Using Albeldo?

[Christa Noel]

Dear all,
i need to create a furniture with two color (green & white). i got note that these color is a main theme of the collection
, so i have to seriously pay attention when setup the color on my scene.
Reference color from client, http://rgb.to/pantone/3272-c & http://rgb.to/pantone/uncoated/7527-u
How to input those value to diffuse color selector regarding corona performance debugging
, which explain that maximum value for white is 180,180,180..?
i know that whiteness value of the image i produced will be much affected by lighting setup & it wont look as same as the client has referenced.
but it is very important to keep the ratio value of white to green.


i hope you guys all understand to every detail of my explanation (cuz, my english is very bad).
thanx!

[Juraj]

, which explain that maximum value for white is 180,180,180..?

I don't think yet that this should be set as some rule in stone. Anyway, one thing to keep in mind that this above is linear value (because 3dsMax color picker is) and other color pickers (like the Pantone website) are in gamma space 2.2(mostly sRGB for simplicity),
so first you have to convert that.
I am not sure if Corona is able to read "VrayColorMap" which can specify gamma space so you can input both linear and sRGB values and also convert them. It also has "VrayICC" node which can specify color profile in case you input texture to so you have consistent environment and no color shift along way
( because 3dsMax is unamanaged app)

Anyway, I do a lot of these jobs for giant furniture company and it's pain to manage. I still do it in Vray using VrayICC (for textures swatches), and clamp the value to about 0.5-0.7 in output. I am not sure how far is it from correct or incorrect, it's really just eye-balled. The color-swatches are basically to me how the color looks perceptually,
but don't represent their physical reflectance/albedo.

I would love to know some precise way too..


PS: Which brings me to..Once the "CoronaUltraSomethingMap" or how they will call their native bitmap reader, it should have Color profiles options alongside gamma. That would make it superior to Vray as it would merge VrayHDRi+VrayICC.

[juang3d]

No direct relation with the problem asked by noel20 but... 180 is the maximum recommended white value? I always though it was 242, because 180 is more grey than white...

What is correct and why?

Cheers!

[maru]

Here:
https://forum.corona-renderer.com/index.php/topic,2359.msg17366.html#msg17366

White acrylic paint, 0.8
white paper sheet - 0.6-0.7

255*0,8=204

I think it doesn't really matter that much if you don't want to be super-physically-correct. If it turns out too dark or too bright you can always change exposure. Just don't use values too close to 255 because it will significantly increase render time.

[Juraj]

Here:
https://forum.corona-renderer.com/index.php/topic,2359.msg17366.html#msg17366

White acrylic paint, 0.8
white paper sheet - 0.6-0.7

255*0,8=204

These values account for overall albedo including specular reflectance, which can be minor but it still can't be used directly like this.

[borisquezadaa]

How much of that albedo is the specular reflectance. One could ask...
Now really... how much?.

[Stan_But]

Energy conservation automatically reduces diffuse reflection now - so we can be calm with using albedo as diffuse level.
About this: http://rgb.to/pantone/uncoated/7527-u - 220/215/210 . White office paper is a white color in CMYK system. So it would be logical to take 220/215/210  from 0.7*255 ( 0.7 is the albedo of White office paper or 70% from incident light )

[Juraj]

How much of that albedo is the specular reflectance. One could ask...
Now really... how much?.

Depending on how specular is that material, so super individual. Could be as little as 0.04 or more from there.

[Stan_But]

How much of that albedo is the specular reflectance. One could ask...
Now really... how much?.

Depending on how specular is that material, so super individual. Could be as little as 0.04 or more from there.

yep
In web some sources pointed to values near of IOR 1.5 (or reflectivity 0.04) for most of dielectrics

[Christa Noel]

Many thanx before for all of your responses guys, i really appreciate it.

but... 180 is the maximum recommended white value? I always though it was 242, because 180 is more grey than white...

What is correct and why?
Cheers!

juang3D, here is your answer: about albedo https://corona-renderer.com/wiki/performance_debugging, http://en.wikipedia.org/wiki/Albedo

, which explain that maximum value for white is 180,180,180..?

Anyway, one thing to keep in mind that this above is linear value (because 3dsMax color picker is) and other color pickers (like the Pantone website) are in gamma space 2.2(mostly sRGB for simplicity),
so first you have to convert that.
I am not sure if Corona is able to read "VrayColorMap" which can specify gamma space so you can input both linear and sRGB values and also convert them.

yesterday i tried VrayColor.. and Yes Juraj, Corona is able to read VrayColorMap, thats help me a lot (i didnt know it before).
i have multiply/scale the color value to 0.8.. it is what i have done for my projects following the advise from headoff, thankyou verymuch headoff.

I think it doesn't really matter that much if you don't want to be super-physically-correct. If it turns out too dark or too bright you can always change exposure. Just don't use values too close to 255 because it will significantly increase render time.

for this case, i think it is really matter. i need to keep both color value ratio. but anyway .. case is fixed & closed. thanx.

[ignicapillus]

This topic has been sleeping for a long time, but I'd add my two cents with a test and and a question.

TEST: VrayColorMap works fine (see test; sRGB values from wwwrgb.to); it would be nice to have a 3dsMax or Corona map to do the job, so to avoid to use another piece of software. My previous workflow was to apply gamma 0.4545 (1/2.2) to the sRGB reference in Photoshop to obtain linearized RGB values; boring process.

QUESTION: Nevertheless I read that Corona does not like high values for whites, and people are using values like [220, 220, 220] or even [180, 180, 180]. Now, RAL 9003 (a standard white color used for household appliances) is [236, 236, 231]. Following headoff approach, I can make it darker, i.e. reducing values to 75%; I get [177, 177, 173], which is similar to gray RAL 7003... so, to preserve mutual appearence, I should reduce *all* colors to 75%, and then compensate the darkening with exposure. Do you think this process is correct?

[ignicapillus]

the second test involves energy preservation; color is exactly the same on all materials, while I changed reflection level and glossines. It seems that some energy is not preserved... materials on the lower row appear darker than colors on higher row, isn't it?

[Juraj]

people are using values like [220, 220, 220] or even [180, 180, 180]. Now, RAL 9003 is [236, 236, 231]

Igni, early in this thread I mentioned the difference between Gamma 1.0 (linear values) and Gamma 2.2 (~sRGB values). You even write that you know of this conversion (by changing gamma in PS to get your translated values).

So there are two things to take in mind in your incorrect comparison:

When people say they use 180 or 220 for value, they mean 3dsMax color picker which is linear.
RAL 236 is sRGB value (everything is gamma 2.2 today because of LCD displays) =  [ 255*((236/255)^2.2) ] = 215 So it would be 180 vs 215, not 180 vs 236 !

Middle grey (50perc. value) is 128 in linear, and 187 in sRGB, it's pretty big difference. When you create textures, it's important to think of it.

Second, but more important, is that RAL colors are simple color exchange charts. They're perceptive colors, not physical values for Albedo. Although they probably might be some direct translation.
In any case, use them with VrayColorMap or CoronaSolidTex, or simple texture placed in diffuse slot.

I never really understood if perceptive colors include illuminant or not. My workflow is to feed renderer with physical albedo, which is the color/value without any illuminant, so I always make the materials slightly darker. How much ? I honestly don't know.
I still eyeball that stuff...it brings me to madness.

[ignicapillus]

Hello Juraj, thanks for prompting me CoronaSolid, I didn't know it.

In my work, I am often asked to test product colors before going in production; for some kinds of products and companies eyeballing is ok, while it's a pain when I must deal with other clients --among all, tobacco companies-- which are crazy about respecting their brand colors. Usually, it involves lots of Photoshop tweaks and digital print tests in typography. That's why the topic is interesting to me.

You're right -- in my example, I forgot to convert the RAL values, though I do it daily. Uff.

I completely agree about eyeballing: I refuse to accept that's the only method! There must be a way!

But sometimes I remember some company marketing director comparing a Pantone palette with a render on his monitor with odd gamma, with neon lights in the room, and commenting: "colors are wrong" -_____-

[Juraj]

Don't even tell me :- D

I do some renders for very big american furniture brand (Herman Miller) and I receive just some swatches. From me it goes through people who watch on uncalibrated iMacs,
often printing it with monitor profile from Photoshop which gives them different tones, then scanning it back and...
All that trouble, on top of me trying to figure out how to convert the swatches so they render out similar. Plus of course, educating clients than lights, angles, and shadows, do indeed
change the perceptive colors compared to studio or reference photography :- ). It's not easy.

You can get exact color from spectrometer, but that will basically give you a calibrated swatch. I still don't know how close or far that is from true material's diffuse albedo.