|
 |
On Wed, 22 Sep 1999 11:26:54 +0300, Peter Popov wrote:
>On Tue, 21 Sep 1999 20:16:01 -0500, "Bob Hughes" <inv### [at] aol com>
>wrote:
>
>>Think I'd use a gradient z pigment instead. I have one RDS (random
>>dot stereogram, not RayDreamStudio) program here that accepts the
>>depth info in Bmp images which I suppose might be simple color
>>shifting, not sure. I need to dust it off and give it a try.
>>
>>Bob
>
>A linear gradient will give a very inaccurate result in points not
>directly in front of the camera. I sometimes use a spherical pigment.
>The problem is that it's hard to get its scale right, especially if
>you have a ground plane, but for indoor or single object scenes it is
>the best choice.
When I wrote my patch I originally wrote it to take the length of
the eye ray (analogous to a spherical gradient) but it seemed to
warp the resulting scene. When I switched to a linear gradient
(length of eye ray dotted with camera direction vector) it flattened
out.
Another thing that can help, if your SIRDS program supports it (I
have one that does) is to use hf_gray_16.
Post a reply to this message
|
|
