does DGS Shader "UV shiny" do anything?

[Brent Turbo]

IMO, the dgs_material isn't ready for use straight out of the box. With that in mind, the best way to get results from it is to crank up the sampling in mRay's render globals. Like a min of 2 and a max of 4. Takes forever to render, but that's the chance you've gotta take when you want expensive raytraced effects.

For testing these type of effects, it's best to set up a test scene (sphere, plane, spotlight) that renders quickly, and then export the material into your scene for tweaking. I recently used dgs_material in an animation, and later ditched the effect in favor of Maya's default renderer because frames were taking 2 hours each at NTSC res.

[joie]

I know everithing about DGS and I use it sometimes.
The only thing I don´t understand is the "UV Shiny" attribute, it seems to do nothing, reflection appears the same with UV Shiny at 0 and as 100000000, and I don´t know why it is.

Anyone?

Thank´s for the info anyway.

:)

[DiGiman]

That is what they are supposed to do.
I have gotten very mixed results with them though. First of all, if you enter a number in either of the U, or V attributes it will void any input you have put into the standard shiny attribute.

Second of all, I have read somewhere that the U, V attributes dont work well on nurbs surfaces. I have gotten decent results on polygon surfaces, though I think the UV's of the surface would have to be set-up, either as a map or projection.

Lastly, the shiny attribute only affects the glossyness of the surface. And the way the glossy works, is reverse of what you would think it to be. The lower the number, the more blurred the reflection.
Because the shiny doesn't affect the specular of the DGS material, keep it low, or off. The specular is refering to the reflectiveness of the surface, and if it is turned up, you wont see the anistropic scratches though the reflection.

Hope this helps....

[trthing]

I've been trying to get this look

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial5.asp

with DGS to no avail. Did someone else try this already with some success?

[Brent Turbo]

Originally posted by trthing
I've been trying to get this look

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial5.asp

with DGS to no avail. Did someone else try this already with some success?

To get that look with a straight dgs_material would be pretty painful. For any brushed metal surface, it's best to start off with an image map of brushed metal -- that gets you in the ballpark immediately. The rest comes down to specular shape, the behavior of the reflections, and perhaps a little bump. All of these attributes are difficult, if not impossible, with DGS.

[trthing]

Oh, well...

I was really hopeful the physica accuracy of MR shaders would lead me to a more elegant solution than that...

Pity... but thanks anyway... :thumbsup:

[brunner]

Have you tried using the mib_illum_ward shader?

[Jozvex]

Originally posted by Brent Turbo
IMO, the dgs_material isn't ready for use straight out of the box.

Why not?

And I've attached an image I made using all DGS shaders and FG for lighting. There are three spheres in a room and from left to right the spheres are:

Left: Shiny = 15
Middle: Shiny U = 15
Right: Shiny V = 15

They seem to create the right effects and look pretty cool! Rendertime was 13 minutes on my 1.5 gighz computer. The sampling was at 3 5.

I know 13 minutes is a long time, but hey practically the whole image is glossy reflections!

:thumbsup:

[joie]

Jozvex, I think is my fault, then, if you want U shiny you have to avoid the "shiny" attribute?, look at this matrix:

if "shiny" >0
"UV Shiny" = 0

else "shiny" = 0
"UV Shiny" = whatever you want

I means that if you want UV shiny you have to let shiny at 0?

hope it makes sense... :P

[Jozvex]

Yes, you are right, that's how it works!

I think the DGS shader's great!

:)

[Tocy]

Hi!
Have you tried the reflection utility by Horvatth Szabolcs? I think it's great for blurry reflections, gives very good result, in a fraction of dgs' rendertime. It has it's own sampling parameter, so you don't have to oversample the entire scene! And has some other cool features. It can be found on highend3d, in the shaders section.
Hope you like it!

[trthing]

Originally posted by Jozvex
Why not?

And I've attached an image I made using all DGS shaders and FG for lighting. There are three spheres in a room and from left to right the spheres are:

Left: Shiny = 15
Middle: Shiny U = 15
Right: Shiny V = 15

They seem to create the right effects and look pretty cool! Rendertime was 13 minutes on my 1.5 gighz computer. The sampling was at 3 5.

I know 13 minutes is a long time, but hey practically the whole image is glossy reflections!

:thumbsup:

Is there a chance of you attaching the scene here for us to take a look? I really didn't get anywhere near your results with _u and _v settings....
It would help a lot! :thumbsup:

[Jozvex]

Ok sure, it's attached!

:thumbsup:

[trthing]

Thank u very much!

Will give it a try here...

How do people say here?

U roxxx :buttrock:

[Brent Turbo]

Originally posted by Jozvex
Why not?

Mainly because I've never had the need to render objects that are just colored, reflective geometry. Try adding some image maps and bump to the DGS and you'll see why I think it's not ready out of the box.

[joie]

I don´t know why but I get different results (or even no results) with different type of geometry, NURBS render well (fantastic), but Polygons render bad and I don´t know why.

Check the image attached, the bottle on the left is made of NURBS, and the right one is Poligon, why is this?, and yes, it has propperly set its UVs.

Please help!

[brunner]

Originally posted by Brent Turbo
Mainly because I've never had the need to render objects that are just colored, reflective geometry. Try adding some image maps and bump to the DGS and you'll see why I think it's not ready out of the box.
Why not? :)
There's no problem with image maps, bumps and dgs_shader :)

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.