hi Andrew, hi All,

Andrew Coppin wrote:
 > It was about June time last year that I discovered how Single Image
 > Sterograms ("magic eye pictures") work. You can actually make 'em
 > out of plain ordinary text! I've done so with Notepad many times...
 >
 > But I figured you should be able to do it with POV-Ray too - after
 > all, it IS a 3D program!!!

as i mentioned in my last post, i am currently developping this
a little bit further. i wrote a perl script to create SIRDS from
image files.

i fixed a lot of bugs (!), and added some new features again.
though it still renders SIRDS only, it now supports

- reading and writing of targa format version 1 image files
   (only those produced by povray, with the settings that i
   usually use; which are 24 or 32 bit (w/o [with or without(?)]
   alpha channel),
- scans the command line for depth map and texture file names
   (it even supports escaped wildcards like "image\*" for depth
   map files, so i can render a full range of frames for an
   animation) - but the texture feature is not yet working.

so i could finish my first six seconds of an animated SIRDS
clip. i hope that i will be able to compress it to some
reasonable size.. without losing too much image information
due to lossy compression.

when i played with one video card's controls, i found that a
rather dark picture with a few bright dots is astoundingly
brilliant when animated! - my eyes also liked it alot, btw! ;)

so, i'll make the images mainly black, and set some
occasional bits along the way. this is quite easy to run
length encode, so i will probably even post a FLiC or such.

i will also add textures to the SIRDS to color them..

and sometime maybe also "real" color, directly from the
depth map file! - as this is either a 24 or 32 bit targa,
so i also have the alpha channel at my disposal - it can
be fed with depth values from a gradient (as described
in news://news.povray.org:119/3e1925c4@news.povray.org ),
leaving the full 24 bits for colors.

i could of course do it in two passes and just render both
layers separately. but neither do i like the idea of doubling
the data amount (well, at the worst), nor the increase in
processing time and disk clicking due to a second file
having to be read and processed. :/

but any way i use the "filter" or "transmit" values to get the
"z-alpha", either i get just some color, and the alpha values are
quite off, or a little more color, and a devastated alpha channel
in most places (especially where objects overlap or, even worse,
intersect!)

you can visit my homepage to see some examples:

   http://www.psico.ch/index.php?page=sis

if anybody can see what i mean and knows how this can be fixed,
please contact me!

[...]
 > PS. Can anyone here see sterograms btw? Kinda pointless if you
 > can't :P

sure. - mine are even moving and turning..! - would be sad
if *i* couldn't! :D


g.

-- 
mailto:gim### [at] psicoch
http://www.psico.ch/
http://psico.servehttp.com/

Post a reply to this message