More neutral than almost all other gray cards

[Asher Kelman]

"Okay, I LIKE THIS thing - lol. This is hilarious - instead of "white balancing" on gray or white spots, I am "skin color balancing" using the skin color patches on this Colorright MAX disk thingy!!!

Oh, just in case it wan't obvious this MAX disk also does the instant normal white balance thru the lens just like the original Colorright or Expodisk, but it also adds the new skin color tool so when you take a picture of it (much like a gray card) you can use the eye dropper tool in your software...to tune in the skin to the right color by clicking on the skin tone patches. Kind of like gray or white balancing - but instead - you are skin color balancing. I LOVE IT!!!"

Peter Gregg
Creator of www.abetterbouncecard.com
Professional Photographer


"No fiddling whatsoever; no trip to PS to worry about overall skin balance."

"...On the engagement shoots I've been shooting with the ColorRight Max, I've had a mix of Caucasian skin types, but also African Canadian and Southeast Asian too. In each case, once I had adjusted for density, the ColorRight dialled in the skin tone, or so close to it that it's a lot less work to get right."

Jamie Roberts
Professional Photographer

Drew,

Could you explain how one deals with a mixed group and one eye dropper click? For an asian skin tone why would that not screw up the Caucasian or Black tones too? Then what happens to the white gowns?

Asher

 

[Doug Kerr]

Hi, Drew,

Please download the Raw files below the shot and click around on the patches to see the results of the ColorRight MAX.

[EDIT] Thank you for pointing this, Doug, I have corrected the original post with the hereunder sentence:

No need to shout in OPF with large fonts. Bolded red fonts should be enough to catch our attention…

[/EDIT]

Wow! The bright color and big font make this look like an e-mail message from a right-wing general! (But I knew it wasn't because all the text wasn't centered.)

But, your wish is my command.

I color corrected the first file ("Single Tungsten Light Source, Straight out of Camera, Auto White Balance") in DPP. I based the correction on eyedroppering on gray patch 1 and skin tone patches 1, 3, and 5. I also eyedroppered on the white of the subject's eye, and I did a final one with the tungsten preset.

The results were really neat, and I never had to touch a slider (good, because I've been suffering from a hangnail).

Here are the results (I cropped the frames to save real estate, and these are of course at reduced resolution). There was otherwise no processing other than the color correction I describe above.

I think this result tells me that the model is Asian and spends a lot of time in the sun.

This is a great thing you've invented, Drew.

Best regards,

Doug

 

[Asher Kelman]

Well Doug,

This is one child. Say you preferred patch #1. How then would you deal with a group of people with different skin color? Also what happens to the white shirts? For real high end work I can't see how one can do color optimization with one click!

Asher

 

[Doug Kerr]

Skin tone

Hi, Drew,

I'm afraid I just don't get the premise of the skin tone target patches on the ColorRight MAX.

According to the literature, the different patches (which will produce color correction that is offset from the "theoretical" correction by different but predictable vectors) are each suited for various kinds of skin tone, in terms of "lighter" vs. "middle " vs. "darker".

That would seem to say that the best offset from theoretically ideal correction (at least for starters) would be known based on the skin tone.

Is that it?

Best regards,

Doug

 

[Doug Kerr]

Hi, Asher,

This is one child. Say you preferred patch #1. How then would you deal with a group of people with different skin color?

I would welcome them all into our group, knowing that they were all God's children.

No, seriously, I would shoot a neutral target in the shot with all of them and color correct on it.

Then if somebody thought the most politically-important person in the shot looked too red, I'd shift it all a little to the blue.

Best regards,

Doug

 

[Drew Strickland]

I color corrected the first file ("Single Tungsten Light Source, Straight out of Camera, Auto White Balance") in DPP. I based the correction on eyedroppering on gray patch 1 and skin tone patches 1, 3, and 5. I also eyedroppered on the white of the subject's eye, and I did a final one with the tungsten preset.

I think this result tells me that the model is Asian and spends a lot of time in the sun.

This is a great thing you've invented, Drew.

Best regards,

Doug

Thanks, Doug.

Lots of people seem to agree. They like it too.

So, I take it you prefer a warmer skin tone like #3. As it says on the site you can use the asian/ medium skin tone settings to produce a "tan" look on a Caucasian. The settings are not there to determine for you whether someone is asian or not. I figure you can figure that out all on your own.

Of course your images above in a non-controlled color environment really don't tell much. People need to download the raw file themselves. In fact, even then most color managed workflows leave much to be desired. Not that I would recommend it, but you could use this tool in a completely non-color managed workflow and get good printable skin tones just by knowing whether their skin was lighter or darker.

The big thing to notice here is that you very quickly found a pleasing skin tone that you like, and it wasn't neutral.

No neutral product can do this for you.

Imagine a wedding with 2,000 to 3,000 images to process. This tool will lessen your workload considerably.

Now, I know what you're getting at Doug. You don't care about anything practical. You want to know how it works. What color theory it is based on. Sorry, not going to give away the store. I'm sure you'll figure it out eventually.

 

[Doug Kerr]

Hi, Drew,

So, I take it you prefer a warmer skin tone like #3.

I never said that. I have no preference.

The settings are not there to determine for you whether someone is asian or not.

I was just pulling your leg.

Of course your images above in a non-controlled color environment really don't tell much.

War, as Sherman said, is hell.

People need to download the raw file themselves.

Why would they want to do that?

Not that I would recommend it, but you could use this tool in a completely non-color managed workflow and get good printable skin tones just by knowing whether their skin was lighter or darker.

Lighter or darker than what?

The big thing to notice here is that you very quickly found a pleasing skin tone that you like, and it wasn't neutral.

I never said I like any of them. But I certainly wouldn't like a neutral skin tone (except of course if the subject was Caspar the Ghost).

Perhaps you mean a skin tone that was theoretically correctly balanced, as if done by reference to a neutral target.

No neutral product can do this for you.

Not even Johnson&Johnson's Baby Shampoo?

Imagine a wedding with 2,000 to 3,000 images to process.

Horrifying thought. I never take more than 1,000 at a wedding.

Now, I know what you're getting at Doug. You don't care about anything practical.

Ah, only six days to the election, and it begins! (Oh, sorry, I got you confused with another guy.)

I'm sure you'll figure it out eventually.

Actually, Drew, I figured it out long ago.

Best regards,

Doug

 

[Drew Strickland]

Drew,

Could you explain how one deals with a mixed group and one eye dropper click? For an asian skin tone why would that not screw up the Caucasian or Black tones too? Then what happens to the white gowns?

Asher

Well, in this case you would have to choose which way you want to go. If you had the whole spectrum of brown (all skin is greater or lesser amounts of brown with slight variations) you would probably be happiest picking the middle.

When in doubt, slight variations from neutral such as patch #3 or #8 might be your best bet. Either one of these are probably going to look and print better than a purely neutral rendering. The dress and white shirt won't show much variance either.

Hope this helps.

 

[Drew Strickland]

Actually, Drew, I figured it out long ago.

Best regards,

Doug

Good deal. Not surprised.

My best to Carla.

Drew

 

[Doug Kerr]

Hi, Drew,

Good deal. Not surprised.

My best to Carla.

Thanks. I'll pass it on.

Best regards,

Doug

 

[Ben Rubinstein]

Not likely to be that helpful unless you continue to use the whibal to get to neutral first.

Of course, you could just try out this new ColorRight MAX thingy that everyone is discussing. It even works on different skin types.

It will make a huge difference, the relative skin tones will be accurate whatever the WB chosen. For example I converted the D700 profile to be used for the 5D (a feature of DNG Profile Editor) and the differences, though subtle, are enough to bring depth to the skin colour tones that just wasn't there with the native 5D ACR profile.

Whether you need truly accurate WB depends on the subject. With faces you often want pleasing rather than accurate, I understand that is the premise of this new gadget being discussed here. However, whatever WB you choose, you want everything else to fall into place colour wise and that can only happen if you have an accurate profile, something Adobe is very bad at, especially for facial tones.

 

[Michael Tapes]

Hi, Drew,

I think this result tells me that the model is Asian and spends a lot of time in the sun.

This is a great thing you've invented, Drew.

Best regards,

Doug

I was going to say you can't make this stuff up, but you did! This is one heck of a thread. It is like being in parallel worlds. I can't wait for the next installment (post).

ROTFLOL!

I wonder how much he spends on wardrobe or if he is late on child support? Hmmmmmnnn

 

[Bart_van_der_Wolf]

Doug, I love your sense of humor! Thanks for the laughs.

Perhaps you mean a skin tone that was theoretically correctly balanced, as if done by reference to a neutral target.

And now to inject a serious note.

Skin is a hairy (sorry for the pun, I mean challenging) subject to shoot with a digital camera.
The outer skin layers (epidermis) are semi-transparent to infra-red. IR is diffusely reflected by the lower skin layers and vein structures. The IR content in the incident lighting depends on the colortemperature of the (incandescent) lightsource.

Silicon based Digital camera sensor arrays are sensitive to IR, and need to be filtered to remove a certain amount ot the IR radiation that is reflected by e.g. skin. It's the amount of remaining IR that's allowed to influence the color reproduction by the Bayer Color Filter Array and Raw conversion software that will determine how skincolor is reproduced, even with a neutral White Balance.

So even under a lightsource of known color temperature, the amount of residual IR that reaches the sensels will pollute the color reproduction of the RGB color layers. That will result in a camera, and Raw converter, specific reproduction of skincolor (despite the visual pigmentation).

That means that the best we can do is to get the White Balance (WB) correct (I prefer a WhiBal or Babelcolor target for that), and tweak the Color Balance (CB) for skin color from there. WB does not equal CB!

Drew's tool just offers a series of off-white patches which allows to abuse the White Balance controls to arrive at a more pleasing memory color we may have in mind with a single click. The correct way to do it would be by using the Color Balance tools on a selection of skin area(s), but that is obviously more time consuming than the quick-and-dirty approach Drew's tool offers. The Blue channel is a likely candidate (varies by camera and lightsource) for reduction of the skin color's saturation, so pulling it a bit more or less might do the trick for a variety of skin types in the same image.

So as long as Drew leaves out any reference to science, his solution is possibly good enough for a quick inaccurate (because it screws up total WB) overall fix, but it is nothing more than that. Whether that justifies the cost is left for potential buyers to decide.

Bart

 

[Doug Kerr]

Hi, Bart,

The outer skin layers (epidermis) are semi-transparent to infra-red. IR is diffusely reflected by the lower skin layers and vein structures. The IR content in the incident lighting depends on the colortemperature of the (incandescent) lightsource.

Thanks for this interesting information about an important matter. I was myself little aware of all this.

Best regards,

Doug

 

[Drew Strickland]

something Adobe is very bad at, especially for facial tones.

No argument with this statement.

Check out C1 for better color profiles Or, the new DNG profile editor you are using. Haven't gotten around to trying it out yet. Glad to know it is helping you out.

However, the point I was making is that there is still a need for a white balance reference tool during capture. Because your profile will make precious little difference if the wb is way off.

 

[Drew Strickland]

Skin is a hairy (sorry for the pun, I mean challenging) subject to shoot with a digital camera. The outer skin layers (epidermis) are semi-transparent to infra-red. IR is diffusely reflected by the lower skin layers and vein structures. The IR content in the incident lighting depends on the colortemperature of the (incandescent) lightsource.

Silicon based Digital camera sensor arrays are sensitive to IR, and need to be filtered to remove a certain amount ot the IR radiation that is reflected by e.g. skin. It's the amount of remaining IR that's allowed to influence the color reproduction by the Bayer Color Filter Array and Raw conversion software that will determine how skincolor is reproduced, even with a neutral White Balance.

So even under a lightsource of known color temperature, the amount of residual IR that reaches the sensels will pollute the color reproduction of the RGB color layers. That will result in a camera, and Raw converter, specific reproduction of skincolor (despite the visual pigmentation).

Couldn't have said it better myself. However, the problem is further muddled by different light sources. Have a look at Melissa's skin tone balanced on the neutral patch in the bottom two photo examples above. Her skin looks quite different with the same neutral patch due to the difference in the light sources. One is standard tungsten and one is "daylight" balanced fluorescent.

That means that the best we can do is to get the White Balance (WB) correct (I prefer a WhiBal or Babelcolor target for that), and tweak the Color Balance (CB) for skin color from there. WB does not equal CB!

I disagree here. The best we can do is deliver pleasing skin tone to our clients. Some images require more adjustment than others. With 6 patches that all push the color ratios in the right direction you are on much more solid footing.

more pleasing memory color we may have in mind with a single click. The correct way to do it would be by using the Color Balance tools on a selection of skin area(s), but that is obviously more time consuming than the quick-and-dirty approach Drew's tool offers.

More pleasing. That's exactly right. Not more accurate. But, you can get that too.

Quick and dirty? Abusive?

How about we call it a very handy time-saving shortcut. No need to get into the curves dialogue in a full blown PS session.

The Blue channel is a likely candidate (varies by camera and lightsource) for reduction of the skin color's saturation, so pulling it a bit more or less might do the trick for a variety of skin types in the same image.

Blue is part of it. But certainly not all.

So as long as Drew leaves out any reference to science, his solution is possibly good enough for a quick inaccurate (because it screws up total WB) overall fix, but it is nothing more than that.

Which is it we want? Accurate or pleasing?

The ColorRight MAX gives you both with one click.

As regards science. I think there is a great need for a reference standard for skin tone ratios. The ColorRight MAX provides just such a reference standard as all products leave the factory with similar color specs. So, someone with a MAX in Canada can talk to someone with a MAX in Singapore and discuss their use of different standardized color patch numbers on the ColorRight MAX.

 

[Jack_Flesher]

Interesting set of remarks...

CaptureOne software (touted on another thread here) now has a "skintone" dropper tool with several skintone presets loaded in it. This is a welcome addition for wedding photographers who want to insure the bride looks the same in all of her images, regardless of what light she happened to be under when the shot was taken. (Ever try to WB a bride lit by light bouncing off a pale green wall that originated via a stained glass window?) Of course using a skintone preset "screws up" absolute white balance (and apparently causes color purists to shudder at the thought of potential inaccuracy) but the end result is the person paying the photographer is happy. Personally, I'll take the hard currency over the absolute science any day of the week...

Anyway, it seems to me Drew's tool provides a similar benefit for those who might desire it and don't have C1 --- Good job Drew!

Cheers,

 

[Bart_van_der_Wolf]

As regards science. I think there is a great need for a reference standard for skin tone ratios.

For digital camera sensors? Ain't gonna happen! That was the crux of my post.

The IR response of the sensor array, after IR filtering, plays a crucial role in the rendition of skin color. Deep skin tissue is a good diffuse reflector of IR, and it's the IR content that's going to throw skin color rendition off. It will expose the blue channel as efficiently as the others, while it shouldn't. That is added in the mix of visual color wavelengths dominated by skin (pigment filtered) reflection.

It is the mix of IR and visual light that needs to be Color Balanced (due to IR contamination) and White Balanced (due to dominant illuminant).

The ColorRight MAX provides just such a reference standard as all products leave the factory with similar color specs.

Does the ColorRight factor in enhanced IR reflection by human skin? No.
Does it know the IR filtration characteristics of the camera it is shot with? No.
It's just a crude tool to throw the entire White Balance off a bit, in order to achieve something else, namely Color Balance for a range of surfaces. It may be used for a quick fix, but it isn't a reference. IMHO, of course.

Sorry, no sigar.

Bart

 

[Bart_van_der_Wolf]

Interesting set of remarks...

CaptureOne software (touted on another thread here) now has a "skintone" dropper tool with several skintone presets loaded in it. This is a welcome addition for wedding photographers who want to insure the bride looks the same in all of her images, regardless of what light she happened to be under when the shot was taken.

Yes, but it is a Color Balance tool, not a White Balance tool.
Look at http://www.phaseone.com/4/?pageid=buy_now , click on the "GO for PRO" tab, and select the Color tool under the Video Tutorials section.

A Raw converter, if done well, can use prior knowledge about a given camera's IR rendering characteristics, a reflective target cannot!

Bart

 

[Jack_Flesher]

Yes, but it is a Color Balance tool, not a White Balance tool.

Uh Bart... As soon as you click a converter WB dropper on a patch of Drew's tool, it is a color-balance tool too --- he's done the color inversion math so the patch delivers the desired result when used in conjunction with a dropper that's calibrated to neutral. The only difference is it's a limited collection of presets for skintone, where the C1 dropper can be used for ANY user-defined color...

We can debate semantics and the science of how they're implemented all day long, but the result is the same -- a happy bride and paycheck in hand

PS: Dealing with out of spectrum color sensitivity is a separate issue.

Cheers,

 

[Bart_van_der_Wolf]

Uh Bart... As soon as you click a converter WB dropper on a patch of Drew's tool, it is a color-balance tool too --- only difference is it's a limited collection of presets for skintone, where the C1 dropper can be used for ANY user-defined color (even Drew's set of skintones )...

Hi Jack,

The important difference is that a White Balance tool affects ALL colors, while a Color Balance tool affects a certain range of colors, while leaving the rest alone.

Bart

 

[Asher Kelman]

Ever try to WB a bride lit by light bouncing off a pale green wall that originated via a stained glass window?) Of course using a skintone preset "screws up" absolute white balance (and apparently causes color purists to shudder at the thought of potential inaccuracy) but the end result is the person paying the photographer is happy. Personally, I'll take the hard currency over the absolute science any day of the week...

Jack,

The problem is not just the girl's face, her dress too and his shirt. I have a white balance of a card held in that light. I correct that then locally correct the face. Bibble 4 will allow local changes so there one could develop from RAW allowing for the dress to be white and the face to have its own reference. We are really talking about time and skill. I could use a Canon G9 on AUTO and take a bunch of great pictures.

If we are doing things on a production line, perhaps it might be "good enough", but it will likely fall short of excellent. I've never seen you do sloppy work and you would never do that for a bride!

I will try it, for sure!

Asher

 

[Drew Strickland]

For digital camera sensors? Ain't gonna happen! That was the crux of my post.

Hi Bart,

Sorry, but I think you missed the point of my post.

By your definition, neither the WhiBal, nor the BabelColor are reference tools either. They are subject to all the variations you mention. According to your definition neutral should be neutral should be neutral. But it is not.

If that is your point. I guess you can try and argue that point.

A standardized reference tool is simply a tool that is measured by some other accepted standard to be x or y.

The WhiBal and BabelColor tools are only a reference in that they are measured by an accepted measurement tool such as a spectrophotometer. Our colors are measured by the same tool and we can easily guarantee a certain variance for each color patch.

The reference here is not the final result. The reference is the tool. You will, of course, encounter variations based on conversion engines, camera hardware, and the like. Sara made this point in her most recent post.

Again, to see this illustrated just look at the two images above of Melissa. When balanced to neutral using the same reference standard of neutral, the colors are quite different in the scene even when keeping all other variables constant. The only major variable that differs is the light source illuminating the scene. They are not just different on the skin.

 

[Doug Kerr]

Hi, Jack,

The only difference is it's a limited collection of presets for skintone

The various patches on the ColorRight MAX do not produce various skin tones - they produce different offsets of the corrected image chromaticity from that which would be attained using the "neutral" target (which we might characterize as the "theoretically ideal" color balance).

That's probably what you meant, but I am always worried about "shortcuts" in language.

Best regards,

Doug

 

[Doug Kerr]

More neutral than almost all yellow cards

Hi, Drew,

According to your definition neutral should be neutral should be neutral

There is a useful absolute definition of a chromatically neutral reflective surface: one from which the chromaticity of the reflected light is the same as the chromaticity of the incident light. We must of course state what definition of chromaticity is to be followed (perhaps CIE chromaticity).

Now, we can have one for which that is true for any illuminant (basically requiring it to be spectrally uniform), or we can have one of which that is certainly true only for a certain illuminant.

A practical meaning in between (ohmigawd, I said the "p" word) is one for which that is true for the range of illuminants that are commonly of interest as photographic illuminants, perhaps the range of "blackbody" illuminants.

A wholly different issue is how precisely, and how accurately, can we determine this property with some real instrument.

But the fact that no instrument is "perfect" does not change the basic nature of the property in which we are interested.

Best regards,

Doug

Older than John McCain, but only for a few days,
and not interested in anything practical

 

[Drew Strickland]

Interesting set of remarks...

CaptureOne software (touted on another thread here) now has a "skintone" dropper tool with several skintone presets loaded in it. This is a welcome addition for wedding photographers who want to insure the bride looks the same in all of her images, regardless of what light she happened to be under when the shot was taken. (Ever try to WB a bride lit by light bouncing off a pale green wall that originated via a stained glass window?) Of course using a skintone preset "screws up" absolute white balance (and apparently causes color purists to shudder at the thought of potential inaccuracy) but the end result is the person paying the photographer is happy. Personally, I'll take the hard currency over the absolute science any day of the week...

Anyway, it seems to me Drew's tool provides a similar benefit for those who might desire it and don't have C1 --- Good job Drew!

Cheers,

Thanks, Jack.

The Capture One approach is somewhat innovative, but still requires a neutral target be in the image to begin to use it properly.

Here is an excerpt from the manual:

How can I define a certain skin tone on my models?
Often it can be troublesome to achieve colors of skin, or an exact color of clothing, especially when photographing in different light conditions.
Skintone is working like the well-known White Balance, but applies
to any color defined in advance.
Mark up “Pick to create new” and select a picture showing the correct colors, pick the color using the Skin Tone picker, automatic you will see the save-as dialogue, then choose a relevant name for your new skin tone.
Skin Tones can be used to define all colors, e.g. the blue in a shirt can be defined by a dedicated studioshot before going out-door for the model shot, and when a Skin Tone is defined and saved you will be able to use the tone in any session.
What are the differences between defining skin tone and setting white balance?
White Balance is created to neutralize the color picked, to a perfect grey balance; this is why the grey-card is needed to exactly point out the grey tone.
The Skin Tone tool is created to adjust a selected color to an already defined
color. by adjusting the white balance of the capture.

You have to first have a photo with the correct skin colors to begin with.

Admittedly, I have played around with it very little. But, so far I have not been able to get anywhere near as good a result out of the tool without a neutral reference in the frame. Even then the included skin offsets leave much to be desired.

So, in a workflow sense. If want to use C1 well with the skin tone tool you still need to use a white balance reference and then try to balance.

The question is, if you have to already include a white balance reference and click on it first for each lighting environment you encounter on your wedding shoot for this system to work- why would you not just go ahead and include the MAX tool that includes neutral and one click skin tones.

The net of the C1 approach is a lot more work for a less satisfactory result.

 

[Bart_van_der_Wolf]

The problem is not just the girl's face, her dress too and his shirt.

Bingo! A WB offset will change everything.

I have a white balance of a card held in that light. I correct that then locally correct the face.

That's how it should be done. One should WB for the dominant illuminant, then CB for a (local) range of colors. A single click 'reference' can't do both at the same time.

Bart

 

[Jack_Flesher]

they produce different offsets of the corrected image chromaticity

Fair enough, we can call them a set of "predefined offsets" instead.

from that which would be attained using the "neutral" target (which we might characterize as the "theoretically ideal" color balance).

Have to take exception here --- a "neutral target" is only truly neutral under a predefined lightsource. I assume you meant to include the caveat of it being a full-spectrum lightsource (ie the Sun), which helps with response linearity, but do not know if we'd agree on D50, D65 or something in-between for color temp? Your "theoretically ideal" white balance is thus problematic for me...

We could next start debating the relative merits of Robertson versus Judd and transform matrices, but surely we'll put everybody to sleep... I know I've already bored myself.

Cheers,

 

[Jack_Flesher]

Admittedly, I have played around with it very little. But, so far I have not been able to get anywhere near as good a result out of the tool without a neutral reference in the frame. Even then the included skin offsets leave much to be desired.

SNIP

The net of the C1 approach is a lot more work for a less satisfactory result.

I agree your tool is certainly easy and convenient. However, the C1 approach is pretty easy too and the results are impressive. It's strong suit is it maintains color uniformity regardless of changes in source-light temperature --- and thus *really* useful for those who shoot corporate logos or trademarks (like the dreaded "Pepsi Blue") where the client insists on exact color reproduction .

But as you indicated, to use it you do need a good starting point. So an initial frame under a clean lightsource (I shoot people mostly under strobes so this is easy for skin) with your preferred form of WB tool -- I use one of the Betterlight sandable cards, but would as happily use yours if I had it -- then a click on an even-toned patch of skin (or whatever item you want to save the selected color for), then save with name of choice. You can further edit tone and application limits if desired.

Cheers,

 

[Doug Kerr]

HJi, Jack,

Fair enough, we can call them a set of "predefined offsets" instead.

Have to take exception here --- a "neutral target" is only truly neutral under a predefined lightsource.

Well, if it is in fact spectrally uniform, then it will be chromatically neutral under any illuminant. (That is, whatever the chromaticity of the incident illumination (under any definitive of chromaticity we care to adopt), then the reflected light will have that same chromaticity.

Your "theoretically ideal" white balance is thus problematic for me...

Well, indeed there is no such thing except for a scene of all well-behaved surfaces!

So I should probably stay clear of the term unless carefully caveated!

Thanks for your inputs.

Best regards,

Doug