MR strange artifacts


#21

Photons or FinalGather trace depth?


#22

try both then if it clears up, increase them back one at a time to find the culprit.


#23

I will do tomorow and keep you updated. Thanks :thumbsup:


#24

Maxing settings will not solve problems. These type of issues often come from wrong settings (it looks like oversized, overbright photons there… probably a photon shader is altering energy levels dramatically). All I can suggest you at this point is try to create a GI solution for a stripped version of your scene. You should be able to isolate and identify the source of the problem (I could think right now that your glass materials are part of the problem). Cheers!


#25

You’re going about this the hard way. Instead of fixing the colored light thingies, just change the concept of the picture. Just say that the little light things are alien energy beings that were trapped in beer bottles for 100,000 years and they were finally released so they can wreak havvoc on the innocent hottie bar chicks.


#26

LOL, indeed. Maybe even add more colored light patches to give the room a “disco” feel.


#27

BTW - I mentioned reducing the trace depth settings because of a similar post on the discreet website. A guy posted a similar render with green & red circles and he solved the problem by reducing the trace depth.

With that being said, I don’t think thats the actual solution…moreso a patch to correct something else wrong in the scene. (either with a material or photons)

Just wanted to explain the reasons behind my post.

Jeff


#28

BTW - I mentioned reducing the trace depth settings because of a similar post on the discreet website. A guy posted a similar render with green & red circles and he solved the problem by reducing the trace depth.
That makes sense Jeff. Reducing depth might force those pesky photons to be lost along their path before being stored.

With that being said, I don’t think thats the actual solution…moreso a patch to correct something else wrong in the scene. (either with a material or photons)

Just wanted to explain the reasons behind my post.

Jeff
I’m all with you there… I would put my money on the material… but I’m known to be a bad gambler, so don’t take my word for it. :slight_smile:


#29

I think i know one more way to try…

Dont you have standart max’s materials with high value of specular?
If your answer is yes, then try to decrease all speculars of all the materials to 0.
There is max’s bug with speculars, but i didnt know that MR has also this problem…
Try… :slight_smile:

And the other thing - i would try to render the scene with the same lightning, but with default gray shaders on all the objects so you can say if it is lightning or shading…

Good luck…


#30

hi there
I did some tests again. Well…I have played around with the trace depth, like Jeff said and it changed the “funky dots” to less but larger dots!?
I’ve changed the architectural glass to plastic. So…no changes.

I’ve also sent Waples change of concept to our publisher which also won’t work. I like the Idea of those trashy aliens kept in a bottle.
BTW: What was the name of the movie with this red “evil” balloon conquering a spaceship.
:shrug:
Whatever…I’m afraid this thread is getting to complicated for me and I’m even more afraid to look like a idiot when you pro’s take a look at it.
I’m awaiting an answer that says something like:
“Man! Just change the Scanline Renderer to MentalRay!”

ok. doggjaww was so nice to spent me some webspace:
http://www.lordzco.com/florian/EducationellBAR.rar
Full Scene with textures, without girls. First time girls work perfectly :smiley:

It’s a scene for a small isometric game we’re doing. OH MY GOD!!! Can the isometric camera setting be the problem? I’ve never seen isometrie with MR.


#31

Florian, if you change the bottle glass materials that you use from the architectural material, to a raytrace, the problem goes away. I’m not sure if MR completely “understands” the architectural materials.

But it only took a few minutes to change all the glass materials over to a raytrace (plus it seems to renders a bit quicker if you use raytrace instead of the architectural mats)


#32

Ive come across this a few times too, look under samples per pixel. for testing I used 1/4 Min. 1 Max. I would get funky colors like this if the samples were low.

Arxangel


#33

Jeff… I’ve used architectural materials with mr a lot, with no problems whatsoever, using GI and caustic photons. However, if you have any scenes where architectural materials give you problems, I’ll gladly take a look at (a stripped down version of) them. AFAIK, there shouldn’t be a problem with arch materials (although I usually assign my own photon shaders to tweak the indirect lighting solution a bit).
Raytrace should be faster than arch. Arch materials use a more complex shader that the raytrace material, so the increased performance seems logical. I always use the simplest shader possible to do whatever I’m trying to do (some native mr surface shaders are quite fast).
Anyway, that’s my 2 cts…


#34

I personally haven’t experienced any problems with them either (but then again, I rarely use them). However, I have never used them in a multi-subobject material as Florian needed to do in this scene. I’m not sure if it’s the combination of the arch.mat being used in a multi-sub map that may have caused this problem or what. But MR definitely preferred a raytrace mat in the multi-subobject material in this case.

JP


#35

I’ve also done lot’s of arch experiments and never seen something like this. I must admit for this rendering it’s really not that important if it archs, raytraced or even standard with 50% opacity. I just don’t understand it.
Someone could serious get injured if he would try to use them in a commercial project that has to be finished in time. (Ask me)
Most arch mats have better quality than “standard” mats. I really would like to trust them.

EDIT: ok…if you are able do build your own, based on the MRshader, you can off course build shader with better quality. I think the arch mats are something like presets. But “inside” it’s MR.


#36

I know what your saying florian. By changing the material over to raytrace, we haven’t really “solved” what was causing the problem in the first place. It’s moreso just another work around.

Maybe someone with a better understanding of what’s going on “under the hood” of MR will be able to explain why the spots appeared in the first place. I do know they went away when I disabled final gather. But I didn’t spend any time increasing the FG samples, etc, since I think you’ve already been down that road.


#37

Yep. My final result is a workaround of everything
:smiley:
…and for all those who take a look at my scene, I know it’s a mess of unneeded mats.
Anyway thats no reason for Max coming up with funky dots. I never subscibed for some Beta testing.
Perhaps there is a problem with the skylight. Because it’s the only lightsource that won’t be calculated without FGather. I remember the dispersion shader (SSS) is also working with every lightsource, except the skylight. Maybe it could have worked with it…sometimes.
Who knows?

SCREAM: WHO KNOWS? :scream:


#38

I will run some tests using arch materials inside multi/sub-obj materials and see if I can repro this.
As for this going away with FG, is there any way Florian or Jeff could send me a VERY stripped down version of the file displaying this? (Maybe just one bottle and the counter). I would take a look into it and let you know if I find anything.


#39

I’m sorry beeing the one who started this thread but don’t have the time to render some experiments. I already spent far too much time to fix that artefacts and the deadline of the project is coming to get me:deal:
If Jeff can’t do, I’ll promise to come up with a basic, but awfully rendered scene at least the day after tomorrow.

once again…thanks for your help. I love this forum :thumbsup:


#40

I’ll gut the scene down to a bottle and table only and see if I can maintain the problem and then post that (provided thats ok with florian?)

JP