Game Development Reference
In-Depth Information
3.6
Discussion
If we compare our proposed technique with the goals we set in Section 3.2 , we
can find that it manages to largely achieve all of the three aspects together:
Artist-friendliness. Artists can easily mark hidden polygons and can accu-
rately model replacement damage geometry. Supporting LODs requires a
little bit of rework, but unfortunately that is how LODs are done.
Flexibility. The technique supports an unrestricted number of damage areas
on a single object in unrestricted locations. The only limitation is that
damage areas cannot overlap (a vertex may belong to only one damage
area at a time).
Rendering e ciency. The only cost added is a few conditional vertex-shader
instructions. The cost is irrespective of the number of damage areas on the
object. Additionally, we do not require any changes to the object's pixel
shader nor its rendering states. There is no need for alpha-testing nor other
tricks that might hinder hardware optimizations.
There are many damage effects that can benefit from this technique. However,
some games might have different requirements that are not compatible with this
technique. In that case, we hope at least that we have succeeded in giving some
ideas to build upon when developing new techniques for damage effects.
3.7
Conclusion
In this chapter, we have studied the possible approaches of rendering authored
structural damage on 3D objects. A review of previous work is made with a
description of the advantages and drawbacks of each technique mentioned. A
high-level description of a new artist-friendly, ecient, and flexible technique
for rendering authored damage is presented, followed by detailed implementation
steps covering the entire pipeline from content authoring to rendering at runtime,
including a few tricks for simulating bit-testing with floating-point mathematics
on shader profiles that do not support native integer operations. The matter of
supporting LODs is highlighted, followed finally by a discussion that compares
the new technique to the ideal-case goals.
There is potential for further expansion of the technique so that it supports
the complex case of overlapping damage areas.
3.8
Acknowledgments
We would like to thank Abdulrahman Al-lahham for his help and time reviewing
this chapter.
 
Search Nedrilad ::




Custom Search