 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
The source code and resulting images are in beta-test.binaries. There seems
to be a difference in the sky-spheres in 3.1g and 3.5. The CSG object
appears to be the same on both tracings. When a media sphere was added
around the CSG object 3.1g behaved as expected. Unfortunately, 3.5 produced
what looked like an enlarged version of the CSG object with the camera right
up against the side of the object. The source code is in the test binaries
group. The media sphere at the bottom of the code was simply commented in
or out and run through both programs. Thanks for your input. Regards,
Kynar
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
It seems to be the cylindrical light creating a blank disk. I can't
understand why.
bob h
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Change the cylinder light source radius and falloff to 2. When you change
the rotation it will disappear at angles like 45*y and 90*y when turned
opposite side facing camera, as opposed to -45 and -90.
I did try without any of the CSG involved, thinking it could be that, but no
change. Making the light parallel had no influence either, only using
shadowless or no_shadow helped.
bob h
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Kynar had said:
> It seems to behave properly until the media sphere is added. It also
seems
> to vary sometimes according to angle when the media sphere is present. I
> have been fighting this at work for 2 days now. Just color me frustrated.
> LOL
>
> Do you have any idea why the sky sphere is acting the way it is??
Hi John. Yeah, I believe the default wave form changed from triangle_wave to
ramp_wave for the gradient pattern.
That media sphere and cylinder light combo is a mystery to me. All I can
tell is that without media, or shadows, that the strange black disc
disappears. So it must be something wrong in the way those work together.
Well, I'm not the one to say it really has a bug since I don't know about
the program source code involved. Just pretty obvious it doesn't seem to
have a fix by trying different values for the keywords those use.
Hopefully someone else that knows something will join the discussion at the
beta-test newsgroup.
P.S. I'll add this e-mail to there.
Bob
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
