A-team test
Posted: Thu Jan 19, 2006 8:39 pm
by Pietro Spampatti
It will be really nice and usefull to really hunderstand the material settings if the A-team will share scene like this one
Tom can you do it?
Thanks
P.
Posted: Thu Jan 19, 2006 11:34 pm
by BEM
Or a little tutorial !!!!!
BEM
Posted: Fri Jan 20, 2006 8:04 pm
by Mihai
Well, first do you know what a weightmap is, and how you use it in the material editor?
Posted: Fri Jan 20, 2006 8:07 pm
by rivoli
Mihai Iliuta wrote:
first do you know what a weightmap is
what is that mihai? a tricky question?
Posted: Fri Jan 20, 2006 8:18 pm
by Mihai
no it's a normal one
Sorry if it sounded strange, but just want to know where one needs to start.
I'll write it as short as I can, if something still isn't clear post your question
Like the caption in the glass image says, use a weightmap (greyscal RGB image) to blend to bsdf layers. You need two versions, one being the inverted version of the first one.
A weightmap controls the influence of a bsdf layer. If you understand this concept then lots of things should fall into place.
About the leaf, there is a detailed post about it in the Maxwell forum, just search for clipmaps and look at the most recent posts.
Posted: Fri Jan 20, 2006 8:20 pm
by rivoli
Mihai Iliuta wrote:
no it's a normal one
Mihai Iliuta wrote:
About the leaf, there is a detailed post about it in the Maxwell forum, just search for clipmaps and look at the most recent posts.
here's another one about blendig glasses using weightmaps you may find useful:
http://www.maxwellrender.com/forum/view ... hp?t=11317
Posted: Fri Jan 20, 2006 8:33 pm
by Pietro Spampatti
Posted: Fri Jan 20, 2006 8:48 pm
by rivoli
figurati, piacere mio.
Posted: Fri Jan 20, 2006 10:06 pm
by rivoli
Adam Trachtenberg wrote:
Each layer needs a weight map? If so it seems like extra work for no reason. It should work like PS layer masks.
it maybe will at sme point, at leas this is what one may guess reading one of mike's post:
We also have a layered-shader implementation known as weightmapping, which will evolve into a traditional stacked-layer shading system.
Posted: Fri Jan 20, 2006 10:14 pm
by Mihai
Adam, you're right, with the tinted glass example only one weightmap is needed. Sorry I got it mixed up with another method for making 'fake' clipmaps.
Posted: Fri Jan 20, 2006 10:51 pm
by rivoli
Mihai Iliuta wrote:
Adam, you're right, with the tinted glass example only one weightmap is needed.
really? i tried them both (1 mapped weightmap and 2 mapped weight maps), and the tinted glass with 2 inverse maps looked better, at least in the mat preview.
Posted: Fri Jan 20, 2006 11:04 pm
by tom
Use 2 weightmaps inverted

Posted: Fri Jan 20, 2006 11:05 pm
by Mihai
hmm.....have to check that, but the weighting should be normalized internally. Or it could be that if you're just using 1 map, at the transition area, the material without the inversed map has too much influence. If both have maps then in the transition area it will look nicer, a 50-50 influence. But using two maps is not necessary.
Posted: Fri Jan 20, 2006 11:12 pm
by rivoli
Mihai Iliuta wrote:
Or it could be that if you're just using 1 map, at the transition area, the material without the inversed map has too much influence.
well, that's what i thought. if the unmapped one is set on 100 the total amount will go, following a gradient, from 100 to 200 (?) where the map goes totally white. if set on 50 it'll go from 100 to 50 (where black). in both cases, and what's in the middle, it doesn't seem correct.
Posted: Fri Jan 20, 2006 11:19 pm
by tom
How do we do normalization in weight values:
Example 1: 50, 50 OR 7,7 OR 123,123 ... means 50% , 50%
Example 2: 50, 150 OR 8, 24 OR 42, 126 ... means %25, %75
Mapped Weight is the same calculation per pixel, so we need to use
complementary maps for blending the layers equally. A basic black-to-white
gradient will require the inverted one on the other layer to participate equally.
If you only use 1 gradient map and the other one at a constant value 5.
The equation will be like this:
Layer 1: 0,0,0,0,1,2,3,4,5,6,7,8,8,8,8,8 (basic representation of pixel intensities)
Layer 2: 5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5 (constant weight value)
Result: 0 and 5 , 0 and 5 .... and so on... then the value 8 will be normalized down, due to 8+5=13.
The correct mapping for this purpose should be:
Layer 1: 0,0,0,0,1,2,3,4,5,6,7,8,8,8,8,8 (basic representation of pixel intensities)
Layer 2: 8,8,8,8,7,6,5,4,3,2,1,0,0,0,0,0 (basic representation of pixel intensities)
Result: 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8