[mverta]

Look at the area inside the mouth...

This is a standard plastic shader with uv roughness of .2

1 four-sided plane emitter, D65 preset.





These started showing up around sampling level 8... this image is sampling level 18.

_Mike

[mverta]

Any thoughts, guys?

[tom]

This problem may be about geometry problems. Since it's a scanned dragon, I assume there may be some surface imperfections like open edges, faces, etc.
Please, submit your scene for debugging as described here:
http://www.maxwellrender.com/forum/viewtopic.php?t=4126

[oscarMaxwell]

Hi Mike, what plugin did you use?

[j_petrucci]

I think it's Maya plugin... it's written on the window caption!

[oscarMaxwell]

true

[mverta]

I actually had a crash at the end of that render, so I have to reset the scene up, but I know exactly what settings I used for material and camera. I'll do that, make sure the problem reappears, and submit it.

And yes, it's for Maya!

_Mike

[Mihai]

is the dragon subdivided?

[mverta]

Not sure how you mean, Mihai... it's polygons - a mesh derived from a point cloud scan, I'm sure.

_Mike

P.S. File Sent

[mverta]

Another angle of the same scene... more small artifacts... visible on the standard plane the dragon's sitting on.




_Mike

[mverta]

Another test... dots in the same general area around the mouth.

Also... I notice that there are a lot of repeated sample levels. There were 5 repetitions of sample level 15, each at about 10 minute intervals. Is this normal?

[Mihai]

looks like a geometry error because the dots appear in the same place on the mesh even from a different angle.

the repetition is normal, it's maxwell telling you it updated the framebuffer (does it every 10 minutes).

[mverta]

looks like a geometry error because the dots appear in the same place on the mesh even from a different angle.

...except on the second render, where the dots appear in the floor as well. The floor is just a simple plane.

_Mike

[Mihai]

maybe because it is a reflection of a black dot If there is something wrong with the geometry of the dragon, like some normals are flipped, then those will render black, ofcourse that black spot will also be visible in reflections.

[mverta]

I considered that... if hyper-sensitivity to normal direction is a crucial part of Maxwell, then so be it, but it should be noted somewhere in bold type, I'd say!

I'm curious as to what the NL team has to say about these anomalies!

_Mike