Ametrine gem material (4.1)

Started by NDenekamp, June 20, 2013, 04:56:26 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

NDenekamp

Another experiment to create a bi-coloured gem variety called Ametrine. A naturally occurring quartz with a gradient between amethyst and citrine

When digging through the forums a while ago, I saw someone requested about the creation of these types of stones, and wether volumetric colour maps were supported.

It seems with the procedural colour mapping, you can achieve something very close to this gem type! and very easy to fine tune colours, gradient, evenness, etc!  The only thing is that the Gem material type doesn't currently support colour maps, so I've had to use a dielectric type.

I've hidden the diffuse centre ball in this one, to better show the two tone nature of the material.. unfortunately I don't have some nice gem geometry to test on, is someone willing to donate one if you're interested?

Am I right in concluding that the procedural colour map textures produce 'through body' colors?


N

TpwUK


guest84672

Oh wow - this is awesome! Can you share the material so I can see how this was done?

Not sure what you mean by "through body" colors. Do you mean colors that go through the entire body, rather than just stay on the surface?

NDenekamp

TpwUk, cheers for the link! I'll see if I can use one of those models!

Thomas, I've attached the .bip file of the scene below.

I guess regarding through body colour, I was wondering if the procedural cloud data only shows when it intersects with geometry, or everywhere within the volume of the object / solid (does keyshot recognize this inside / outside of a volume?)

N

guest84672

Thanks for the scene. You can take the mapping tool and manipulate this further. I also dialed down the 64 ray bounces. 12 are quite sufficient ;-)

I think it depends on the material - but the question is best answered by our Chief Scientist.

NDenekamp

Testing on a gem stone from grabcad. Trying out the focussed caustics, but I think I need higher sample levels?


DriesV

#6
Quote from: NDenekamp on June 21, 2013, 04:56:53 PM
Testing on a gem stone from grabcad. Trying out the focussed caustics, but I think I need higher sample levels?

Aha, caustics!
Exactly my cup of tea... ;)

Is that an 'advanced settings' render?
If so, the 'Caustics' slider is going to determine the quality (sharpness) of your caustics. Even though the slider only goes to 10, you can manually crank it up to 100. It doesn't have as big an impact on rendering speed as it sounds. ;) The 'pre-render calculation' (the phase before the prepass) is going to take a bit longer though.
Also, realtime rendering can give very good results with caustics...

Dries

Ed

#7
Very nice NDenekamp -

Procedural color map - I'm anxious to try this.  I've been looking for a way to duplicated titanium anodized finishes where you see a color shift if you look at the material off axis - much like looking at the colors in a soap bubble, or the shifting colors in an oil slick on water.  See subtle color shift in reference photo below.



Ed

NDenekamp

Hey Ed,


I have looked at this finish as an experiment, but it can transitions through quite a few colours (pretty much the entire rainbow!) whereas the current procedural colour maps support only a few colours per material.

Perhaps the finish you see sometimes that involves mainly yelowish to blue / purple can be approximated... I will add it to my list of materials to experiment with further, and let you know how I get on!

Niels

Ed

Niels -

Titanium body jewelry can shift through a wide range colors, but that is because sections have been anodized at various voltages, so it is purposely multi-color.

The effect I'm after is where the metal is anodized at a constant voltage in order to produce a single color.  However when the surface is viewed off-axis, there will be a subtle color shift at the edges of one octave (I hope octave is the proper term here).  So if the titanium is a round bar and anodized blue, you can see a purple tint at the edges.  If the titanium is a round bar and anodized magenta, you can see a yellow tint at the edges.

Ed


NDenekamp

I see what you mean.

Isn't such a look already achievable in with the 'Metallic Paint' material? I've had a quick go with the magenta > fall off to yellow one, see attached.

Is that what you were going for?


N




NDenekamp

Also, Dries, I've re-rendered the gem in both advanced render mode and by giving the realtime window a good while to clear up, with the results below.

For the advance render I set focussed caustics to 100, GI to 5, shadow to 3 and samples to somethings quite high - can't remember.. others as default.

In your experience - other than the 'focussed caustics' slider set to 100 in advance render settings - what influences focussed caustics? does the # of bounces matter, shadow quality, material glossy samples, the ground plane material used, type of HDRI, etc?


N

Ed

#12
Thanks Niels -  I had not considered using metallic paint for an anodized metal  (I rarely go outside the Metals tab :)

I made a quick test on this titanium ambigram ring below.  The off-axis color shift is subtle and realistic for this metal treatment.  Every little bit of added realism helps.

Thanks for the tip and sorry to hijack your gemstone thread - it's all interesting stuff.

Ed

DriesV

#13
Quote from: NDenekamp on June 23, 2013, 11:56:10 AM
Also, Dries, I've re-rendered the gem in both advanced render mode and by giving the realtime window a good while to clear up, with the results below.

For the advance render I set focussed caustics to 100, GI to 5, shadow to 3 and samples to somethings quite high - can't remember.. others as default.

In your experience - other than the 'focussed caustics' slider set to 100 in advance render settings - what influences focussed caustics? does the # of bounces matter, shadow quality, material glossy samples, the ground plane material used, type of HDRI, etc?


N

1) The quality (shapness, unblurriness) of the caustics is solely determined -this is at least my experience- by:
*the caustic slider (advanced rendering)
*render time/sampling level (realtime rendering)
For advanced rendering, the other quality sliders have no direct influence on the caustics' quality.

2) The appearance of the caustics can be determined by several things:
*# of ray bounces (this is especially apparent in more intricate caustic setups, where focussed photons are bounced around a lot)
*material glossiness (roughness): the more 'perfectly finished' an object is, the sharper its generated caustics will be (e.g. a matte lens cannot converge light into a point, a finely polished lens can)
*caustic catch surface material: e.g. dark surfaces absorb light energy (aka caustics hardly visible); e.g. polished surfaces reflect caustics (no caustics visible on these either); a diffuse (or matte) surface with a light diffuse color catches caustics well. You need a surface that can scatter the incident flocussed light (caustics) to make them visible.
*lighting: e.g. a strong pin light (aka sunny day) generates sharper caustics than an overcast high falloff sky

Dries

NDenekamp

Ed, no worries! As you say it's all interesting stuff. I'm actually considering opening just a general (focussed on the new procedural possibilities) materials thread, so I can just post whatever material I'd like to share and open the floor for questions, share tips, etc. Also so I won't clutter the forum too much with random new threads.

As for the ring render. If you copy the parameters I posted along with my test render, but bring down the 'metal coverage' value, you can somewhat influence where the fall off from magenta to yellow occurs, should you wish to exaggerate the effect or bring it out a bit more in specific views.


Dries, thanks for the tips! at the moment I just added a ground plane, but left the material as the 'ground' type. I'll try it with a white diffuse plastic. Does the sample level for the material receiving the caustics matter much?

Too bad the license for KeyShot that I have doesn't allow me to modify HDRI's.

I'll give it another go later on.


Niels