UVW Render element wrong output?

[juang3d]

Hi there.

I'm trying to use the UVW render element in Fusion to apply a texture to an object, however the texture output is very different when I apply that texture in max and when I apply it in Fusion, check the pictures test_UV_Fusion.jpg, the left one is the max render output, the right one is the FUsion output.

I'm not sure if there can be some problem with the UVW output, Fusion just gets the U and the V, not the W, check all the attached pictures and please, tell me if there is a problem here.

I also tried to enable the built-in UV option of the EXR format, but it returns nothing :P

Some ideas?

Cheers!

[juang3d]

If I render the UV channels from the EXR using Scanline, I get the correct output, check the picture, this is composed in Fusion.

Is there something broken with the UVW Render element?

Cheers.

[juang3d]

BTW this is using 1.00.02, I have not been able to install the latest DB.

Cheers.

[juang3d]

Hey guys, is this something already knwon and that's why i'm not receiving answers?

I would like to post this on mantis, but I would like to receive a confirmation first, just in case I'm doing something wrong in max or in Fusion to interpret the UV channels.

Cheers!

[racoonart]

If the uvw channels from scanline and Corona differ then it's a bug. Uv channels are absolute, so they should be the same in all renderers. A case for mantis then ;)

[juang3d]

Cool, yes, they are completely different, in the first post you can see the UV from Corona, and here is the UV from Scanline.

Ok, I'll post it right away.

Cheers!

[juang3d]

Ive discovered a thing, I reported this to mantis, but I'll write my discoverings here too, in mantis the problem is the issue 0001050.

It seems there is a problem with gamma removal from the UV's in the render element of corona, that and a mixed value, being R = V and G= U, instead of R=U and G=V, wich is the standard scanline output.

So if I mix the GR (UV) + Gamma 2.22 I get more or less correct results, but since I have comp different thigns there are small errors like borders, so solving this in comp is not a correct solution, since it's data, so an error can create a big problem, check attached files in the zip, hte maxes are 2016 and the exr's are multichannel, also you can see the results in the JPG's

Check the attached pictures, instead of the zip, if you want it you can find the zip in Mantis.

Cheers.

[Ondra]

could this be caused by UVW channel having negative values? If it has neg values, they would be mirrored just as world position pass currently is

[juang3d]

Nope, no negative UV; in fact my UV's area complete planar mapping envolving all the object, they are simple UV's because it's just to put a logo on the object.

In Mantis you have a zip file wich contains a max file from the example explained there (the one with the word corona written on it). In that case they are just primitives with the default mapping.

Cheers!

[juang3d]

BTW Ondra, bear in mind that if there were negative UV values, then the scanline UV's should be wrong also, or at least show some kind of strange thing, right?

I think it's just a problem of gamma interacting with the UVW render element and the thing that you put U in G and V in R, instead of U in R and V in G like the scanline does.

Cheers.