Another general question IM vs. DMC and white interpretation

[Alexander Petrov]

Нi, i've read all Vray for Rhino manuall and Vray Complete guide several times, but i didn't find an answer of my question. So i noticed that there's a principle differencies between IM and DMC white interpretation. DMC is a bit darker than IM, and i didn't know what's the reason for that.

Here are two renders, of my last project. The first picture is rendered with DMC+LC, the second IM+LC the rest settings are equal!
As you can see, the first image is more darker and gray than second. The white is not real white, the rug is gray and nondescripted :-\
But the first picture have more acurate shadows, and details thanks to my favourite DMC ;D
The second picture have fantastic light beige rug, and the white is clearly white :
I've tryed everything - to adjust color mapping, brighten textures, F-stop, ISO.......everything possible, and nothing ??? Yes the global brightness of scene was increased but the rug and white walls stills grayish :-\ What is the reason for that, please help me :'(

[Micha_cg]

Re: Another general question IM vs. DMC and white interpretation

Hmm, interesting - did you have seen - at the underside of the balkony it looks like no light is catched from outside. If this is the case at the room ceiling (not visible at the images) too, than this would explain the big brightness difference. I have the feeling, so the question could be, why the DMC solution catch no light at the balkony underside?

[dalomar_cg]

Re: Another general question IM vs. DMC and white interpretation

This is really just a best guess, but I think the issue is (more or less) that the IR process is slightly more "additive" than a DMC process is. IOW, DMC is extremely local, and depending on how accurate you're solving, that could determine how much light actually gets retrieved by a given pixel. IR on the other hand is averaging/merging a number of different illumination points, and therefore may be "capturing" more light than DMC just due to the multitude of different points that are sampled. I'm not sure whether that's actually the case or not, but that's my best guess at the moment. FWIW, the DMC looks a bit better in my book...there's a bit more that's washed out in the LC version. Either way, those are nice images

[Alexander Petrov]

Re: Another general question IM vs. DMC and white interpretation

Thanks for reply guys Infact this is my global problem, not only in this case. In other cases i've compensate that with Photoshop - curves, levels, exposure. But now i want to determine it generally in Vray ;D
Otherwise, the scene is pretty simple - as you can see, there's some furnitures, TV set, generally i'm using Vray sun and sky system - adjusted manually, one big rectangular light outside to the window (invisible checked), nothing specially
Phisical camera - shutter300; f-stop 7; ISO 300
Color mapping - Reinhard 1/0.6 for IM+LC , or 2/0.8 for DMC+LC affect beckgr, clamp output, sub pixel - all checked ;D
Gamma corection is 2 not defaults 2.2 + secondary ray bias (default 0.001)
Image sampler - ADMC 4/100 noise 0.002, AA filter i did'nt use generally
Displacement - edge 0.6, subdivs 1280, ammount - controlled by texture multyplier, Bbox, dependent, tight - all checked

[Micha_cg]

Re: Another general question IM vs. DMC and white interpretation

Do you use "store with IM" for the rect light? I build a simple scene, a room like yours - "store with IM" caused an effect like your and without this option, both solutions are same looking. This option makes, that the light isn't used for the DMC.

If you like, you could post your scene (you could delete the object you don't like to share) and we could test it too. Or you send it me per email.

[Micha_cg]

Re: Another general question IM vs. DMC and white interpretation

IM vs. DMC ... same brightness of shadows and "direct" light (IM with "store with IM" at the rect light)

I used avery fine IM, max rate 1, because I liked to test too, how good is the speed of the IM at highest detail quality - not bad, the DMC is still noisy and the IM shadow is nice detailed. (that was not the question, but ... )

[Alexander Petrov]

Re: Another general question IM vs. DMC and white interpretation

Hi again ;D I've made some tests, look the results











Postprocess in Photoshop with Blur tool



Yes Micha, the scene is a bit brighter

[Alexander Petrov]

Re: Another general question IM vs. DMC and white interpretation

As you can see, the "store to IM" is not a decision :-\ It seems that DMC presents a super acurate detail, infact the gray color is obtain to fibres of the rug. DMC repeats very accurate Displacement texture, especially hers dark zones. The end result is grayish vision. On the other hand, IM blurs that dark zones on the rug fibers, and they looks brighter and soft - like a wool. So my tests shows that, and i'm very disappointed :'( It seems that for interior renderings IM+LC is the right decision :

Micha check your mail

[zuper]

Re: Another general question IM vs. DMC and white interpretation

Great explication!!!

[Micha_cg]

Re: Another general question IM vs. DMC and white interpretation

Alexander, I got the scene. Some small observations first:

* extrem small bucket size could cause longer render times
* glass isn't needed at the windows
* I suppose so you know it - the displacemenr rug is a render time killer
* the LC sample size could be a smaller like 0.005, so finer details could be better shown

But the different light distributation of the IM and the DMC is a problem here too (if I use your setup without and with modifications). I don't understand why my first test of the simple scene was without any difference. OK, next test, I load my setup file from my first test to your scene and .... now DMC and IM match, the same light distributation!

I have no more time today to find the difference between your and my setup. But it's good to know, that a solution should be possible.

Here my setup file (it should work with any HDRI or disable HDRI and use a white env color with intensity 2.5 + hide window glas and window light):
http://www.simulacrum.de/download/Se...st_test.visopt

... a last idea: you used clamp output and subpixel and I don't used it ... no, this is not the reason.

Here the test images with my setup file, both show the same burnout at the window, no difference, anything like expected. Something at your setup file cause the problem, but what? ???

[Micha_cg]

Re: Another general question IM vs. DMC and white interpretation

... the image sampler cause the problem. If I change your DMC 4/100 to classical 1/6, than both calculation method show the same result. I would say, let us ask Vlado.

[Alexander Petrov]

Re: Another general question IM vs. DMC and white interpretation

... the image sampler cause the problem. If I change your DMC 4/100 to classical 1/6, than both calculation method show the same result. I would say, let us ask Vlado.

Michaaaaaa.....I love you ;D You are my angel Yes it proves that the Image Sampler is responsible for the problem : Similarly i've burned with shame :-[ Im so stupid : i should be guess for the reason........its sooo simple :-\ Anyway, i would just say you one big THANKS BUDDY

DMC+LC/ Adaptive DMC min3/max4 noise 0.002 AA Area 1.5