You should be able to pack as many bones + weights as you want in your vertex. If you're on relatively modern hardware (DX10-class or higher) you may want to branch on the weight being > 0 before fetching the bone matrix as an optimization.

eh?, i was under the impression that overall it's cheaper to do the mathematics, then to do such branching? granted modern hardware is more capable of dealing with such branching, i was simply under the impression that it's overall cheaper for the gpu to do the matrix math, then to do any branching on dynamic data sets.

You should be able to pack as many bones + weights as you want in your vertex. If you're on relatively modern hardware (DX10-class or higher) you may want to branch on the weight being > 0 before fetching the bone matrix as an optimization.

eh?, i was under the impression that overall it's cheaper to do the mathematics, then to do such branching? granted modern hardware is more capable of dealing with such branching, i was simply under the impression that it's overall cheaper for the gpu to do the math, then to do any branching on dynamic data sets.