see trekkie's thread on the problem at the corner of the staricase.
this may be a different bug.  is it still there with box filter = 1?

Quote -  see trekkie's thread on the problem at the corner of the staricase.
this may be a different bug.  is it still there with box filtering?

That render took 12 hrs. I will turn down the exp to say about 1.8 and I will try with box while at work tomorrow and see what happens. The floor is highly reflective. Not sure if the exp being at 2.2 washed it out. The floor reflected better in PoserPro. Even being washed out the back of the hair looks much better then in PoserPro. The camera is very close and set at 25mm

turn down RT IC and IDL IC to 50, incr. shade rate to 0.5 to speed it up a bit.

For test renders on IDL, go even lower on IC.

When IDL is enabled, RT IC is irrelevant, BTW. There is only one IC in use at a time. If IDL is off, the RT IC matters. If IDL is on, the IDL IC matters.

When you use the basic Poser render settings dialog, not the D3D one, the IC value is automatically sent to the correct one under the covers, and some math is done on it.

Also for test render, you may want to set RT bounce to 0, because you don't really need to see reflections to see how the IDL is doing.

OK but, lets say I wanted to do a final render....even if the skin sucks, I would like to get rid of the black spots.

Is it possible that the bias on either lights or the reflection node is the problem? (ok, maybe not if they don't appear in the without IDL render)

I haven't seen spots like that since Beta testing in July. The difference here is you're looking at them through a mirror. I never did that.

When I was seeing them in the past, they often would change by moving the camera a tiny bit.

This is one of those problems where if you are able to send the exact scene to SM, they find the reason really fast and fix it in the renderer. Every time I reported a generic rendering artifact without an attached scene, it was hard for them to reproduce. With a supplied scene including specific camera position, it is easy to fix.

Of course, sending copyrighted content to them is a bit of a problem, even if they already have that content.

Can you remove the woman and still get them? If so, save that!

Let me render one more time with revised settings. If I still have spots, I will contact support. John at SM has got to think I am a pest :)

As an FYI, render still in progress...but lowering the RT cache to zero and the IDL cache to 50, increasing the pixel samples from 3 to 5, changing to box, and lowering the exp from 2.2 to 1.8 and I am seeing the issue where the two walls meet in the upper right hand corner as reported by others. Whereas before I didn't see those. On the surface it appears one of those changes invokes the dirty corners where 2 walls meet.

I did move the camera a couple of degrees also to remove the open area in the very upper right, The walls of the changing room are shorter then the walls that surround the entire scene and it looked odd to me.

It's not really after some tests !

A:

I don't understand what you said.

"It's not really" What do you mean by "it" here? And it's not really ... what? It's not really ugly? It's not really fast? It's not really doing anything? What do you mean?

And the IC=x, are you talking about the standard Render dialog IC or the D3D Render Firefly dialog, which has two IC values. Which?

Anthanasius, even if the skin and the lighting need work...what I am after is the spots with high end render settings.

1.) how did they happen-get there (can I trace down their cause).
2.) can I repeat them in a different scene

If I can repeat the spots, then I will report a bug to SM tech support.

If it is something unique to those walls, their textures, the procedural floor texture, the reflections, the camera angle or the lights, that can't be repeated, then I dismiss it as an odd circumstance and move on.

Bumping (rotating whatever the smallest amount it moves) the camera higher, cut the spots down to 2.