Texturing Plastic
Hello! I'm curious about one thing: which technique would you guys use for texturing plastic?
I have to create a small plastic soldier, like the ones in Toy Story. The model is done, but i'm analyzing how to make it look good. I have completely green diffuse, baked normal and no idea on specular.
I have to create a small plastic soldier, like the ones in Toy Story. The model is done, but i'm analyzing how to make it look good. I have completely green diffuse, baked normal and no idea on specular.
Replies
You can do this with either a grey specular or maybe even a specular that's a more neutral saturation of the inverse colour of your diffuse (or maybe even slightly blue? who knows!).
Then as shown above those are fairly glossy - the specular highlight is very tight. In general, "hard" plastic like the soldiers are made from has quite a tight specular, so make sure your glossiness value represents that.
MoP beat me to it.
If you look at the image the surfaces that are flat/smooth and perpendicular to the camera have broadish highlights, were as the reflection in the cracks/folds is much tighter. I suppose a normal map would help break it up, but without view dependent shading it wouldn't look as good, as it's not really as evident in games engines.
I would say an asset which is all one type of plastic does not need a gloss/exponent map. You would only need one if it's made of multiple types of plastic which have different highlight widths.
For those army men I'd just try to use a shader that has diffuse, normal, specular and a constant gloss value, plus the environment cubemap reflection masked by fresnel as I said earlier.
So im going to contradict myself and agree for that particular object, you could get away with just a uniform value across the surface, but still think a gloss map would help.
As specular is basically a cheap form of reflection, i think it makes sense to try an replicate these tighter/broader hotspot reflections using a gloss map.
Although your suggestion about fresnel with a cubemap, i think would work very well.
I think alot is overlooked in the lighting stage aswell.
If your lighting is really hard and hasn't been diffused someway then the highlights are gonna be much stronger with harldy any reflections as they'd be washed out, were as with soft lighting as in the picture the highlights are coloured because its picking up the reflections of its environment.
hello there,
do you have some real life example of materials that produce broader or tighter highlights depending on the view angle? Seems weird to me but it could be, and it would be easy to simulate even without a cubemap. - and Mop fresnel is something else, it's all about intensity, aren't we talking 'glossiness' here? (fresnel does help a lot for plastics tough
I like the idea of using some sort of ao/cavity map as spec power map, it makes a lot of sense for small objects that are likely to be rubbed during hours by greasy little fingers
Not quite happy, but getting there.
There are 3 light sources and an environment map for the ambient light + reflections as well which I could not fit in the screenshot. Using a Mental Mill generated CGFX shader.
I don't think many materials change their glossiness by their viewing angle per sa, but maybe stuff like hair, skin and things like brushed metal that have anisotropic reflections. Basically any surface that has grooves of rough surfaces.
I guess this is more revelent to Fresnel as thats what it's supposed to do isn't it? Mask out channels according to the angle of incidence?
By view dependent, i was refering too how the specular changes its position on a object as the camera moves around the object.
I.e. the spec highlight still shows when not hit by direct light.
If you load a sphere up in your 3dapp and apply a blinn and de-activate the specular's "view independent" option, you should see the specular hotspot disappear as you rotate around the sphere.
Well this is kinda unrealistic as the spec would always be visable to some degree no matter what angle you viewed the sphere as it relies on the camera angle and not the lighting.
What i was trying to imply was that by using gloss maps to break up the uniformity in shading as the camera moves.
So have tight highlights in the creases and more broader highlights on the larger surface areas.
Maybe im talking out my ass
Specular is (roughly) the dot product of the light vector, the surface normal, and the camera vector. So it's definitely dependent on the camera angle.
But in this case I'd just use a blurry reflection map. The specular colours you see in the original image are reflections of the blue background and the orange ground.
There also seems to be a bit of subsurface scattering going on. So it'd be good if you've got some way to simulate this.