Debunking the color of Vaders Lenses for his ANH Helmet

Doug -

I get the spirit of what you're after, but it doesn't quite work that way because you're not truly adding or subtracting wavelengths in a contributory sense as would happen with actual light beams, you're just brute force rolling the final composite around in the spectrum. In other words, there's no Photoshop way to approximate dispersion or intereference, you have to actually fire light particles in a physically-accurate simulator to know (this is what we're doing.) I'm curious about your reference photos, though... what are they? Is there an actor in in the first one, but no actor in the second one?

_Mike

Mike,
Yes the photo on the left is a still from the movie and the one on the right is
an empty stunt helmet.. If you want PM me your E-mail and I'll send you
the pic if you want it unedited..

I totaly agree with you.. My experiment was a bit crude and was only given
to show what I believe IMO to be the characteristics that add to the final coloration
of the lenses (face, and lights, or filters, or both) :)
will you be taking photos of your experiments to see how the lenses react to the different
filter settings in a camera?
If you can find a way to duplicate all the correct lighting conditions along with
the correct camera filterering. You should get very good results. Of course IMO :)
 
STEP ONE: Approximating the Lighting


First of all, as Doc says, "Forgive the crudeness of this model." Geometric precision isn't important in this test, nor is absolute accuracy with regards to lighting. What's most important is that we get the general angle, proximity and intensity of the lights correct, and we've done that.

Vader_Stage_1.jpg




I have NOT addressed the lenses yet - what you see here is simply the same lenses from the previous test. (Sort of on the way to the right color, though, huh?)


A bit about the science here: These images are being rendered in an unbiased spectral renderer. Unlike traditional CGI, where you just "fake it" until it looks right, this engine strictly adheres to the physics of light propagation and can't be altered. It accounts for all common propagation factors, like dispersion, diffraction, Index of Refraction, Fresnel, surface roughness, anisotropy, transmission, attenuation, reflectance, etc.

What this means is that if I modeled a prism and fired a virtual light source at it, it would do all the prismatic dispersion of light. In fact, here's an example of that happening using this engine:

Thomasan_Prism03.jpg


Everything in the above image is CG. So what we'll be doing is assigning a material to the lenses, in this case, acrylic, for which we have known photometric data - that is, having been previously measured in a lab, the particular ways acrylic interacts with light are known and stored in a table which this engine references, to obtain an accurate solution. We use this in product visualizations all the time, so that manufacturers know exactly what something will look like in the real world without actually having to build it.

We also have photometric data for human skin, for a wide variety of races. We'll be plugging in the average data for North American white male on the head inside the helmet.

We can also change the temperature of the lights through the entire range of Kelvin, but film set lights are usually only in a few frequencies, which we'll try. There are no colored filters on the lenses of the cameras in any of our reference shots. So what we'll really be left to focus on then, is the coloring of the lenses themselves. Now that this part of the simulation is ready, I'll begin that stage.


_Mike
 
:confused... Guys...:confused If we spent all the time and energi to confirm the coloration of a lense and put all that same energy and time on solving alternative power source. I belive you guys whould make allot of cash.

Anyway. Admire the effort you guys are putting into this.
Is it not the final ressult that matters.?

But i just wanted to say i belive you guys are taking it a bit to far now. :$
However i think it is exiting to se the experiment. :angel
 
I think this is great. This place is full of opinions and speculation and hearsay often times leading to arguments but on this thread we have hard science being used. Diggin it.

:confused... Guys...:confused If we spent all the time and energi to confirm the coloration of a lense and put all that same energy and time on solving alternative power source. I belive you guys whould make allot of cash.

Anyway. Admire the effort you guys are putting into this.
Is it not the final ressult that matters.?

But i just wanted to say i belive you guys are taking it a bit to far now. :$
However i think it is exiting to se the experiment. :angel
 
that´s one of the most scientific threads i´ve read here in the last few years and i REALLY love it!
thanks Mike, i think we all appreciate your will to put all your time and efford into this...

:):thumbsup
 
As you can see I don't post alot here and when I do it's not about costumes, but I would like to say thanks for a really interesting and absorbing thread.

Cheers,

Dan
 
The simulation still needs a lot of refining, though we're closing in on it. We discovered the acrylic lens has a coating on it, and had to simulate that first. (This is evidenced, by the way, by the non-tinted (white) reflections seen on the lenses, while the lenses are obviously colored to some degree. If not for the presence of at least one more layer to the material, reflections on tinted transmissives take on the color of the transmissive. So a truly yellow lens would have a yellow-tinted reflection/highlight. In this case, the lenses are both tinted, plus have the non-tinted reflection. If you actually look VERY closely, you can see both reflections inherent in coronas around highlights - you see 3 reflection characteristics, actually: a sharp, tinted highlight, a sharp white highlight, and a broad "greasy" reflection typical of plastics for a lot of reasons.) Here is the simulation thus far under 3 lighting conditions. Getting there...


Vader_Lighting.gif



_Mike
 
Could this be applied to the Hero Stormtrooper bubble lens question? Would be great to get that confirmed too.
 
Coloring-wise? Possibly, depending on references. In the case of Vader, we have multiple sources confirming the contributions of lens tinting, and we know there was plain human skin under there.


_Mike
 
I think it would be great if you could make another "movie" showing the process you went through to establish the contribution of flesh tone to the final lens color.
 
I'll do another movie showing the whole process once I feel the results are "conclusive," within whatever margin of error we end up with - it's not hard, the process is actually quite simple and the physics are basic.

_Mike
 
Regarding the trooper lenses, I think the prison scene is the best spot for reference.

Regarding the illustration. That's looking very interesting. Can't wait to see and learn more.

Regarding the movie, would it be possible to do it in another format than .mov? If not, then that's okay.


Thanks for doing all this work! :thumbsup
 
Yes, more information and videos coming. I had a chance to visit the Archives last week to see what I could learn firsthand.

_Mike
 
Back
Top