Maxwell Render

First attempt in Jewelry rendering.

Awesome!!! - First attempt? You have done a great job!! Great clarity on the diamond!

_________________
Tea_Bag's Maxwell Gallery: viewtopic.php?t=29110
Portfolio - http://www.dvc3d.com (Under Construction)
Texture Packs - http://www.dvc3dtextures.com (Now Live)

Excellent job!

_________________
Next Limit Team

great job for attempt, diamond very clear what did you use for lighting
good to see some jewels

Actually, there's something missing in every single jewelry render that hurts my eye and quickly hits the wall. The diamond cut should have very low rough micro bewels on the edges of facets. I think you can imagine how much realism it could add. Here you can see it:

http://www.wellpromo.com/upload/upimg79 ... -19579.jpg
http://c0373252.cdn.cloudfiles.rackspac ... n_zoom.jpg
http://treesflowersbirds.files.wordpres ... iamond.jpg
http://www.feng-shui-products-decoratin ... 410091.jpg
http://diamondstudearrings.files.wordpr ... iamond.jpg

-edit: Also the gold part needs more subdivision as it shows faceting in this closeup.

_________________
Next Limit Team

Hey Tom, actually most of those pictures were of materials other than diamond. As a Gemmologist one of the defining indicators between diamond and its imitation counterparts is the absence of those micro bevels. Those small bevels are created in polishing in materials softer than diamond. Looking for those small rounded edges is one of the first diagnostic criteria that you are looking for when trying to determine if a stone is diamond or another material.... if they are present you know you are not looking at a diamond. Granted at an extremely high magnification (electron microscope) you would likely see very small rounded edges. Not showing off just figured I would chime in

Neeper... really nice render! The diamond looks too dispersive to me but your lighting and metals look great!
Cary

_________________
Cary James

CONTACT
www.contactdesign.ca

Hi Tom

You are so right! That's what I told a friend if mine after doing this, I was like, hmm there's something wrong, maybe diamonds aren't perfect!? I need to make some micro bevels on the edges of the facets, you are right... The ring is from Falling Pixel, and was a .max file, converted into NURBS. It is too rough as you mention, I smoothened the geometry a bit, but it can't get perfect.

This was actually done really fast, just as a test, as I saw there's a lot of companies looking for Jewelery renderings the moment, and I realized I never did one, so just to see how hard it was...

All the best, Kim

Hi Cary.. Hmm.. so maybe I was wrong in what I just wrote above, but I actually thought the same as Tom, as you usually see these tiny bevels on, well guess "diamonds"...

-- I used a Forrest spherical HDRi files.. It was a bit geenish, so I had to correct the colors afterwards, but it gave some interesting dispersion in the diamond.. Maybe a bit too much as Cary James mention... But it can be controlled, I have 2 layers of Diamond, one with, and one without lens scattering, and then composed in PS.

Yes, I admit they were not ideal examples. My intention was showing them off clearly.
Agreed. In fact I knew somebody would come up with this fact but still, nothing -even diamond- can have absolutely perfect cut.
I've said "micro" (1/1000 mm) and maybe it's still above the toleration in producing diamonds. Can you tell us what's the biggest acceptable crease size in diamond production?
No, it's really welcome!

_________________
Next Limit Team

Hey Tom, it is really hard to say what is the largest acceptable crease or bend size.

First inspection of a diamond is usually under a 10X loupe (doublet) that is spherically and chromatically corrected (like over at the WIP forum) so that you have a clear field with no distortion. Under that initial 10x observation you should see no discernible rounding of the edges when inspecting a diamond.

It is hard to quantify because I know where you are going with your pictures and descriptions, that slight blending shows up in every other cut and polished faceted gemstone. The samples pictured ranged from molded and cast glass display pieces to natural sapphires with small rounding. I have not heard of any gem lab actually trying to measure the amount of blending and I doubt that it would be measurable by any means outside of an electron microscope.

For ultimate realism maybe you could blend the edges of the 3D model using a very small 0.0001mm blend but I am just guessing. The biggest thing for me that is missing in the jewellery/gemstone renders is surface reflectance (or lustre), diamond has an "adamantine lustre" and quartz like amethyst has a "vitreous lustre" These are hard properties to quantify but when you look at light reflected from a perfectly polished diamond facet it is quite different than light reflecting from a perfectly polished amethyst facet. I find this to be the biggest issue when looking at jewellery renders (my own included). But you are right the diamonds are a little too perfect both in cut and in surface reflectance that I think adds to the 'unreal' or hyper real look of most jewellery renders.

Just my two cents but I would love to discuss surface reflectance... and sorry Neeper if we hijacked your thread !!!

_________________
Cary James

CONTACT
www.contactdesign.ca

I see. The fact is, theoretically that crease is still there. Because, you know we're dealing with photons. So, it's impossible to jump to another surface in a different angle without meeting that theoretical tiny bevel we tend to omit. I'd really like to see the difference even that 0.0001 mm would (or not) create in the final look. I think it's something worth to check as it may lead us losing some valuable amount of rays escaping easily by the critical angle when modeled perfectly like above. Just an idea

I think lustre is all about minerals/crystals and I guess they can add further birefringence and/or alter the fresnel in a chaotic way. You can experiment improving it via blending your diamond with metallic and dielectric components in tiny different weights. Again an idea...

_________________
Next Limit Team

Thanks Tom, I think I will take you up on your mini challenge I too would like to see the effect. May take a couple of weeks but I am going to check it out.. will post results when finished

_________________
Cary James

CONTACT
www.contactdesign.ca

You would have to have a very accurate modeling program to achieve such tiny bevels. I don't think Max could do it. Max has problems dealing with such small tolerances. I think modeling at 100x scale, then scaling the object down before render might work. I don't know. Interesting, I want to try it too.

_________________
Brian Looney
Maxwell + RealFlow
http://lunatic-studio.com/
Help! I'm dispersing into a haze of probability!