ok i think my analysis is correct. working out what is actually happening with defective stereoscopic images generally involves getting a headache.
when i render a stereo pair from a shade map, the parts of the scene behind a window all appear to have the same depth as the window. comparing the two renders in photoshop its pretty clear that the view through the window is identical in both views, whereas it should be perspective shifted like the rest of the scene.
initial low res tests suggest that excluding the glass from the shade map fixes this, but until i do a -long- high res render i cant be sure.. there is a suggestion of some artifacting, but its too low res to be sure.
does this make sense?
when i render a stereo pair from a shade map, the parts of the scene behind a window all appear to have the same depth as the window. comparing the two renders in photoshop its pretty clear that the view through the window is identical in both views, whereas it should be perspective shifted like the rest of the scene.
initial low res tests suggest that excluding the glass from the shade map fixes this, but until i do a -long- high res render i cant be sure.. there is a suggestion of some artifacting, but its too low res to be sure.
does this make sense?
Comment