"Jim Holsenback" <jho### [at] hotmail com> wrote:
> I can see that you've been very busy there aQ.
Not really. Running out of ideas. Fixed the car yesterday, think it was mainly
the distributor cap. Gotta get her hot and really check it.
> It seems that the zip file i
> downloaded to work on is out of date with the images you've posted. Perhaps
> a day late and a dollar short, anyway here's what I've done.
>
> I've cleaned up the source, added four quality levels (debug -> best). This
> image was run a Quality=2 it took about 26 minutes. I didn't have much luck
> with the curved reflectors (convex/concave) so in this setup they are flat.
> They also have a front and back, it didn't seem to make sense to have the
> entire object reflective. It just adds to render time. I would also reverse
> the order of the lenses. It seems to make better sense to make the beam
> bigger with the convex lens then back to original size with the concave
> lens. The other way around and I think photon spacing in the global block
> would have to be even smaller. I'm wondering if curved surfaces might be
> better as iso_surfaces, and with the correct formula made to be a true
> parabolic surface rather than the hit or miss approach with differencing and
> uneven scaling spheres. An improvement more along esthetic lines would to
> frame and make stands for the reflectors and lenses rather than have them
> just basically floating in space. Lastly I opted for a single light source
> (the laser). Both the light and scattering media in the container are white.
> I used absorption media to turn the beam green. I'll post the source over in
> scene-files.
Interested to see what you've done.
With nothing else to do, left this one running last night; 9.5 hrs,
1 + f1 + f2 between lenses. Divergent light must mean a greater focal length for
lens1 by ?
Post a reply to this message
Attachments:
Download 'lasereflect_00_1.png' (190 KB)
Preview of image 'lasereflect_00_1.png'
|
