Shading network
Now that you understand the necessity of having good UVs and a skin shader that reacts appropriately and can give the appearance of fine skin wrinkles when it deforms, you are faced with the challenge of rigging it up altogether. Luckily, the process is made extremely easy using built-in shading nodes. Similar to how deformers work automatically on every vertex in a 3D model to deform it, a shading node works on every sample in an image to process it much like a compositing operation during render time.
Basically, in order to make this whole process work as it should, your shading algorithm should echo exactly your deformation algorithm.
With this knowledge, you can go ahead and create a shading network that emulates the behavior of blend shapes. Then, use the multiplier that controls the blend shape attribute for a particular modeled shape to drive the multiplier that controls the calculated difference of a displacement map painted especially for that same modeled shape.
The simplicity of the blend shape algorithm can be built directly into the shading tree, as long as the original map is always started from the base map when modifying the new maps that get created. This is a perfectly reasonable requirement, since blend shapes already work that way. So, then all you have to do is subtract the original map from the newly painted map, making sure you subtract them in the right order so you are not getting incorrect negative values. You now have a difference vector, just like a blend shape has. Then you take the output of that subtract node and multiply it by the actual blend shape value. Next, all you need to do is take this output and add it in with all the other difference outputs from any of the other blend shapes, as well as the base shape. The outputs of this single add node should then be connected to the displacement and bump map's input channels.
The network explained above is delineated in the following Hypershade network. It uses a simple example with only two blend shapes for the sake of clarity. You can have an unlimited number of inputs coming in to that final add node that plugs into the displacement shader, and they will all work exactly as they should.
The shading network used to extract proper displacement values j. Movie: Watch the movie drivenDisplacement.mov from the support files to see an animated example " Ti'iv of this technique.
Get the difference for this shape map
Get the difference for this shape map
Add all the differences to the original map
Blend shape node and animation
Feed the results In the displacement and bump maps
Multiply the difference with the blend shape value
- tesson 17 Driven displacement
The rendered final effect
The rendered final effect
The rendered movie from the support file is only a test with low quality render settings. If certain approximation settings in mental ray's displacement subdivisions and raytracing settings in the Render Settings were made higher, there would be even richer shaded detail in the geometry.
Average user rating: 5 stars out of 1 votes
Post a comment