Here's a stupid Rhino vase a made a while back for my chimney scene. One thing I had a problem with was null-shadows, the black areas that Bryce's raytracer produces when it "runs out". This is Take One, unsmoothed and rendered at RD of 4. I didn't let it finish AA (almost at the bottom), because it was garbage.

And here's the next one, smoothed 180 degrees and rendered with softshadows and a RD of 6. (TIR depth is equal to the Ray Depth in all of these renders...)

Another RD 6 image, this time with a small light rig to cut down on the render time. Notice the black null shadows at the top of the scene. There is no haze or atmosphere on these images, so far. Which was a problem, what if these null-reflections (sorry, I think I accidentally called them null-shadows earlier) were actually reflections ofthe black, nonexistent horizon? I like the lighting a bit better with the light rig, as opposed to Brycean soft shadows. (The softshadow render previously shown was at 64RPP, it took some time.)

And this one, I did last night, same scene. I turned the atmosphere on, set it to some reds and oranges (custom mode, of course), set some light orange haze and fog, and rendered it at a Ray Depth of 18. TIR was also 18. There are no null-reflections, as the shadow color is the same. The render time was just over 2 hours, on my little Duron 1.3 with 512 MB-DDR. All of these are pure renders, no post... I finally got rid of those little bastard null-reflections! Thanks for viewing...

Good study! I can't get over how real these look...

thats freakin amazing---these turned out awesome.

Which brings up an interesting question. What if you placed a white image on a large 2D face and placed it just behind the camera out of view so that it reflects off the glass vase, would that get rid of the null reflections? How about a white dome over the scene? Great demo and very nice renders!

Very thought-provoking experiment and renders. I really like that last render you did with a "roll-your-own" sky. I think it's very realistic. Excellent work your Lordship!

#4 does it for me! Execellent render....!

Available on Amazon for the Kindle E-Reader

All the Woes of a World by Jonathan Icknield aka The Bryster

And in my final hours - I would cling rather to the tattooed hand of kindness - than the unblemished hand of hate...

The final image really is the best. It looks most realistic in color and reflection. 8) Catlin

Aye, thanks for viewing! Quest, null-reflections come from rendering with a Ray Depth that is too low for the object being rendered. Of course, Ray Depth has no effect on materials/objects that are not transparent or reflective. If you had a white background/sky shpere, you would still see black nulls. I believe the null-colors are based off of the Master shadow color, in the Sky pallete. One more render should clear all of that up, back in a few...

Okay, I made a mistake, but it had nothing to do with null-refractions or whatever. I had my lighting setup backwards, meaning that all these image WERE rendered with softshadows in Premium mode, EXCEPT the first one. Other than that, this image will answer Quest's Question! Ray Depth of 2, TIR also 2. I also turned the Master Shadow Color to an purple, in case that had anything to do with the color of the empty reflections. It did not. Empty reflections, like quantum singularities, are black, regardless. Now I wonder if empty reflections have any ambience in True Ambience modes...

I like this one too. It looks like a beautiful smoked glass, which means that null reflections are not all bad. 8) Catlin

So would a thinner vase need less TIR/Ray depth? Or does the thickness make no difference in which case we could just say "if you want nice glass use TIR set at 18". BTW #4 is beautiful.

Ang, there really is no way to judge such a thing. Bryce's render report wouldn't, for example, say how many bounces a particular object in the scene used... Or an average, or anything like that. But consider that this glass object is an actual Rhino object I modeled. It has an inside and an outside face. So, if you shot a ray straight at it from a perpendicular angle, it would need 1, 2 (for the first side) then 3 and 4 counts to emerge from the other side. And every angle of incidence that is NOT 90 degrees dead on, would need progressively more Depth to let the ray emerge. For example, a light ray going almost straight down along an edge? The more Ray Depth, the merrier! Considering that this model is fairly high-polygon, and smooth, I'm thinking that 18 is probably as high as necessary, although mathematically 16 would make more sense...