Type - the method used to apply
2D mapping (landscape) - this method
bases the displacement on a texture map that is known in advance.
The displaced surface is rendered as a warped height-field based on
that texture map. The actual raytracing of the displaced surface is
done in texture space, and the result is mapped back into 3d space.
The advantage of this method is that it preserves all the details in
the displacement map. However, it requires that the object has valid
texture coordinates. You cannot use this method for 3d procedural
textures or other textures that use object or world coordinates. The
displacement map can take any values (as opposed to 3D mapping,
which will ignore values outside the 0.0-1.0 or black to white
3D mapping - this is a general method
which takes the original surface geometry and subdivides its
triangles into smaller sub-triangles which are then displaced. It
can be applied for arbitrary displacement maps with any kind of
mapping. This method can also use the displacement map specified in
the object's material. Note that with 3d mapping the displacement
map's range of values must be within the 0.0-1.0 range (black to
white). Values outside of this range will be clipped.
Subdivision - this method is similar
to the 3D mapping method, with the difference that it will apply a
subdivision scheme to the object, similarly to a MeshSmooth
modifier. For triangular portions of a mesh, the Loop subdivision
scheme is used. For quadrangular portions, the Catmull-Clark scheme
is used. Other polygons are first converted to triangles. If just
want to smooth the object, without applying a displacement map, set
the Amount parameter to 0.0.
Which method to use? In previous V-Ray versions, there was a great
difference between the performance of the two methods, with the
2D mapping being faster in many cases. With
the introduction of dynamic geometry handling in V-Ray 1.45.xx, 3d
displacement has become a lot faster for similar or better quality
compared to the 2d mapping. Still, for large displaced surfaces like
oceans or mountains, the 2d mapping method might work better.
Also the 2D mapping method keeps the displacement map in a
precompiled state in memory. Large displacement maps can take a lot of
RAM. It may be more efficient to use 3D mapping in that case, since it
can recycle the memory used for the displaced geometry.