c10_h006@jhunix.HCF.JHU.EDU (Stdnt 06) (04/07/89)
For those of you who were interested in my original question about
n-dimensional rendering, I'm told this is a definitive description
of the process.
From: tr@cbnewsh.ATT.COM (thomas.rosenfeld)
See Communications of the ACM, Vol 10, Number 8, Aug 67 for
" A Computer Technique for displaying n-Dimentional
Hyperobjects" by A.Michael Noll.
Feel free to post this, for some reason my reader wont let me folowup.
Tom Rosenfeld
{attunix,homxc,mhuxa}!pixels!tr
However, I can't get a hold of it. More or less, though, I only have one
remaining question.
The algorithm I'm using is to translate and rotate the origin so that it
is located at the viewer's position, and the axis of the dimension I'm
shaving (dividing out) points at the old origin.
What I want to know is, do I need to re-rotate the origin for each dimension,
or is one rotation enough? That is, do I need to reorient each axis to point
to (intersect) the old origin as I divide it out, or is the initial orien-
tation to the first axis enough?
If my description is too general, drop me a line and I'll repost with more
detailed mathematics.
Gunther Anderson
ins_agwa@jhunix.bitnet