Hey all! I work for a forensic engineering firm, specifically for automotive crash reconstruction, and one of the bigger struggles we've encountered is getting accurate lighting. In particular, we've had a string of cases involving visibility levels in low light conditions, where the ability to perceive, or not perceive, something like a pedestrian helps determine who is at fault.
Now, I know that Arnold is supposed to have accurate GI, inverse square falloff, etc. However, near as I can tell the light intensity doesn't correspond to real units. We are able to measure the intensity of light sources in terms of lumens, candles, etc. Is there a workflow that will allow us to say that, alright, for a light source of this intensity, it would be exactly this value in Arnold?
Secondly, is there a way to be able to match the exposure of the camera in Arnold as well? Say, if you know the camera, lens, have all the data like F-stop and ISO and shutter speed, and match that in Arnold? We already are taking multiple exposures and comparing them on site visually, and recording the exposures that most closely match the lighting we are seeing. The end goal would be to be able to say, with good deal of certainty, that the light levels in our renders matches what would be perceived by the human eye.
I know that this is likely a bit of a broad subject by this point in the post, I am more looking to be pointed in the right direction. I haven't managed to turn up anything helpful on this subject. In particular, if anyone knows of any academic papers on the subject those would be especially useful as they can be cited in court.