Harold Baize wrote:
 > Hermann, Great. I'm a stereo enthusiast too. Stereo is why I
 > started doing POV-Ray. I also would be more interested in output
 > as JPS for use with LCS glasses than anaglyph output.
 >
 > Do you know of SAB's stereo patch?
 >
 > http://sabix.etdv.ruhr-uni-bochum.de/sabpov/
 >

As I came acros it, most of the work for my patch was alredy done. I
didn't try it, but from the description at the webpage it seems, it is
rather focused at providing different output formats.

Anyway, fisheye-type perspective is very important to me, and for this
I had to create a new projection type that can be made steresocopic
over the whole viewing range.



 > .....and then POV 3.5 came along and the patch is for 3.1g.
Yes 3.5 !!
 From looking at the MegaPOV sources I guess it will be possible to
adapt my patch to it. But post-processing hooks in at the same
locations where my patch hooks in, so I decided to first do it for
3.1g and wait for the release of the 3.5 sources. I'm verry impatient
to be able to try out photons and iso surfaces, as you may guess. :-)


 >
 > How do you approach the stereo window?
 >

My approach is: let the "up" and "right" vectors specify the image
dimensions (i.e. dimensions of the resulting window) in pov ray units.
So the user may chose the pov ray units to represent real world units
as he/she sees fit.
To give an example, I use 1 pov ray unit = 1 m and use

camera{location 0
        direction 1.75 * z
        up 3/4 * 1.8 * y
        right    1.8 * x
        translate ....
        look_at ....
        }
to get a f=35mm view correctly adjusted for a typical home/club
projection screen of 1.8 meters width. I.e. if you magnify the
resulting image so it is 1.8m wide, the max.deviation (for objects
located in infinity) results to 65mm.

Of course you may use the "angle" keyword, but by using "direction", I
know precisely where in my scene the stereoscopic window will be located.

Regards,
Hermann

Post a reply to this message