| Author | Message | ||
     Anonymous Rating: N/A |
I want to manually convert a few 2D images to stereo, and I'm wondering, taking in consideration human sensitivity to stereo depth perception, how many different depth layers should I use to get a good effect? Another thought: it would be a good idea to have a program that, similarly to the way random-dot stereogram programs are, would take the image and a greyscale image you can draw as a heightmap, and generate a stereo pair. I've never heard of such a program, though. One idea I came up with is to use a 3D program and create a heightfield then texture map the image onto it, and render two views. Now, when I render from 3D, should I have the two cameras look straight forward, i.e. aligned in parallel, or converging at the object? |
