We are currently on a patch which requires a rather accurate depth sorting of the objects that are being rendered. We are running into the problem of artifacts when the perspective is moved(culling is not cool sometimes! :-)). We’ve been experimenting with depth buffers but i was wondering if wbuffers are available, somehow, in vvvv. Also if anyone has experience with this type of approach?
yep, had that on 3d particles most, you need to re-shuffle objects by the length they are to the camera… But not quite sure in the part you are talking about perspective, so can you post couple screenshots of how artifacts look like
Thanks for the fast answer. We have two main elements on our renderer:
a cluster of textured sphere and a nurb model. The model is placed at the center of the cluster. Our issue is that when we rotate the model we get those artifacts. We have tried to cull the faces of the model but the result is not satisfactory because of the type of texture we are using(we need to use both faces of the model).
you give points, it gives you length,
“!” means length ::]
so your problem that you need texture on both sides of the mesh?
quite odd problem, think i’ve seen thing that what you talking about if you use some transparent shader you see moire inside, first what i’d try is to make a copy of same object with inverted normals and cull maybe…
but looks like you need something like transmittance technique