Sep 05, 2003
Post id: 153757 Report Item

very good start ickyptang,...

for "ior" physically correct values can be found on the internet. Am i am right that there are no such values for

glossy and specular ?

Sep 05, 2003
Post id: 153758 Report Item

Very good. But theres just one thing i have to comment.

While physcs state that theres something that can be called 0% light or absence of visual photons theres no such thing as 100% light since theres no known high bound for that (ok space IS quantized so there might be one but its ridiculously high but ill leave that to hardcore quantum physicisists... hence the hdri format!)

While people tend to limit their things in 1 to 0 range there are cases whare the range is in fact bigger!

Take a very typical outdoor scene where a car has so shiny reflections of the sky. You might alos notice that the ground is very dark in theese reflections. This is due to the fact that sky and clouds on it is usualy so much more brighter than the surroundings.

To make things even more complicated the eye doesent work on a totaly liear scale... or even on comparable one, it is autoexposing wich means that it will adjust. So do atest and se if you can spot this its hard is only noticable for biref brief moment. Look at some resonably closeby trees at the edge to the sky. block the sky aout of your mind. Now concentrate on the sky for a moment without paying attention to the trees. Now switch back if you have a GOOD color memory (most of us dont), or are extremely atuned, or try a lot (this last one migh actualy be just a mindtrick). you will notice that the vivd green is a shade of much darker shade than you initialy tough.

oka nyone interseted in duplicating this contact me... Ill asnwer the 20 first ones!

Sep 05, 2003
Post id: 153759 Report Item

In my "100% light" section I was just illustrating that WHATEVER the value of your light source, you can not reflect more than the initial value... a 310% reflection of the light source more than triples the amount of starting energy, which is completely impossible in the real world.

Also, I know that a 0-1 scale when dealing with light is false. These days I only use Mental Ray's physical light, and trying to get ANY result using a 0-1 scale with a physical light is something of a joke ; )

Thanks for the insight, though.


Sep 05, 2003
Post id: 153760 Report Item


thats a nice chart for index of refraction values...

As for "correct" values for specular and glossy, I don't think there is such a thing as "physically accurate" values, since its still just a fake to achieve real-world looks.

In reality no surface has simply 1 standard value for the amount of microfacets on its surface... There are simply too many variables to take into account.

For example, even 1 piece of ivory or wood or whatever from the next, taken from the same source, will vary in its "look" if you really examine it.

There MIGHT be a list of common materials and the best ways to fake them using a diffuse/glossy/specular approach, but its probably just easier to think of/look at the material you are trying to simulate and analyze what properties it has.

For example, an white-plastic surface may appear as (mostly) white with a slightly blurred highlight... Therefore you set the diffuse value to a similar white color, put a glossy value fairly low, so you get a highlight but not much reflection, and set the shiny appopriately...

As a side note, you may want to try to keep the values of your diffuse and glossy/speculars beneath 1 (i.e. in HSV mode if your glossy has a .05 V, make your "white" diffuse a .95 V...)

You don't HAVE TO do this (in that nothing prevents you from doing otherwise), but I have found it helps to keep a "real" sense to your shaders.

Hope that helps : )


Sep 05, 2003
Post id: 153761 Report Item

Actualy some realworld surfaces reflect more than 100% of light coming in well sortoff. Since they actualy convert for human eye outbound light mostly ultraviolet to visible light....

Many pure white surfaces to be exact.

Hence blacklight effect with blacklight! and other realy cool colors like ambulance orange out of monitor gamut offcourse but not film!

but dgs is not very close to how it realy works it is after all a sipmlified light modell like labert (the standard old method beig quite far away on how it works) etc... just more accurate nut quite tough. And still off for many surfaces im afraid. lots of things you and me se every day it cant do!

but i guess this is academic tongue.gif

Sep 05, 2003
Post id: 153762 Report Item

Well, ultraviolet is a whole other realm... we are just dealing with visible light here, as thats all standard cameras can capture, and 3D programs only mimic standard cameras... they don't even typically mimic the human eye... just cameras.

Moving on, nothing really bounces more light it recieves, rather, in some instances surfaces can create light of its own via various processes which are catalyzed by light itself.


That site explains how black-lights work... the surfafces being struck are not bouncing more light than they recieve, rather, various phosphors coating the surface react to the light as it strikes them, and these phosphors emit light on their own.

The real world is a much more complex place than it might initially seem ; )

DGS was not designed to simulate extra surface attributes like phosphors and such, rather just to simulate how a light source will affect the surface it is striking...

But, in the end, there is no way that anything in the CG realm will 100% replicate the real world, as the real world is far too complex...

DGS does a damned good job of simulating the most simple lighting scenarios...

phosphors and other factors aside : P


p.s. yes, DGS is off for "many surfaces" and doesn't do "lots of things you and me see every day" but its just specialized for the more common conditions...

Thats why there are other shaders, like dialectric, which handles more specialized "light passing through transmissive media" situations.

There is no "be-all, end-all" shader, DGS just works very well for many common situations, as it was designed to do.

As a side note, if anyone wants to design a Mental Ray UV Light shader and a phosphor simulating shader, that would be quite a project ; )

Sep 05, 2003
Post id: 153763 Report Item

Yup. I mean a incoming lightbeam might excite a elerctron to jump back and forth on its path hence making a qyantified backlash in form of a photon.... Bounce it back and fortha few times. Interference the beams. Lag the emission etc etc. Complex very complex.

Anyway snow is one of hardest surfaces to do beacuse it does most of all the ffects that you can imagine its subsurface scattering varying contains some uv to light conversion, may contain some interferencin factors etc.

So maybe a good snow shader would be a good start.

3d progs dont even mimic a camera very well.

PS. maybe its time for me to shut up and hit the bed!

PPS. The way standard lightmodells work in physics are wrong! theres a few correction factors to account for. But i cant say more because its a trade secret. not 3d but rather ultra ultra high tech sensors!

PPPS. one more thing any owner of a extremely good optic know that a surface can pass light and reflect that is summed more than 100% but thats jus some lag inside things so...

Sep 05, 2003
Post id: 153764 Report Item

I realize that the 3D world is nothing but a massive fake... my point is that the traditional fakes completely and totally ignore the way real light works.

DGS and other such shaders and effects (such as GI, physical light, whatever) attempt to bring 3D into a realm closer to reality... its still not going to be "real" or truly "physically accurate" simply because the entire 3D world is based on a system that was invented to mimic reality in a completely artificial way.

Plus its still something being artificially created and simulated in a computer... primarily for the entertainment industry, which is not terribly grounded in reality anyway ; )

I would think that its pretty obvious to most people that the 3D world is not the real world : )

Oh, and as for your trade secrets, if they're so secretive, you shouldn't even mention that they exist! ; P

Just kidding.

Have a good night, Joojaa.


Sep 05, 2003
Post id: 153765 Report Item

Not realy trade secrets the implemantation however is. Hawent worked for the company tough. Have been there for about 10 excursions however and they showed some stuff thats realy wierd.

Sep 07, 2003
Post id: 153766 Report Item

How do you actually USE DGS shaders in Maya? How do you create, assign, etc..


Sep 08, 2003
Post id: 153767 Report Item


Sep 09, 2003
Post id: 153768 Report Item

Thank you, I am well aware of my nerdhood, and flaunt it accordingly ; )

Anyway, as for how to create/assign DGS in Maya, I plan on making a full tutorial soon... for now, though, you can look at an old "tutorial" of sorts I did here...


its on there a little ways down.

I intend on doing a new one, definately with line-breaks to make it easier to read, and preferably with images and such, but for now that'll get you the basics on how to start.

For now, I hope that helps. : )


Sep 09, 2003
Post id: 153769 Report Item

Would like to see that tutorial very much.

Here is one render with DGS shaders for wine

and glass, light is physical and connected via

explicitre connection to material and photon shaders.


Harsh highlight on glass is not a reflection/diffusion pane, but area rect physical light. I need to make it

more organic without losing other good things, thats next to do.

Sep 10, 2003
Post id: 153770 Report Item

lookin' good... by the way, to my knowledge you actually don't need to connect the light to the DGS_photon, since the photon shader is there simply to tell the photons how to behave... it in no way affects the direct illumination.

As a matter of fact, if you don't connect anything and render you will see the GI effects without any direct illumination at all... it can be a weird effect to say the least ; )

But, it certainly does not hurt to connect the photon materials to the light, its just extra work... anyway, I am glad you shared your work, and keep going with it : )

I doubt my tutorial will have any info in it that you don't already know, but even so, I will make it as soon as I get the chance.

Have fun DGS-ing!


Nov 12, 2003
Post id: 153771 Report Item

Hey, folks, I have been horribly busy with school, and thus have not had a chance to put up (or even make) my full DGS tutorial. With any luck I will be able to get to it soon, probably once I get on a break from school...

I know, I know, I am like 3 months late, but bear with me ; )