Fractured Amethyst

LuxRender 0.8
4,2 kS/p

Its looking good. I haven’t seen so many diamonds and other gems so I don’t know exactly how they look.
I know what a amethyst is (I actually have one) but I’ve never seen one in this shape, so I’m not sure how it would look.
I would however believe that the color would be more purple-ish and cleaner on the inside (more transparent).
Maybe you could also check how it looks with translucency on.

I think that the gem looks really classy. GJ! :slight_smile:


Yeah there are some fractures just white and high SSS with scattering g=0 direction (like milk). I did lower it for more transparency.

The absorption specter is quite transparennt, maybe i need to rise it up a bit like 2x - 3x.

Ohh, shiny. I also don’t know what an amethyst should look like apart from other generic gems, but it does seem a little solid to me. The fracturing looks good though. I’d certainly pick it up if I saw it on the sidewalk.


Yes it’s not really fractured but fractured because to use multible materials / values on all the pieces like absorption, scattering, dispersion, textures.

Pretty nice concept, though yeah, there is a reason why the diamond shape is mostly used for diamonds. It has to do with the refractive index of diamond that together with the shape optimizes the inner reflectivity and thus the glimmer-effect.
Also the color is way too weak (for most representative amethysts)

Yours look nearly like non-defective crystal

It only becomes purple under a certain kind of radiation.

It looks like reall model in wiki has much larger scale, absorption is a bit higher,
SSS is a lot higher. Also i need to work on that how to get that white smooth to sharp gradient milk stone looking thing. Maybe it’s not pure amethyst, it also looks like double refraction effect (like moonstone has).

Actually i need smoke inside volume and that smoke would be scale texture for the absorption specter and or for the scattering amount and direction.
How can i define smoke inside complex volume so it will not be visible outside volume?

the white purple-white stripes are a nice effect that are probably relatively difficult to get into a nice looking shader.
Basically, it’s unpure amethysts that grew under varying external radiation, sometimes looking more like quartz and sometimes more like amethysts.
I guess, some kind of blend-material, textured by a little higher contrast stripes of roughly equal width and direction would kinda work.
Though, those often are so turbid that it’s even less realistic in terms of how you’d sell them, to use the typical diamond shape…
I remember that I did an Amethyst render a long time ago in Indigo…
Wow, my English improved a lot since then. And already back then I thought, it was good <.<
Either way, the renders over there are roughly how it should look like (I guess), in case of clear, pure, “perfect” amethyst. The crystal shape is more realistic in the sense mentioned above, as well.
(There are amethysts that grow in both directions at the same time, though they are quite rare)

The color I chose was very empirical: I chose what Wikipedia defined as “Amethyst” as a base color. Additionally, I accounted for the transparency by filling an image with that “Amethyst” color and used “color to transparency”, to “remove” once black and once white from that color. The RGBA values that resulted from that, where what I based my RGB values for the color on.
Then I blended (in case of the first to images in the thread above) two materials, namely the “bright” one (removing black) and the “dark” one (removing white) to obtain the images you see.

Apparently, the Wikipedia articles did change a little since then as it doesn’t directly give RGB values for “amethyst.”
Though, just in case you’d like to do a similar thing, xkcd had a pretty interesting empirical color survey
which resulted in a table of the 954 most common color-names, together with their respective average colors.
One of them is amethyst: #9b5fc0
Seems like a perfect fit in my eyes.

Note that my renders still turned out a little too transparent. I directly based my absorption coefficients on the Alpha value. Probably, they should have been multiplied by a factor… or simply chosen arbitrarily until they have the right look and feel.
The IoR I used was directly from Wikipedia.

Hi kram1032

For the color (absorption) i used measured spectrometer data from
It’s also here for indigo

It’s quaite transparent, so maybe i did wron when i scaled the spectra with values 0.1 - 0.5 - 1.0 - 1.5 - 2.0.

In luxrender i have the data in txt file.
wavelength: nm data: absorption (1/m)
379.6595 91.8840519
380.2573 91.4293355

719.9030 39.3684117
720.5050 39.3385413

Then i scale for some sub fracture the same data with different value like that in materials file:

Texture “AmethystFrac-Abs” “color” “tabulateddata”
“string filename” ["…/Scene/00001/003 - Absorption - Bolivia Amethyst.txt"]

Texture “OOOWEEE-2” “color” “scale”
“color tex1” [0.50000000 0.50000000 0.50000000]
“texture tex2” [“AmethystFrac-Abs”]

The gem measures are:
diameter of 3.3 cm
In height 1,74 cm

I got the colors better by tuning scattering and absorption, but it’s very noisy to post here now.

This reall amethyst gems are soo much better looking arrrrrrr :evilgrin:

Back when I rendered my Amethyst in Indigo, it had troubles with transparent data. It only was able to do reflective data.
There already was diamond and water and such but they turned out to be nearly black. I guess the reflections where correct but no refractions^^
Using meassured data can’t be beaten, I guess… Though then the data must stem from quite pale amethysts…
Heh, directly comparing the data renders and my own Amethyst material, I actually must have gotten REALLY close :slight_smile: Mine seems a little more pinkish.

How can you do messured IoR and still adjust absorption? Isn’t that included into the messurement?

Yes measured IoR can be used in LuxRender from nk file for transparent volumes, or if you have Sellmeier’s 6 values. Actually even Cauchy A and B works. But from nk file, i think it only reads n, not the k (extinction coefficient) and other values. Maybe i’m wrong.

How can you do messured IoR and still adjust absorption? Isn’t that included into the messurement?

They are 2 different measurments.

I know how it works, but… oh, all the k are 0. didn’t see that before.
Well, then you can only guess the absorption :-/

I’m not sure i got you right, but absorption is in different file.

Huh, but then it should be right.
And in that case, how could you adjust absorption (without touching those messured values)?
The k are a messurement of absorption aswell…

I scale absorption like that:

Oooh, there is a multiplier, I see.
Well, then everything is clear.

Any progress on the actual project? Heh. I guess, this went a little offtrack <.<

I did restart rendering meny times after tuning this and that material.
Then i removed 2 Gb of memory from my main machine and installed it in another machine now i have enough memory to use slave to help to render (2.4 x speed up + another 2.1 x speed up by removing very slow glossy translucent material).
Still the main scene is 2,9 x slower compared the scene where i had just the amethyst.

A lot noise, it’s just @ 34 S/p

I think i get ~ 300 S/p in 1 render day that’s for me 15-16 hours.

Not much of a scene but i’m more interested in the caustics and the materials, what will they look.
So for that i use nice rendering settings for caustics like:
BiDir and high path depth (128) and close to none russian roulette threshold (0.98) and low largemutationprob (0.1) and high MLT maxconsecrejects (11264).

Also i added very small light source in addition to HDR, BiDir makes nice caustics with that.

Then i scaled the noise size of some shaders that texture the scattering of volumes.

I also tried to texture the IOR from blender shader but no success so far.
I did it because usually there are mixed substances in reall life that have different optical properties.

There is only 2 image file textures now (HDR and bump on mp3)

So this is more of a caustics and transparent material test…
A nice one though.

Thanks :eyebrowlift: