Everything related to http://resources.maxwellrender.com
#209754
Tom, Mihai and all .r2 file gurus.

Is it possible to use the r2 file system to replicate a ramp shader?

If I layer multiple bsdfs with different 0' & 90' colours, which use r2 files that overlap, would that give me the same control as a ramp shader effect?

eg:-

layer 1 r2=90 60
layer 2 r2=65 30
layer 3 r2=35 0
etc

There aren't any r2 files in the material database, which have a second numeric different to zero, so I'm really guessing here.



In my tests, a material will render with r2 values above 90'. Is there an angle limit for the fresnel effect?

Many thanks in advance,

Tim.
User avatar
By 3dtrialpractice
#209761
just throwing in toms reply form his velvet thread..

as you said that there arnt r2 files with second value other than 0...

What you seem to propose in your setup probably may work!?!.. Great idea BTW!

I asked a question and his reply - may help you with the second number q before someone else with better idea towards your specific q replies to you..

--------------------------------------------------------------------------------
TOM QUOTE:

Only 2 numbers allowed and they should be seperated by a single space.

Like:
25 0

This means:
Hold frontal reflectance between [0-25) and ramp up fresnel between (25-90]

The second parameter (0) represents the fresnel coverage versus roughness, see above graph for more detail. Here 0 means apply fresnel for all roughnesses.
User avatar
By Tim Ellis
#209765
3dtrialpractice wrote: The second parameter (0) represents the fresnel coverage versus roughness, see above graph for more detail. Here 0 means apply fresnel for all roughnesses.[/i]
This is the bit where I think my idea falls down. My misunderstanding of the second value in the r2 file.

My thought was it is the lower limit of the reflectance angle, not the fresnel coverage versus roughness value.

So how and when would I use the second r2 value? What's the numeric range calculated?

Cheers,

Tim.
User avatar
By Mihai
#209766
The second value controls how the first value is influenced by roughness.

The first value is a falloff angle between 0 and 90 reflection colors. So if you specify 10 for example the transition between the two colors will happen much more abrupt way. Putting the second value at 0 will keep this effect, even if your material has a high roughness.

I'm not sure you will be able to achieve what you want, you may still need to use some weightmaps.
User avatar
By 3dtrialpractice
#209901
tim- ya thats why tom answered my q is cuz I too wanted to specify the angles in both numbers....

Mihai- you suggest using weightmap. .currentlyh there is no way to make weightmaps be based on "facing angle" to the camera right? .. maybe we wshould put this as a wish... to have a check box that mas the shading or weightmaping be relative to facing the camera.. cuz right now weightmaps rely on the set uv layout.. so maybe have the uvs be projected form camera onto another uv channel? but this would work on still frames only.. not animation?

any suggestion on how to to afacing towards the camera weightmap.. ? or am I missing somethig in the basic essents of the material shading/weightmapping that already does this


but still i think tims got a good idea to have multiple bsdf layers with differnt r2's "blended".. a while back i wanted to- but all the different r2 values in one r2 file.. tim's approach would be to just have multiple layers with dif r2 setups..
User avatar
By Tim Ellis
#209971
Thanks for clearing it up Mihai, I thought my idea was slightly flawed.

3dtrialpractice, can you not map from the camera view before rendering with Studio? Not sure of your 3d app, but I can UV map project from view, with Blender. I load the camera view and unwrap.

This has been my only possible solution for this type of mapping. Alas it falls down if you rotate or move the camera, but for a static shot it might work.

I do some more tests...

Tim.
User avatar
By Mihai
#210038
Yes, with weightmaps you can't control them with camera facing angle, they are simply applied to uv's. These kinds of things is were procedurals would be very handy.
User avatar
By Tim Ellis
#210212
Mihai, surely fresnel works to the camera anyway, in a similar fashion to procedurals?

My 'ramp shader' tests.

The idea is to start the fresnel curve for each bsdf layer, at a different angle, thus faking a ramp shader. Due to the r2's method of affecting only the 90' reflectance colour & angle, the 0' relfectance colour needs to be the same for each layer and we use the 90' colour for the different sections of the 'ramp shader'.

Multiple bsdf layers, each sharing the same 0' reflectance colour, but with different R2 controlled 90' reflectance colours.

Here's a small R2 test scene I've created. http://www.4shared.com/file/10701138/1c ... files.html

Included are a couple of multiple layered R2 mxms which are shown below:-

R2 mxm 1:-
Image

R2 mxm 2:-
Image

Mxm 1 has a low roughness level, hence the reflections.

I've also included all my R2 files from my tests.

The strong colours are for visual reference to show the effect.

More tests soon.

Cheers,

Tim.
By Mr Whippy
#210710
I've tried using this to create the "Rage" type paint effect, where it goes from light green, through to blue to purple.

The problem is that the first layer needs to go from green to blue, but then the second layer with the pink falloff also needs to be green to pink, so the blue interim colour is mixed with green up until the new falloff point, and even then the pink is mixed with blue at the edge...

Ultimately the green is the only colour that comes out as you want it, with the others dull mixes.


In reality we layer the paints with a thickness, and as we look through more material to the edge we get it's colour coming through, not through cheating with falloff's and Nd's.
The real paints thickness is what controls this in my example.

I made the paint work well if I modelled it per layer as per real life but it was noisy and took a long time.


What we really need is a proper assesment of what is happening with these materials so we can simulate them as they occur in real life, not just bodge them with R2 files and reflectance colour tweaks.

Hopefully in a future version we will get proper layering using thicknesses and some other blending controls of the BSDF layers to make ones further up take colour presedence etc...

Right now the edge falloff is a mix of all the other 90 degree reflectance colours.

Dave

what about gpu maxwell q project?

SS Pinto Bean

Hi Tommy, Great stuff - love it~! Thanks for pos[…]

Never No More Studio Lighting

Hello Mark! Very good tips about the camera setti[…]