Issue by benjamin-heasly
Friday Jul 17, 2015 at 17:48 GMT
Originally opened as https://github.com/DavidBrainard/RenderToolboxDevelop/issues/21
Mitsuba factoids give us lots of useful "ground truth" images about scenes, like albedo, distance-to-camera, etc.
PBRT doesn't have the same kind of factoids.
But for the same scene and camera geometry, Mitsuba renderings and PBRT renderings come out very similar. So the "ground truth" about Mitsuba pixels might be applicable to the "ground" truth about PBRT pixels.
We should check on this! Does this ever work? Under what conditions?
Friday Jul 17, 2015 at 17:48 GMT
Originally opened as https://github.com/DavidBrainard/RenderToolboxDevelop/issues/21
Mitsuba factoids give us lots of useful "ground truth" images about scenes, like albedo, distance-to-camera, etc.
PBRT doesn't have the same kind of factoids.
But for the same scene and camera geometry, Mitsuba renderings and PBRT renderings come out very similar. So the "ground truth" about Mitsuba pixels might be applicable to the "ground" truth about PBRT pixels.
We should check on this! Does this ever work? Under what conditions?