[zdeno]

Maybe you could help me about it?

I would but Brian was faster (apple-polisher ). and it is better because I have only account on evermotion forum and cgsociety is better.

thx for .mxs Master

[tom]

Thea Render and Indigo Renderer contributions on CGTalk are added to first post.

[simmsimaging]

I'd like to see the better Vray version you said you have on hand - is that going up sometime soon?

b

[tom]

I'd like to see the better Vray version you said you have on hand - is that going up sometime soon?

I just think this is closer to the reference image.
http://img85.imageshack.us/img85/9678/vray20minppt.jpg

[simmsimaging]

I agree the material is closer for sure, although I still think it could be rendered a bit better Why not put that one up instead of mine?

Aside from Maxwell I can see a lot of consistencies between the other render engines. Maxwell still looks the best though.

b

[tom]

I agree the material is closer for sure, although I still think it could be rendered a bit better Why not put that one up instead of mine?

I need to ask permission from its author first. Also, I was waiting for your approval. So, now I can proceed...

Aside from Maxwell I can see a lot of consistencies between the other render engines. Maxwell still looks the best though.

The biggest characteristic differences could be about different BSDF/roughness implementations. So far, Indigo and Thea are more or less doing very good. My main complaint about Vray result is different. Pay attention to the refraction around shoulder and neck area. Vray has a quite different look in that part like the material has a different Nd or it's not just a rough glass. Can you see this too?

[simmsimaging]

....The biggest characteristic differences could be about different BSDF/roughness implementations. So far, Indigo and Thea are more or less doing very good. My main complaint about Vray result is different. Pay attention to the refraction around shoulder and neck area. Vray has a quite different look in that part like the material has a different Nd or it's not just a rough glass. Can you see this too?

I can see a definite difference in the "internal" detail that that area that these engines are showing that Vray is not. I have played with other material variations that did better in that area, but then other aspects didn't match your reference as well so I discarded those. Since my *personal* goal is just to get something that looks good I was satisfied that something could be had from Vray, and most of the others too, although I would still be very surprised if any were to surpass Maxwell.

b

[tom]

I can see a definite difference in the "internal" detail that that area that these engines are showing that Vray is not. I have played with other material variations that did better in that area, but then other aspects didn't match your reference as well so I discarded those.

Interestingly, the Mantra render out of Houdini has a similar look about that detail.

Since my *personal* goal is just to get something that looks good I was satisfied that something could be had from Vray, and most of the others too, although I would still be very surprised if any were to surpass Maxwell.

I really try to be objective, accept the current results successful and avoid to think Maxwell is the best, because my primary goal is to show there are others pending which are quite suspicius about such an achievement.

[simmsimaging]

Here's an example of what I mean - some areas are better, others are.... different. To me either one would be acceptable for a project though, there there is a smoothness to the Maxwell version that I haven't seen matched and that I would prefer to get.

(note the higher render time though )

b

[tom]

Now, it looks better than the previous attempt. What was wrong making the neck/shoulder reflection different? Also, is it possible to get rid of that cyan tint and match the background blue better? To minimize the differences I mean...

[simmsimaging]

It's not that there was something wrong per se, just different ways of achieving the frosted effect create different options. I have tried various levels of reflective and refractive glossiness, and also one that achieves the effect entirely by bump mapping, to better simulate the way it's done in Maxwell etc. They all create slight variations.

Anyway, I can match the blue in Photoshop and re-send it if you like.

b

[tom]

I'm confused. So, do you use bump mapping?
(Btw, thanks for the color matching suggestion. I'd not like to see any kind of post-processing in this test.)

[simmsimaging]

No problem - I will re-render it and try to match the blue value more closely.

This example is not using bump, it's reflective and refractive glossiness in a VrayMaterial. I have tried using bump as well though, and the result is similar, but different in other ways. Just a lot of ways to get there really, but none are truly identical.

I have some today so I'll re-do the dragon based on this last example.

b

[tom]

Thanks for your time Brett! Your inputs for providing a better V-Ray output is very valuable.

Meanwhile, see what I found when playing with the hue on this monochromatic scene.
The blocking artifacts are JPEG compression but, what about the glass itself?
With Maxwell (on the left), I can make 360 degree rotation without a single bit of loss in quality...

[simmsimaging]

Not really sure why that happened. I just tried it with the new one I rendered and it seems to adjust fine. Dunno?

Here is a new one. I re-adjusted the background tone to better match the colour and density of yours. It's not perfect, but closer.

Also decided to just crank up a few settings far beyond what I normally use just to see what happened. Along with that I turned on anti-aliasing, which I normally don't use (but should for something this lo-res I think). The result was that the image is a lot cleaner, but the render time really shot up. I'm pretty sure it could be tweaked down a fair bit, but felt this would make the Maxwell users happier to see

b