home *** CD-ROM | disk | FTP | other *** search
- a) It's not actually my teapot (the original data was
- designed by a chap called Martin Newall, and is now one
- of the standard items in computer graphics literature).
-
- b) The original data was in parametric bicubic Bézier
- patch form, and the thought of rendering it directly
- made my brain overheat; so I converted the data to
- polygon format (64 polygons per patch).
-
- c) I rendered the polygons using my own Gouraud-
- shading routine (which incorporates a simple pixel-
- dithering algorithm) and a quasi-Phong shading model to
- give a metallic sheen. There are 77 frames in all,
- depicting the teapot rotating about two axes at once.
-
- d) Each frame took a couple of minutes to generate
- (even with the whole rendering process running in ARM
- code, this is still pretty slow compared to dedicated
- graphics hardware).
-
- e) The frames are individually compacted using an
- extremely tacky run-length-encoding procedure of mine
- and take up far more space than need be.
-
- f) As a result of the rendering methodology, the end
- result is covered in digital artifacts, both spatial
- and temporal (particularly noticeable if you watch the
- spout). The decompactor only runs at 25 Hz and so
- this doesn't help either.
-
- So despite Jonathon's enthusiasm, perhaps you can
- understand why I don't shout about it from the hills.
- I guess it's not bad for a quick sketch, but there's
- certainly a lot of room for improvement. I might have
- a go at ray-tracing it at some stage, but at the moment
- there are various things like earning a living and
- studying for a degree which get in the way !
-
- Zeus