I am not sure if it me or the function.
I am was using the RiPointsPolygon and it seems to work until complex mesh where used. Not matter the size of the mesh it would still get smashed into a 1x1 sphere of entangled chaos.
The Cone was actrally points up and fatter, while the monkey head well got turned into a sphere, then there is the hole in the floor....
I went ahead and switched to RiPolygon, and everytthing was fine. Where I am lost is if The Parameter data is fine, why would the face length and vertex index list make any object smashed into a ball of 1x1? I would have expected to see it with alteast the oringal points. I am nore sure if this is a problem with your function or me as I have written my file w/o the RiPointsPolygon and it seems to render fine, with the same parameters.
There's little I can say without seeing the code (and possibly also the input data).
So can you post a small code example or the relevant parts from your script that show how you call RiPointsPolygon() and what the input looks like? It might also be useful to see the generated RIB snippet...
- Matthias -
RiPointsPolygons(geo.nvert,geo.vertids, P=geo.P, N=geo.N) #, Cs=geo.Cs)
I get an error when trying to download the zip....?
But anyway, the interesting question is what geo.nvert, geo.vertids, geo.P and geo.N looks like. Well, I take it this is from the script you also posted at the Aqsis forum, isn't it? So I had a quick look at it and found a typo in geometry.getGEO():
for vert in obj.verts:
You accidentally appended the normal to P and the vertex to N (it should be the other way round) which explains the squashed look.
By the way, is it guaranteed that the individual faces are convex? If not, using RiPointsPolygons() is not valid and might lead to unexpected behaviour (i.e. artifacts), you should then use RiPointsGeneralPolygons() instead.
Out of all the stupid mistakes I do this had to be one of them lol, thank so much.
Sign up for the SourceForge newsletter:
You seem to have CSS turned off.
Please don't fill out this field.