> If you frequently use such constructs, I suggest writing a macro.

No, I think one of the in-house mathematicians should work the formulae to
handle images as textures and post them in the documentation.

Making tests now I can understand the faded border in some pictures I remember
seeing from time to time to be the added pixel columns when the image is
distorted only a few lines beyond its normal dimensions. It does not show up
with a single pixel column added, but when it is ten or more it is noticeable.

Basically we need a f(w,h|camera<...>) to be translated into texture {translate
<...>} and object {transform <...>} so that the resulting image be at x,y with
width and height w,h. I think it needs somebody very conversant with
scene/camera assumptions and coordinate systems to convert the whole issue into
pixels.

Post a reply to this message