Share on other sites More sharing options...
Matt McDonald Posted July 28, 2011 Author Share Posted July 28, 2011 I put a material overide on everything and the spots disappeared. Next I hid geometry until I figured out what was causing the spots. It was one of the cars underneath a bright light. Then I replaced those materials one by one with matte black materials. One I swapped out the glass the spots went away. In that seen at least it seemed to do the trick. Link to comment Share on other sites More sharing options...
Justin Hunt Posted July 28, 2011 Share Posted July 28, 2011 Have you analized that material it see why it was causing the error Glad you have nailed the problem jhv Link to comment Share on other sites More sharing options...
Matt McDonald Posted July 28, 2011 Author Share Posted July 28, 2011 The only thing possibly odd about the material was the use of IOR for refraction and BRDF for reflection, everything else was normal. Link to comment Share on other sites More sharing options...
Justin Hunt Posted July 28, 2011 Share Posted July 28, 2011 can you post the material and render settings, I want to see if I can reproduce the error jhv Link to comment Share on other sites More sharing options...
Matt McDonald Posted August 1, 2011 Author Share Posted August 1, 2011 Justin and I exchanged several emails back and forth outside of the forum so I want to follow up and bring this thread to a complete state...there seems to be a couple of ways to fix the trouble I was having. Most importantly IMO was "speckle reduction" I had it set to "none". I believe that I have read previously that you can do with out it and get a slight render speed bump. However, in my scenes (which were lit by very bright photometric lights) speckle reduction was necessary to eliminate the dots. The other options include reducing the illumination level of the lights or moving them slightly further away from the geometry. Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now