Here's a few wips for my challenge entry showing the difference between normal and TIR.

The top version has TIR set, the lower doesn't. I actually prefer it without.

What do you guys think?

Also here's the thumbnail to my final(probably) entry.

OK. Time for another noob question. What's TIR? Triple infinite redundancy? Toothpick in reservoir? Talent injection renderer? Topical image replacement? Tropical investment reserves? Please enlighten.

All of the above. Total Internal Reflections It's what makes diamonds look real and if you're in a fish tank why you see a reflection on the underside of the surface of the water.

Ah, yes. Thanks. I see it there in the Render Options.

In this case, I like it bettwer without it. TIR usually works well when there are many facets, such as in an imported .obj vase or something... Great work on this, though! The lights are wonderful...

So THAT'S what TIR is (don't think I've got it in Bryce 3.1 (it moves). :-)) And yes - I agree, the version w/o TIR looks better. (Trying to sort out "why?" in my head, I reckon it's because in real life you very rarely come across TIR; so when you DO see it it sort of looks unreal. Must be great for certain special effects though.) Just my usual mental muttering... Cheers, Diolma

There's one thing that caught my eye: the bump on the wood. It just doesn't look right to me. Here's my thinking: If the wood is not polyurethaned it would have some bump that follows the grain. If it has been heavily coated so that the grain doesn't affect the surface it would still have some bump... probably along the lines you're using, but perhaps with a higher frequency and a lower bump value. Just my buck fitty.

You're probably right. I was lazy and used the same image as the bump map, so it doesn't quite work. Thanks for pointing that out. ;-) This was about 70% and I reduced to about 40% for the final render, so hope that looks a little better. Too late now though. Bored of this picture and moving onwards and upwards.

Well, actually, the final is where i noticed it. In the tests there's a slight following of the grain at the edges of the refractions on the wood. In the final, that's gone. The bump in the tests works a little better for me... for what it's worth... since you've moved on. Case closed. ;o)

Just out of interest, yeah, I know I had closed it. But how do you think I would fix that? Lower bump value didn't work. I could go way down on the bump and increase the reflection and say it was varnished. Higher frequency? Different bump map? Learning all the time.....

As a novice, i don't really know what to suggest. But here i go anyway. ;o) One possibility is to take the texture into an editor, change it to greyscale and see if you can tone down the random bumps and bring out the grain more. I don't know how to accomplish that, but it's something i might play around with to see if it helps. Increasing the reflection might interfere too much with all the lovely refractions and reflections already happening in the image. You could try using a different material for the bump - like, use the Blue Bump texture, crank up the frequency and turn the bump value way down - that would help make it look like the normal variations in most surfaces. That's all i can think of at the moment. Perhaps someone else has the perfect solution on the tip of their cerebrum.

Ok, since I'm a student of physics/optics etc, here's the explaination. When light waves hit the surface of a transparent object, some pass through, some reflect, obviously. However, here's the deal, when the wave passes through an object the refraction index determines how quickly the light can travel (or any transverse wave) in the material. For physics sake and the sake of the explaination this will be called n. n determines the final angle the light will take based on its initial angle after it enters or exits the surface. This works as follows Sin(angle)initial / Sin(angle)final = n. Then you can simply solve for n in order to get the angle ie arcsin(angle) final = Sin(angle)initial * n. Which gets converted to Angle final = Sin(sin(angle initial)) * n. Now, when you enter glass the light gets bent towards the normal line (the perpendicular bisector of the surface. IE a virtual line drawn coming straight out of the surface) This causes a refraction index that is greater than one, water is 1.33, air is ~1.00 as is a vacuum. NOW we get to total internal refraction. Ok, here's when and why this happens, when a light beam is traveling from a surface with a high refraction index into a surface with a low refraction index (ie water into air) the light is bent AWAY from the normal line. At a certain incident angle the light will bend so far that it will actually become paralell to the surface and therefore will not escape the surface, even greater angles will result in even more internal reflection. However, that is the point of total internal reflection. Now, in materials with higher indexes of refraction the angle this requires is much less, as seen through the formula above. This is why diamonds sparkle so much more than cubic zirconia, because the refractive index is so much higher. Also why diamonds are cut at certain angles to project light through the top faces thus making it appear brighter than it would be otherwise. Pretty cool stuff, I'd draw you a diagram, but I'm really lazy.

Oh yeah, and on the subject of the image. The reason you probably prefer it without tir is that this kind of image does not really benefit from it. If you have a brighter environment with gems cut a certain way then by all means use it, but if not... don't. Also, your refraction may be up too high, is it on 133 or 155? The more refraction, the greater tir will effect it, in this case, very negatively.

It looks like the TIR version needs to have a greater Max Ray Depth. I'm pretty sure that's why some of the surfaces are dark and not reflecting the scene becuase when Bryce hits the MRD, it just colors the surface. I've never understood how Byrce chooses this color but it's usually a shade of gray. For glass/TIR scenes I always set the ray depth quite large, around 100 MRD. This way I can be sure that no rays are cut off too soon. Just becuase it's set to 100 doesn't mean that every ray gets bounced that many times, the ray trace bounce still stops when it hits something "solid". Excellent image; Dan

100 is quite large. I usual don't go above 8. Maybe I'll try a re-render tomorrow on a higher setting.

yes i liked the wood too in the above pics just change it back to those settings, and without tir is my fav too as i think diamonds is just another carbon lol and they arent forever as we all know carbon burns.

Aye, aside from what Ornlu and the others said, Diolma, you ALWAYS encounter TIR in real life. At a level that is inifinite... Bryce, or ANY renderer, cannot handle this quite yet. Which is why in Bryce 5, you can set the depth of this effect, the higher the depth, the longer the render... At 100 Ray Depth/TIR, this image would probably take a much, much longer time than at 8, for example...

Attached Link: http://www.rockhounds.com/rockshop/gem_designs/refractive_index/

whoops, it's not part of the 3d lapidary site, but that's where I linked to it from

I'm doing an image at the moment which has ray depth set to 50 and TIR set to 20. So far it has taken 2 days and it's still going although I don't think it will be much longer. I must say that when I've stopped it to save, just in case my cat pulls the plug out again, it looks really good, even if I do say so myself. Another reason for the time is that it is at a size of 4000x3000. 8) Catlin