Poser 5 had a similar problem to this, I don`t know the solution but maybe someone else can remember?

You know, I never got this problem in P5. One of my gallery posts, "The Naked Truth" used a similar setup with no trouble.

Just a guess, but it would seem to me like each mirrored plane in the scene is trying to reflect not only the object of the woman, but also each other's reflections of that object. What are your raytrace bounces set at? The problem here, as you might know, is that in real life if you reflect a mirror in another mirror, the reflections can go on indefinitely. Obviously, this is a problem for computer software, which is limited by the number of reflection bounces it can trace. Have you tried setting the bounces lower/higher to see if changing the reflection depth in either direction helps this?

maxx, the current bounces are set at 4. You could be right about increasing the amount but I don't think so because my gallery post managed to render ok in P5 with bounces set to 3. The other thing which makes me think that's not the problem is that if it were so, surely none of the secondary reflection would show. It's as if there's a planar "cut" somewhere in the reflection which stops anything further. I'll try it anyhow but I'm not hopeful.

I see what you're saying, Sam, and you're right. I couldn't find the P5 version of this in your gallery. Do you have a link so I can check it out?

Attached Link: http://www.renderosity.com/viewed.ez?galleryid=792850&Start=19&Artist=SamTherapy&ByArtist=Yes

Quick question: are those walls you're using there "single-sided" square props? I just tried duplicating this in P6, and the reflections seemed to come out correct. I'll post my sample in a few min. I just used the ground plane, and two single-sided square props (resized), with a chrome material on each. It hasn't finished rendering all the way yet, but so far, I don't see the same problems you have there.

Yup, I'm using the one sided square.

Surely the image of james right next to the real one shouldn`t be all black, like a shadow?

"Surely the image of james right next to the real one shouldn`t be all black, like a shadow?"

It's the reflection of his backside, and since there's only one light source illuminating his front, I'd say it's correct (there's no light coming from behind, and Poser does not bounce light rays off of surfaces like mirrors, such as would be the case with reflective caustics).

The only thing I'm not sure of is WHY it's even there in the first place. I'm not sure it should be. Message edited on: 07/13/2005 00:19

It could be a reflection of a reflection.

"It could be a reflection of a reflection."

That's what I said earlier, but it's incorrect if that's the case. I would expect it to also be reflecting on the other plane in the same way, where the two planes meet in the middle.

Message edited on: 07/13/2005 00:34

Some differences between tests: The original picture appears to have the exact, or almost exact, center vertically and horizontally to the corner join of the mirrors. Wondering if some kind of math anomaly isn't being generated that way. Subsequent tests don't have that same angle. I'd still be interested to the see the first one with a larger number of bounces...

"I'd still be interested to the see the first one with a larger number of bounces..." Me too. Remember, it's raytracing at work here... an accurate rendering method that plots a view of every pixel in a scene through a virtual camera's lens. Rays are cast from the camera, taking into account the location, strength, and quality of each object in the scene. Reflections should be accurate and realistic regardless of POV, however there definitely seems to be something strange going on in both the original and my first test render. At least to my eyes, and compared to the results in other raytracing engines.

I don't quite see what's wrong with #9 - looks alright to me. Am I overlooking something?

There's nothing wrong wth number 9. I can get Poser to do something similar to the ground plane by using auto render settings and setting it to the second notch that uses raytracing. Could you post a screengrab of your render settings? Also, it may be the properties of the squares, or their shader settings, so you might want to post screengrabs of those as well.

It looks like there's some clipping going on when things get too close to the mirrors. I can't remember. Is there a raytrace bias setting, like shadow bias? I've noticed odd clipping problems with raytraced reflections before depending on the smooth polygons setting. You might try turning that on and off for the scene and the figure and see if it makes any difference.

Something else - go back to the 1st picture and look at the reflections at floor level. The floor molding is reflecting back black - but isn't even a darker color on the apparrent mirror objects. I'm curious - are we being hoaxed here? :) (He said with good will and humor.)

First, my image seems to be caused by having too low of a Reflection_quality setting. I upped it from .1 to .5 and it's improved. Not perfect but getting there. I'm trying a render with a quality setting of 1 to see if it gets it right. Why the low quality setting in the first place? you ask. Well, there was a thread a few weeks ago about improving the look of reflections by dropping the quality and boosting the Ray_Bias. Seems that it only works in certain situations. As far as maxxx's experiments go, the reflections he's seeing are correct. Like my gallery pic (linked above) he is seeing reflections of reflections. Get a couple of mirrors and a Barbie doll and try the same scene for real. You'll find the Poser version is reasonably accurate. Remember also that mirrors (in real life, at least) reflect light, so the rear of an object would be illuminated somewhat.

Ray_Bias was the culprit. I dropped Reflection_Quality to 0.5 and Ray_Bias to 0.2. Raytrace Bounces are now set at 2 and everything renders just dandy. Thanks for your input, gang.

"I don't quite see what's wrong with #9 - looks alright to me. Am I overlooking something?" Well, looking at it again with fresh eyes this morning, it does seem correct to me now, based on the position of the character and the camera angle. My eyes were the only thing "off" last night. LOL. Glad you got it sorted, Sam.

Good sorting, SamTherapy!!

Droppping Ray_Bias back to its default value improved the image even more. Thanks again, Lads and Lasses.