Why is this aliasing? I thought aliasing is when you under-sample.
But in Catmull-Clark we are increasing the samples. Is it because the new samples don't approximate the surface correctly ?
keenan
@a4anna Good question. More broadly, aliasing happens anytime you have a signal processing strategy that causes two different signals to end up looking the same, i.e., it makes one signal look like another signal that it is not. In this case, it's our interpolation strategy rather than our sampling strategy that fails: by using Catmull-Clark subdivision to interpolate the positions and normals of a triangle mesh, our nice smooth signal ends up looking as though it actually came from a highly bumpy signal.
Why is this aliasing? I thought aliasing is when you under-sample. But in Catmull-Clark we are increasing the samples. Is it because the new samples don't approximate the surface correctly ?
@a4anna Good question. More broadly, aliasing happens anytime you have a signal processing strategy that causes two different signals to end up looking the same, i.e., it makes one signal look like another signal that it is not. In this case, it's our interpolation strategy rather than our sampling strategy that fails: by using Catmull-Clark subdivision to interpolate the positions and normals of a triangle mesh, our nice smooth signal ends up looking as though it actually came from a highly bumpy signal.