White Balance Part II: Accurate, or Convincing?

By Carl Weese

The first part of this series described what constitutes an accurate white balance, and how to get there with a digital camera. Part II will deal more with visual judgment and the idea of "convincing color." Here are a pair of files, a test and a picture, shot under winter overcast morning light, viewed together in an ACR filmstrip window:

There's a big white balance error here (the overexposure and even the burned out red channel, which is actually caused in part by the white balance error, don't matter because the gray area of the card has plenty of tone). A number of digital cameras I've used get confused by a subject with a lot of green, especially under cool lighting. At risk of anthropomorphizing, it seems like the AWB software says "Wow! look at all that green, we must be in fluorescent light, better crank up the magenta on the Tint slider." This is usually accompanied by a Temp setting that is lower than normal daylight, when it ought to be higher. The WB here is 4800/+19. That's actually a setting you might find correct for a mix of fluorescent light with cool daylight. Shooting on a rainy day in an evergreen forest I've seen AWB go for color that reads out 4100/+35, which actually is about right for warm-white fluorescent tubes.

The "real picture" was shot on AWB as well, but the software has done much better. The numbers are 5000/+1. The grass and the winter forested hillside in the background strike me as quite convincing. The AWB had much less green subject area to deal with and the slice of overcast sky gave it something to grab onto. It "feels" about right for the cold winter light, but there's a problem. What drew me to the subject was the stone of the monument, which has very unusual colors, at least for around here. So let's select both files, click the eyedropper on the WhiBal card and then switch the main view to this picture again.

This nails the unusual colors of the monument, though it has lost the sense of cold winter overcast. Still, for this picture, I'll take accurate presentation of the stone over a more convincing rendition of the light. The numbers are 5500/-7, which is in the expected range for overcast light recorded with this camera.

One of the first things you should do to begin using WB calibration is to shoot multiple tests with your equipment in standard situations like clear sunlight, and overcast light. Be sure to do these tests out in the open where there is no reflected light from bright colored objects (including your clothes!) to contaminate the reading. While in theory sunlight should be 5000/0, the raw development software is seldom that perfectly calibrated to the camera. The Pentax K20D used here gives neutral color under clear sun at 5300/-6 in ACR, while an Olympus E-1 I've used for years comes in at 5400/+14. It can be interesting to know how far from standard the color of the light is in a given situation, and for that you have to know what your camera's "standard" numbers are.

Another pair of pictures, this time shot in pretty weird conditions. It was a dark day, light rain falling, and the sidewalk and street were heavily overhung by a green canopy of leaves—there's a row of trees growing at the top of the retaining wall. AWB has made a valiant effort to cope, turning in numbers of 6750/+2. AWB has compensated +1450 temp and +8 tint compared to where it would set for ordinary sunlight, because it knows this sure ain't clear sunlight. Actually, it really felt that cold and blue standing there in the rain, but these colors are pretty livid and I know that they'll look even stranger printed than they do on the monitor.

So here's the test frame after clicking the card with the eyedropper. Look at those numbers, 8800/+4. That's very weird light indeed. The card and the hand look good. However...

This is all wrong. The test card reading has been applied to the picture file. Okay, those are accurate renditions of the subject colors, but it totally misrepresents the scene. Everything seems bathed in a glow of clean sunshine. The color of the light (sunny) is totally at odds with the tonal quality of the light (heavy, soft overcast). The whole point of the shot is lost. How to get back the feel of the day and the strange location?

Some time back I was working on a set of pictures made in overcast light and through tedious trial and error finally arrived at white balance settings that seemed convincing to me. Some of the takes included WhiBal test shots so I reverse-engineered things a bit. I applied the adjusted Temp/Tint settings from the good-looking files to the test card files, then used the eyedropper to measure the r/g/b readings on the test card. It turned out that the white balance that gave convincing portrayals of overcast light had r/g/b values where the g was a couple points higher than the r, and the b was about five points higher than the g. Then I went looking through my archive folders of raw files for pictures that, after laboring over, did a good job of presenting really dark, stormy light—pouring down rain or threatening snow. Repeating as above, I found that for these g ran about five points over r, and b about ten points over g, instead of the dead-even numbers of an accurate gray card calibration.

Right after I'd finished doing this, Oren Grad asked me to look at some snow scenes under overcast skies he'd struggled over to get the color just right before putting them up on his website. I thought they looked perfect on screen, so I pulled them down to my desktop and tossed them into Photoshop so I could get the eyedropper on them. The light gray snow showed r/g/b values such that g was a couple points higher than r, and b was about 5 points higher than g. To have a colleague confirm your results by request is nice, but to find him doing it by independent discovery means you might actually be on to something.

So, I selected the test file again from this screen, then hit "Select All," then clicked the Temperature slider to activate it, and placed the cursor on the test card. Next I used the keyboard down arrow to tweak the temp setting lower, watching the numerical rgb readout of the card and the thumbnail of the other file. When the temp setting got down to 7250, the color of the thumbnail was starting to look pretty good and, sure enough, the r/g/b numbers for the test card were 174/179/188. I switched the main view to the picture file and got this:

Much more like it. So here are three white balance versions, one set by the camera's software, one set as accurately as possible using the test card, and one set to what strikes me as a convincing interpretation of the scene as I experienced it. None of them is "correct." You don't have to agree with my choice. You might think the scene works best at another intermediate balance, or even farther to either end of the scale.

In Part I, I mentioned that portraits probably ought to have accurate color, but I'll modify that to say that I suspect a commercial portrait photographer won't get any complaints if she routinely tests for accurate white balance and then bumps the temperature up 300 or 500 degrees. Warm is happy and friendly.

Household tungsten lamps have really low color temperature. Even though our eyes adapt to this so that we don't generally perceive household lighting as yellow or amber, pictures in living room lighting often don't look right when the color is fully corrected for accuracy. It's the other side of the coin from my rainy day under the trees picture. Maybe this is because somewhere in the back of our heads we think, or feel emotionally, that indoor light should be warm. Paintings from hundreds of years before the invention of color photography often depict fireside and candlelight scenes with a really warm amber palette far removed from an accurate presentation of subject colors. As with the previous example, a test exposure that tells you what white balance is accurate is a helpful reference even if you decide that a warmer rendition will make a more convincing presentation.

If you are in light that is likely to fool the camera's AWB and you plan to use the camera's histogram for exposure evaluation, it's a good idea to select a manual balance or attempt a custom white balance to get the color closer. Even though you can correct the raw file for white balance based on your exposure of a test target, the camera histogram, based on a badly color-balanced camera JPEG, may be so far off that color correction will end up clipping one or more channels. If the WB is set close to accurate color, the histogram will be a much more accurate exposure indicator, and you can still fine-tune the raw files for color using the test card exposures. Note, changing the white balance setting in-camera doesn't have any effect on the raw data, but making the JPEG-based histogram more honest can avoid mistakes based on a faulty histogram.

You can also save specific white balance settings that you find recurring in your work, like those clear sunlight numbers I mentioned earlier. With PS, you can save a white balance setting in ACR as a "custom subset." You can then apply it to new files either in ACR or Bridge. It can be informative to toggle between what the camera's auto white balance came up with and your standard daylight or overcast settings, for example. If you do a shoot in a place with weird light, you can save a custom setting for that, and it will prove useful during your editing. You can always delete the setting to keep the panel less cluttered, after you're done with that shoot.

White balance is both a technical tool and an aide to expressive printing. Use of a neutral test target makes quick work of getting the most accurate color your camera can come up with in a given situation. At the same time an accurately corrected test file gives you a helpful visual reference point for expressive use of color that gets across your personal interpretation of a scene.

________________________

Carl

See more examples at Working Pictures

Send this post to a friend

Comments

[Louis McCullagh]

I hope this is not a too pedantic restating of above.

Using a grey card or expodisc gets you to a "pseudo-real" neutral colour base. It does not give you the REAL colour.

The 'REAL colour - outdoors' changes with the seasons, time of day and weather conditions (cloud).

If you colour balance using a card/device it removes the "seasonal" "time of day" and "weather" colouration.

Example: A photo at sunset if colour balanced using a device/card will end up looking as if it was taken (in colour terms) like a mid day scene.

A card/device tells you what to do to get an image back to a neutral "pseudo-real" colour base it does not give you what the light colouration was at the time you took the photo (it does not tell you what to adjust the colour by to remove (a)the camera colour bias and (b)the colour of the outdoor light.

Notionally you could calibrate your camera using a known light eg 5500K but that is only part of the story as it can be that your camera needs different calibrations measured with a range of lighting eg tungsten, sodium, flourescent etc etc.

Nitpicking is mostly a waste of time because you most often can't control the colours when your viewer sees it.

The factors will be the colour of the light they are viewing under, the paper the image is printed on, the fidelity of the reproduction by the printer (machine/person) or simply the monitor calibration/quality.

All you can aim is to be in the ballpark using the grey card/device and the ADJUST back to the light colour at the time the image was taken (assuming you liked/want to show what the light was like at that time e.g. sunset, sodium light etc.)

A question.

Has anyone used a standard set of adjustments to change a neutrally balanced image to a sunset/tungsten etc.(or are the ACR adjustments just perfect ;-) )

Two+ adjustments!
Beware Photographing people on lawns or under trees or anywhere that the light is reflected off a non neutral surface.

You will find that under chins /arms etc are a different adjustment to the face. This is because (I presume) you take your reading (from device/card) pointed from the face back to the camera position.(a fill-in flash might be helpful).

You need to know what colour you want at the time of shooting and be aware of what is hindering you from getting what you want.

[Michael Tapes]

Carl,

Excellent investigation of a complex topic.

Michael Tapes

[Adam]

Carl,

The pictures you selected do an excellent job of demonstrating your point, but I'm not sure that I fully understand your technique for getting something that approximates your subjective view of a more appropriate rendition of overcast or stormy light.

(1) First you set a "neutral" white balance using the eyedropper/gray card settings from your reference picture.

(2) You adjust the temperature slider until the picture looks about right.

(3) You check with the RBG sampler to see if the G is about 5 points above R and B is about 5 (or 10) points above G.

I'm not sure I understand why (3) is really necessary, if you are satisfied with your result from step (2). Assuming you get pleasing/subjectively appropriate lighting out of step (2), but that the results don't match your expectations in step (3) with regard to the relative RGB values, would you then ignore the result of step (2) and keep adjusting the white balance to get a result that satisfies step (3), or would you stick to your step (2) result? I'm not trying to be provacative, this is a sincere question. I'm trying to understand how you value the subjective step (2) vs. the somewhat more objective step (3).

In a related question from the other side of the coin, is there any way to go directly from (1) to (3)? If you save a custom white balance setting in your RAW converter, it just saves the actual color temperature setting, right? If you just apply the same color temperature to other pictures in slightly different light, I would expect the results to be "off". What you need is some sort of macro or action that takes a "neutral" result and then adjusts the RGB values relative to eachother to match your preferred ratios. Is that possible?

Last question: it looks like you are using the keychain version of the WhiBal card. Is there any benefit to using a bigger card (or any disadvantage to using the small card)? In theory I suppose that if your card is big enough, you could use it to set a custom WB in the camera, but that seems like it would be impractical in most situations, since the card would have to be pretty big. Are there any other advantages I'm missing?

Best regards,
Adam

[Dierk]

Yip. Another subject where WB'ing becomes problematic is sunsets/-rises - one really doesn't wnt to correct them towards neutrality. A grey card should look anything but neutral under these circumstances.

This holds true for all photos in which the [strange] light is the actual subject, from modern functional architecture, like subways [British usage!] or railway stations, to mixed lighting, coloured lights and night shots.

Bottom line: It's your decision, not the environments.

[Martin]

Carl,

thanks for this informative articles.

"Accurate, or Convincing?" that's the real question.
Here my thoughts about it.

We can neutralize certain sources of light by selecting the WB for neon light, lamps, shade etc. Or with a white balance on a suitable medium exactly on Kelvin, we can prevent the light dyeing its environment after own liking.

But what the heck with it? What is the reason for our neutralization illusion?
Naturally, our eyes have the ability to see white as white in each color of light. The experience library of our brain makes it possible. Even mixed light situations we adjust on a perfect way.

Only the view from our illuminated living room window outside into the dawn, permits us, to see the blue hour, or to look from outside into warmly illuminated dwellings.

But already briefly after we entered the pure area of a wafer-producing company we see nothing more from the intensive-yellow lighting, which prevails there.

What do we make now? Do we place the yellow light neutrally? Do we let the employees carry white or yellow work clothes? Does a the Blue Hour become a Grey Hour?

I make an experiment for some time: if I photograph outside with daylight, I set the camera on 5400 Kelvin, in artificial light to 3400 Kelvin. And actually, in the light ambiance swings again, sometimes coldly, sometimes warm, sometimes blue, sometimes green.

I make exceptions with flash photographs and mixed light. Because my experience is, that in these situations the AWB of the camera supplies more plausible results. Perhaps, because electronics makes better of a shady business than I could do.

Honestly, if we see a photo, in which the natural lighting effect is conserved, we do not offend as an avoidable technical lack.
On contrary! If I often regarded neutral colors within my own photos as a relating to crafts, I could regard the pictures of colleagues more relaxed and notice, that not only light and shade, but also “ambience” set accents.

Courage for ambience!

(Please excuse my bad english)

[Carl Weese]

"Has anyone used a standard set of adjustments to change a neutrally balanced image to a sunset/tungsten etc."

No reason not to, as long as you realize that the idea is to get a quick starting point to deal with future pictures. The article mentions my starting points for overcast and stormy light. Another example is fluorescent where I find the sense of the light is often best if I get the r and g perfectly matched but with 5-10 points higher b reading. I want to get rid of the green cast, but find a cold look feels like the actual light. For some reason I never seem to shoot under tungsten lights but for anyone who does, a bit of experimenting could find a custom starting point. I suspect it would be neutral r/g with a lower b reading.

[Carl Weese]

Adam wrote: "(2) You adjust the temperature slider until the picture looks about right.

(3) You check with the RBG sampler to see if the G is about 5 points above R and B is about 5 (or 10) points above G."

In the example, these are not two separate steps, even if I didn't make that clear. I'm watching the picture thumbnail and the numerical readout both as I adjust the WB sliders. The visual clues and the numbers are a sort of reality check on each other. So this is a single step. The next step is to switch to a full screen view of the actual picture and make the final fine-tuning as a purely visual judgment (in this case I didn't make any further changes). The weirder the light or the stranger the subject colors, the more valuable the "abstract" numerical values obtained from the test card will be.

[Carl Weese]

"Last question: it looks like you are using the keychain version of the WhiBal card. Is there any benefit to using a bigger card (or any disadvantage to using the small card)?"

The benefit of the keychain card is that it's always there in one of your pockets. The disadvantage is that it can be difficult to get a nice big image of it with a short lens without bringing it within inches of the camera. I actually use the shirt-pocket size more often.

[Aaron]

White balance is something I struggle with. I'm often tempted to pick up some neutral device (all of which promise to make WB easier). For situtaions where you just want to reproduce acurate color, I get it.

However, for lots of other photos where perfectly neutral color isn't what you're shooting for, why use it at all? If you're fiddling around with the color balance until it looks right, how does starting with a neutral point help with that? If possible, can you explain the "reverse-engineering" stage a bit? What are you gaining from having a neutral reference point?

Thanks!
Aaron

[Eduard de Kam]

Carl,

I just do not agree with using a white balance target under natural light, wether sun or cloudy. The grey balance tool has just one function: make the color of the lightsource disappear from the picture. That is fine if you wont to reprooduce the colors of a painting under whatever light, or make a portrait there.

But when I take pictures outdoors, there is always the color of the light that I wont to see in my pictures because that is what produces the feeling for the place or the weather.

So I never use the AWB,m it is never right.
I use the presets of the camera, for daylight most of the time. The cloudy setting is to warm mostly, but for any camera make but Canon one can change this setting to a more blueish color.

Of course, as we are all shooting in RAW, we can make presets in the RAW conversion software to do basically the same. I use the same white balance setting for all my outdoor pictures, unless its late in the evening. This also solves the sunset sunrise colorproblem, the colors just come out right.

[John Camp]

When film users deployed filters or gels to balance the light or get the effect they wanted, they actually had quite a range of implements to either screw on their lens or cover their lights with.

I wonder it WhiBal or some other lighting group could come up with a set of virtual filters -- just a set of RGB numbers -- that you could apply after using the Whibal card to get a neutral position. In other words, all your photos would be neutral, but then with a figurative flip of the wrist, you could get sunrise/shadow, sunrise/open, sunrise/stormy, midday/shadow...etc., with the conditions described in words, rather than in temps. It does seem to me thatthere is some value in starting with a neutral position all the time... I could see the possibility of dozens of virtual filters -- maybe even , a sort of Tiffen color drawer, based on degrees kelvin, that would get you very close to the actual (subjective) light, starting from a neutral position.

Say...could this be a big moneymaker?

JC

[Carl Weese]

Aaron and Eduard,

If sensors reacted to changes in ambient light in a way that mimicked our senses we could leave the white balance at daylight and be done with it. But they don't.

Individual perceptions will vary on this, but in general light that varies substantially from neutral sunlight will be recorded in a way that looks exaggerated when we view the picture. We sense that open shade is cool, but not as cool as the camera records it at a 5000K white balance.

This is a subjective call, and "struggle" is an excellent term for what happens when a picture taken under unusual light "looks wrong" and we try to find the right WB for it by just diving in and playing with the sliders. This is especially true when both the temp and tint sliders need substantial adjustment--how much of which to use? This can turn into a real whirlpool. A test card exposure gives a helpful starting point and primarily makes it quicker and less frustrating to find "convincing" color.

As several have noted, one exception is when the light itself--not something illuminated by the light but literally the light itself as when photographing a sunset sky--is apt to look just right if the camera is set to manual daylight WB.

[Kurt Shoens]

The LAB color space offers some advantages, both in evaluating color and in applying a similar shift to different images with a different starting white balance.

Since LAB separates color from luminosity, you can see patterns, casts, etc. in the values and ratios in the A and B channels more easily than in RGB. Plus, even in the absence of neutral reference points you can infer (or sometimes guess) the color of ambient light source(s) and respond as you like. You can see LAB readouts in the Photoshop Info palette without changing the color mode of a picture to LAB.

You can also apply color shifts across the luminosity of an image (if that's what you want) with curves on the A and B channels. If you want that same shift in multiple pictures, you can copy the curves layer from one image to another.

(If you don't want the color shifts applied uniformly across luminosity, curve in RGB or CMYK instead.)

[Carl Weese]

JC,

The problem I see is that unless individual camera calibration is perfect, all those virtual filters will be inconsistent. My K20 and my E1 are a couple hundred degrees Kelvin and, more dramatic, nearly twenty points of "tint" different in clear sunlight. Apply a "cloudy day" virtual filter that really just sets 6000/0 and...oh, Adobe already has half a dozen of these in the White Balance pulldown menu .

Add to that the fact that interpretation of convincing color is subjective--your notion of deviation from neutral rgb to convey a sense of overcast light could be quite different from mine--and it seems more sensible for users to do a little experimenting and make their own set of virtual filters (for each camera they use) which can be saved as custom subset->white balance, in ACR.

[Rod Graham]

"If you are in light that is likely to fool the camera's AWB and you plan to use the camera's histogram for exposure evaluation, it's a good idea to select a manual balance or attempt a custom white balance to get the color closer. Even though you can correct the raw file for white balance based on your exposure of a test target, the camera histogram, based on a badly color-balanced camera JPEG, may be so far off that color correction will end up clipping one or more channels. If the WB is set close to accurate color, the histogram will be a much more accurate exposure indicator, and you can still fine-tune the raw files for color using the test card exposures. Note, changing the white balance setting in-camera doesn't have any effect on the raw data, but making the JPEG-based histogram more honest can avoid mistakes based on a faulty histogram."

Carl, it's been bothering me that I'm told to pay so much attention to the histogram when it's actually based on the jpeg image and not the raw. You clearly state why and how we should be doing this. Thank you.

[Tom K.]

My head is spinning.

**note to self** Study Black and White photography and edit exclusively to B&W on every image for the rest of my life.

**note to readers** I'm color blind. I can see color......just not like a non-color blind person. I can't tell the difference between red, green and brown in most situations.

B&W could free me from the shackles of color angst.

Superb column. I have some thinking to do.

[ILTim]

Interesting article! I have struggled with this, particularly with overcast. Finally I shoot raw+jpg and leave Daylight WB set almost all the time, that way I have a jpg reference that while sometimes exagerated, does show the 'mood' of the light. Being neutral is so far off at times that I wouldn't recognize the picture I took.

I loathe residential lighting, and I dont mean as a photographer, I really hate artificial lighting. While my eyes might 'adjust' and see normal color at home, I'm still QUITE aware of the odd lighting. Making an indoor photo neutral freaks me out! Sometimes I'll neutralize it (if it seems sensible), but often I'll leave it at/near daylight orange cast and all.

I hadn't realized I was cheating myself with the histogram though, thats a great point. Anyway, my view of reality (after chasing accuracy/neutrality) is that Daylight WB is most realistic.

[Adam]

JC/Carl,

Actually, I think JC's idea would work, I just think that it probably wouldn't be necessary. Here's why:

If I understand correctly, John is suggesting that the filters be applied AFTER a "neutral" white balance has been set using, for example, a WhiBal reference shot. This step should eliminate the impact of white balance differences between cameras. What John is suggesting is that you should then be able to apply virtual "filters" that would adjust the white balance for shade, open sun, etc. In theory, these would just be adjusting the temp/tint values BY a fixed amount, not TO an absolute value. So for "Cloudy Day", the filter wouldn't just set 6000/0, it would set X+1000/Y+0, where X is the temperature and Y the tint of the picture AFTER a neutral white balance has been set.

I think that some programs can already do this (Ctein has been supportive of Picture Window Pro, which I believe has virtual filters named after their real-world Wratten counterparts, and Tiffen has produced Dfx Digital Filter software, which I imagine does something similar). These programs surely include more extensive filters and capabilities, but for the kind of thing John is suggesting, it seems to me like it should be easy enough to just enter the correct value in the raw converter directly. So if you have a "neutral" temperature of 5639, it seems like it would be easy enough to just enter 6639. Of course, you would have to experiment first to see if +1,000 is actually the value that you think best brings the neutral setting up to an accurate representation of a "Cloudy Day". But once you find an increment you like - whether +850, +976 or +1,147 - you should be able to use it for all other pictures that you want to have the same atmosphere.

I think.

I'm sure somebody here will correct me if I have just demonstrated my complete ignorance of this subject...

Best,
Adam

[Ignacio Soler]

Thanks for the post. Very well written and easy to follow. Unfortunately, you need a WhiBal card to get your technically accurate starting poin and they are not available where I live (not anything like it).
I figured an alternative method and it would be great if you could validate it for me. I could photograph an object that I think is neutral grey in colour with direct flash (and no other light source) and the camera set to flash white balance. In ACR I then click with the eye droper on the object and if it gives me the same colour balance as the flash white balance, then the object is neutral grey, a different colour balance means its not. I will try different objects until I find one that works. I understand this will only work if the flash colour balance is set to exactly match the colour of the light emitted by the flash, but you would think that's what manufacturers would do, right?
Thanks for you help,

[Kurt Shoens]

Tom K, I commend LAB to you. I'm in the same boat and learning about LAB has made all the difference to me. I'm much more confident with color now (and my hairline has stopped receding and I have friends now!) and also understand my color deficiencies. I can't color correct as well as the experts, but I can certainly outperform what I used to do.

The great thing about being color blind is that you can ignore things like monitor calibration and ambient viewing light color.

The not so great part is that you have to try stuff and see how normally-sighted people react ...

[Carl Weese]

Adam, I had a hunch about your latest suggestion so I played around with an overcast WhiBal shot. It looks like the correction factor can't be applied in a linear fashion. This shot at neutral was 5850/-8. To get my "stormy light" r/g/b ratio measured on the card took -600 Temp and +3 Tint. But the shot used in the article illustrations, which began with a neutral of 8800/+4, required -1350 Temp and no change in Tint to get the same r/g/b ratio. There may be an arithmetic handle for the non-linearity, but I suspect it's easier to just jiggle the sliders to get the desired r/g/b ratio on the card.