Strange shadows

[dandelO]

Here's a small .obj sequence TG test render using a cube, there are 3 light sources, all identical except their colour and position. The yellow shadow is wavy around the edges and looks like it uses a much lower render detail than the rest of the scene, while the other two look fine. What's going on here, I wonder. http://youtu.be/8Tbo8YY8npM

* Another object sequence TG tester, just while I'm posting, is http://www.youtube.com/watch?v=m4Kaewrjzi0
Both crap, just play files but the shadow issue is strange.

Trying to make a kind of Jenga tower rendered in TG but the collision physics is giving me a bit of annoyance in Carrara in that when I begin stacking my Jenga pieces the structure immediately explodes when I run the bullet simulation, regardless of physical scene settings like collision distance and other various physical properties that can be set relating directly to the objects. Another hassle is that I can't find a sequence exporter for Carrara so it means exporting Collada files to import into another application to get the object sequence in .obj format that can be used in Terragen. A lot of pissing around, in other words.

[TheBadger]

Thats weird Martin. The problem and the first animation. When I looked at the first animation, I had to watch it three times. My eyes just did not understand that the shadows were shadows, because they were made of color. Whats strange about the problem is it only effects one shadow.

[Oshyan]

And there is no difference in settings between the different lights? This difference shows up in the raw rendered images? Very weird.

Nice deflating anim btw.

- Oshyan

[dandelO]

No, Oshyan, all the sunlights have pretty much the same settings, they are set at 0, 120 and 240 degrees, all at the same height and all at the same strength, the only difference is the colours, one red, one blue and one yellow(if I remember correctly, I'm not at home just now but I do know that they're all the same, barring colour and position).
The error is present in every frame of the sequence. Odd, eh?

I don't have the file here but if it's any help to use for troubleshooting purposes I'll post it tomorrow when at home. I will triple-check it again beforehand, in case there is something I've just overlooked but I really can't see it or think of a reason for it so far, it's a strange one.

[Oshyan]

Yeah, I'd love to take a look if you get a chance to send it.

- Oshyan

[Matt]

Is the ground plane an object that's being ray traced, and is your minimum samples per pixel less than one? I ask this because the difference is in the shadow from the blue light, and blue is considered less important when doing adaptive sampling. Maybe you just need higher anti-aliasing settings?

Matt

[dandelO]

I think that's probably it, Matt. The ground is just the planet node but since it isn't displaced I think I used RTE on it and the sampling is quite probably very low. I don't plan on going back to it for any reason but thanks for that info', I just thought that was a pretty strange difference when I noticed it rendered. That clarifies what is likely happening nicely. Thanks.

[dandelO]

I ask this because the difference is in the shadow from the blue light, and blue is considered less important when doing adaptive sampling.

Confirmed. Adaptive sampling is below 1 minimum sample per pixel and the planet is ray traced.
Thanks.

[bla bla 2]

Trop drôle les vidéos