Is there anything similar to the validation filter in subtance designer/painter in Vray?

V-Ray V-Ray for 3ds Max General
21

I wouldn’t say it is pointless ..At least according to the following link having too high albedo will affect rendertime /noise amount :
https://coronarenderer.freshdesk.com…hat-is-albedo-

I remember reading a blog post on Viscorbel site comparing how GI looks more natural when using a correct values for diffuse but it seems that Austris deleted it for some reasons ( you can find a copy using any archive website under this url http://viscorbel.com/vray-materials-part-1-diffuse/ in 2015 )

for me following PBR rules have 2 main advantages over if-it-looks-cool- then-it-is-right the first one is having constant workflow so all my shaders will work together no matter when and where I created them.
and second my materials will looks good under any light conditions as it wasn’t built for a specific scene lighting but quite the opposite ..I build my lighting based on these materials
I’m not that good in math and physics and I usually eye-ball it , but for me having “right” color range made my life easier.
..at the end it might be just an illusion that makes me feel comfortable and confident so I do my work well :wink: :stuck_out_tongue:

22

No one is saying getting the Albedo just right (ie. generally lower than expected. Cfr. one of my threads from 2005 or so.) is pointless, on the contrary!
What we are saying is that substituting exact data (f.e. vrscans and similar approaches)for an eyeballed arbitrary value isn’t going to solve a thing.
Further to this, there is *no way* for a renderer to know what the material is supposed to be (paint, bricks, grass, wood…), and so there can be no guideline given on its Albedo.

If marking in false colors values above and below two arbitrary thresholds makes anyone feel better, as i said, i can easily script that, and even give you the choice of false color to use.
But a guide to shaders’ albedo levels it won’t be.
Deciding that is going to be a task for a human mind for a while still.

23

That’s fair and understandable . personally I wouldn’t go that far .

Yes, this will make me very happy : ))

24

It was a figure of speech, which completed in the next two lines: it’s not a tool as such.
EDIT:to further elaborate, it wouldn’t currently work behind refractions, and would force the addition of a number of REs to the standard render, while being an offline operation to be performed after the render finished.
I’d rather *not* spread bad ideas: the LUT explorer alone will haunt me to the end of my days. :slight_smile:

25

Perhaps new guidance should be for everyone to set their black point to RGB 127, their white point to 129 and mid grey stays at 128? :smile:

26

All of this could be avoided with some presets, or some official tutorials. Example, we are chaos group, look, we’ve made this material, we made it this way because of x reason and you do this to it too because of y! Just common materials, nothing fancy. All of the science and technical detail explained. I appreciate this could be a vast wide ranging endeavour but it’s absolutely fundamental.
​​​​
Just look at the amount of information out there on official websites for Zbrush and SD. The Vray courseware pales in comparison at almost every level and in fact it does not even offer a material orientated tutorial. You can understand lighting correctly and how to use a camera but all of that is futile if your materials are not backed up by some official baseline.

I’d consider myself an intermediate to advanced level user, around 8 years professional experience, yet I am still not confident in creating simple materials correctly. My materials change sometimes drastically from scene to scene, and I spend so much time tweaking them to fit. This leads me to to believe I am not not creating them correctly, so I’m I’m sure you can understand when I came across something like a validation filter, it got me excited.

​​​​​
​​​​​

27

When I used the validating material in SP, it’s pretty handy, especially in that real-time context. You pop it in at the top of layer stacks and it gives you a vibe of if any albedo and reflection value is out of defined range based on their PBR guide. If I know the HDR or lighting used is somewhat calibrated, then I can use that validating material to just quickly check if I stupidly set some out-of-whack value in my shaders.

I don’t really see the connection between that and material presets as I am NOT looking for material presets or calibrated material when using that. Even when it shows red, it only tells me that it’s “out-of-range”. I wouldn’t deem it as I absolutely cannot use that value when it shows red. For users who are new to authoring materials or looking for quickly dialing in reasonable values as a starting point, assuming he is working in calibrated lighting, the validating material isn’t a bad idea.

Again, speaking in real-time context in Substance Painter.

28

Just so i don’t abuse words, i went in and scripted the tool i was mentioning earlier.
It creates an “Albedo” out of diffuse and reflection filter, and then colorises based on two threshold and colors i chose: yellow for values below 10% (0.1f, ~25i) and red for those above 90% (0.9f, ~230i) As Jason was mentioning, a tool like that may well be useful in specific contexts, and even then just as a *very* generic guideline.
Notice the results below, and how little descriptive of this scene’s issues with shaders’ albedo they are.
Below, from left to right: RGB, Albedo, Colorised Albedo, Colorised RGB

rgb.jpg
rgb.jpg867×577 228 KB
albedo.jpg
albedo.jpg867×577 168 KB
colAlbedo.jpg
colAlbedo.jpg867×577 175 KB
colRgb.jpg
colRgb.jpg867×577 232 KB

EDIT: i obviously didn’t alpha premult, nor excluded self-illuminating parts. for speed’s sake only.

29

I fully understand after reading through this thread, that it may have to be generic.

SRGB TO LINEAR VRAY, who knew eh? Not me, how would I know such a thing before hand? I’m not a scientist, I’m an aritst! Right? I’m glad I learned something though. Does knowing this make my images better? Not reallly.

What the hell are those guys in Corona doing with it all then? Are they just wrong? Maybe you should just tell them eh? The guys at allegorithmic too, pushing this metallic PBR workflow (which coincedntally is clearly the way its all going, its even now fully supported in next) that they are wrong in their methodology too?! Don’t film studios work in linear? They must do right? Studios the world over work with allegorithmic dont they? What baseline/validation do they follow? Can you see why I’m confused? Try and explain it to me.

Honestly? I’m just glad there’s some conversation about it all. Theres a vast space of offical nothingness on this subject from the gods that be. You guy’s need to start catering more information to the grass root user, i.e. actually telling them what they are doing and why. It’s honeslty embasrrasing that after 8 years of professional usage I cannot 100% say to myself I know how to create a material correctly. There are no official guidlines, nothing.

30

…

31

Well, i am as confused on PBR as the next guy.

If i was to attempt a historical perspective on the subject, i’d say PBR is new to the realtime world (read: videogames), which until very recently had no LWF (or even a standard for gamma, for that matter.), nor any valid, and quick, BRDFs to show for.
Facing those issues for the first time, realtime adopted a few of the proven techs naming them as they saw fit.

Ask yourself this, in other words: if the only “Physically Based Rendering” was Substance’s, what had everyone else been up to until then?
What were we rendering with V-Ray (or mR, or Brazil), before the moniker came out, with physical sky and sun, and camera, and…?
Corona came to the party *much* later than us, with zero baggage to keep supporting: it jumped on the ready-made set of features and adopted them without much fuss.

Adoption of a *workflow* is a more delicate affair: often features are given out to not hinder people’s ways of working, but not as game changers (for the host application).
In other words, we would be “PBR” even with glossiness instead of roughness (really, they are mathematical inverse.), and could make -identical- metals before the metalness parameter was added.
Having those in the UI, however, will allow people coming from PBR-based applications to more easily translate into our shaders (i am not convinced, code in hand, this was ever the case with us, but there you go.).

As to who’s right, and who’s wrong, well, that’s a subtler question still, but it so is in life, generally. :smile:

32

I am wondering why people are turning crazy because of all this “PBR”-nonsense. Engines like V-Ray, Brazil, heck even Mental Ray were physically based more then a decade ago. It is new to all the game-related stuff, yes. But nothing to get excited about if you were doing raytracing before.

33

It improved the realtime world extremely.
I don’t understand why offline rendering guys always be so arrogant about this.
Nobody ever said offline rendering weren’t physically based before.
Of course I do understand the term is misleading for noobs but why is everybody so offended about this? Why are we talking about PBR here at all?
This has nothing to do with the PBR approach in particular.
Albedo was always important.
It is an actual part of the real world and one of the most important ingredients for the simulation of light in any render engine that’s even close to be so called physically based.
So please stop misusing this thread for PBR bashing and start an extra thread about how long you guys where rendering physicallybased before PBR emerged in the off topic section were it belongs..

34

That’s what I keep asking myself all the time.
Take the maps you created in Substance Designer or Painter and stick them into a VRayMtl and you’re good to go. It will look the same. As Lele stated, even before the roughness and metalness craze started. Yes, I do use Designer and Painter. But I don’t NEED to use this stuff to create a good render. It’s just another tool in the palette. It makes creating materials easier, but not necessarily better.
*<--- Looks at his join date and now feels old*

35

They found scene referred linearity and sRGB fifteen years later, along with BRDFs other than Blinn.
Called it Potato.

I don’t understand why offline rendering guys always be so arrogant about this.
How’s stating what is what being arrogant?
Should we gaily adopt anyone else’s lingo, each time someone discovers hot water and calls it “the warm stuff TM”?

36

You are right of course. But why to get up about it in a discussion about thermometers? That’ll lead us nowere.

To get back to topic.
If I understand you right, you are saying: Look, this scene has big albedo problems which won’t get identified by the threshold colors.

But one could now take the pixel info to identify the wrong albedos by measuring them. Which right now, isn’t easy.
We could put the material in a scene with a domelight at 1 and a sphere for the material. That’ll take ages for a scene with 50 materials.
Or use compositing like you did, which requires us to save the image and put it in an external program. Also, we couldn’t get the right albedo of none VrayMtls like ALsurface fast sss and so on. At least I don’t know how.
The threshold colors would just be a bonus, not necessary for the element to be useful, but handy. The workflow to use them would be:
I want to create an asphalt material, so I go to Albedo - Wikipedia f.e and type 0.04-0.12 in the threshold values and start to mess around with a bunch of textures and can always have a direct feedback whether I’m in the ballpark or not.
The thresholds could be 0-1 by default to not make everyone think a value over 220 is never correct.

Another solution for the problem would be a material preview window which offers a mode for an albedo test scene as discribed above.

37

“One” is someone i never quite got to know and meet, eheh.
In maya, there’s debug shading with IPR (soon to come to Max as well), which ought to simplify this kind of investigation.
In that type of context, providing for Albedo ought to be easier, and once LPEs are in, it may even be possible to get something out of refracted bits as well (which currently won’t work.).

I’m not against helper solutions, i’d just like them to work right, not intermittently, and not not at all. :stuck_out_tongue:

We could put the material in a scene with a domelight at 1 and a sphere for the material. That’ll take ages for a scene with 50 materials.
This would be the most exact method, and it could be scripted too (values above 1 for any surface-only shader would be wrong, so it could be automated.).
Notice that to break reciprocity (or energy conservation, or to have shaders above 1.0) takes some effort in V-Ray, so the benefit of such an approach is marginal, and likely going to benefit only a small(ish!) subset of shading tasks.
It would still not be able to say *anything at all* about the validity of the Albedo interval chosen for that type of shader.

Perhaps adding simple material presets (much like for the fastSSS2) is the best way forward.

38

Right. I’ve some time to test this out.
As I’ve stated, I’m at a mid to advanced level, but I will attempt to go through this in a way that a user at any level can follow. An approach that seems limited/lackadaisical from Chaos Group generally.

So, I’ve set up a scene with a sphere and vray domelight set to 1, as well as a 3ds Max Physical camera. Camera settings do not matter in this instance.

I’ve applied an asphalt texture (I wouldn’t call it fresh but close enough I would think) to the sphere. I’ve then added a VrayRawDiffuse element in the 3dsMax render elements tab in render settings.

I press render. Default settings.

Once rendered, my understanding is that I now need to obtain the float value from the ‘Pixel information dialog’ box and cross reference this with an albedo measurement reference, such as the one seen here on the right hand side (Fresh Asphalt= 0.04). So, ensuring I have flipped from RGB Colour to VrayRawDiffuse element in the frame buffer I now need to sample a few areas of my rendered sphere by right clicking different areas of it.

Colour Float around 0.04 - 0.06 in my example.

Now. Questions.
1. I wouldn’t want it exceeding 0.04 too much right? The value would obviously vary from area to area dependent on the variation of tones in the texture? But, it would be better to have those values vary (on average) below 0.04 rather than above it?
2. What would be the quickest, most streamlined way of tinkering with this on the fly? I’d personally go and lower the RGB Output value in the 3dsMax texture output properties, and tone this value down until the float value level in the Pixel information dialog is around the values I want. 0.04.
3. In terms of reflection/glosiness/bump maps. How can I ensure the values of these are effective? What I mean is, how can i ensure that the lows/mids/highs are correct in range? Is this just via eye? This is why I want to nail down some albedo values as i cannot sometimes tell whether it’s my diffuse/albedo textures that aren’t correct, o the additional material maps for reflectivity, bump etc etc.

01.jpg
01.jpg1104×453 63 KB

02.jpg
02.jpg970×515 53.8 KB

39

From what I gather, one side is mostly interested in having the ability to sanity check albedo values, and the other side is mostly opposed to setting specific thresholds that indicates when a total albedo value is “wrong”. Is that fair?

Would an albedo render pass that doesn’t get huffy when it thinks values are wrong, but just impartially presents albedo as well as it can be acceptable compromise? That way you can inspect the pixel values and rely on your own judgement as well as references like the one Vlado posted.

40

I’d call this a fair assessment, yes.
The gap is between artistry and engineering: the terms “right” and “wrong” clearly hold different meaning to either party.

Would an albedo render pass that doesn’t get huffy when it thinks values are wrong, but just impartially presents albedo as well as it can be acceptable compromise? That way you can inspect the pixel values and rely on your own judgement as well as references like the one Vlado posted.
It wouldn’t work when anything was behind refraction, currently.
Without that limit lifted, an Albedo is just a diffuse filter plus reflection filter.
If the user’s that advanced which they want to be picky on specific Albedos, i’d assume they can also plus the two REs, but perhaps that’s optimistic/old school?