Manual White Balance - how to?

[Hening Bettermann]

Hi!

Even though I am happy with AWB, that is without comparison, so I want to investigate. I wonder how to use manual white balance.

1-how do you expose the white card, for zone VII or VI? (I remember that in connection with camera profiling, in some context we are advised to pick WB from the second-whitest patch of the ColorChecker, not the whitest)

2-Which WB do you set for the shot of the White Card? Auto, or the nearest, e.g. cloudy ? (I think of daylight only; and of a WB I would use in post-processing; for a transition, I would like to compare manual to auto WB)

3- A product like the ExpoDisk by ExpoImaging, Inc.
http://www.expoimaging.net/product-overvie...ywords=ExpoDisc
looks like it might be handier to use than a white card. Does anybody have experience with it?

Thank you for your response!

[Panopeeper]

I remember that in connection with camera profiling, in some context we are advised to pick WB from the second-whitest patch of the ColorChecker, not the whitest

The point here is not to use a raw clipped patch, as that would be misleading. ACR does not allow picking WB on a patch containing saturated pixels, but I think most cameras do not care for that (see your Uni-WB template).

Which WB do you set for the shot of the White Card? Auto, or the nearest, e.g. cloudy ?

This plays no role. The shot of the WB card can be used either as a WB template in-camera, or for picking WB on it in raw processing. The WB of that shot is irrelevant in both cases.

It is possible that you misunderstood something in this relation, and that may be the reason for the strange WB coefficients with your neutral WB template. However, this is speculation on my part.

A product like the ExpoDisk by ExpoImaging, Inc. looks like it might be handier to use than a white card. Does anybody have experience with it?

I don't have experience with either; thus only a remark: make a distinction between the color temperature of the source of illumination and of the scenery. For example I live at the ocean, the coast is mountainous with forests. The setting changes the illumination, depending on the direction you are shooting, for the open water and the forested mointain side reflect the original illumination differently. Thus, measuring the color temperature of the sunshine (or cloudy sky) does not show the actual illumination of the surrounding scenery.

[DarkPenguin]

I don't have experience with either; thus only a remark: make a distinction between the color temperature of the source of illumination and of the scenery. For example I live at the ocean, the coast is mountainous with forests. The setting changes the illumination, depending on the direction you are shooting, for the open water and the forested mointain side reflect the original illumination differently. Thus, measuring the color temperature of the sunshine (or cloudy sky) does not show the actual illumination of the surrounding scenery.

Being under a forest canopy is a problem, too.  Where do you point it?

This reminds me, I need to buy a whibal.

[Panopeeper]

Being under a forest canopy is a problem, too.  Where do you point it?

Let's go one step further: is this kind of WBing always useful? I mean, does an object, which reflects all visible light waves equally (i.e. it is "white" or "grey") have to look white in all settings?

When in the Lower Antelope Canyon, in Arizona, I shot a white card for later WB. When back home, developing the raw images, I tried to WB them based on that shot. The result was horrendeous. Although the resulting color matched the sand (we took a tiny amount in a small plastic bag), it did not resemble at all the scenery as I saw it, because the entire canyon inside was not lit by the "original" sunshine but by the reflections on the walls. Ultimately, I WBd based on my memory.

When shooting in a night club or bowling alley or other place, where the illumination is intentionally unnatural, like black light, should one WB so, that a white shirt become white? That would ruin the mood. I admit I have no idea how such a shot should be WBd if not purely subjectively.

[Hening Bettermann]

Yes we are into a theoretical problem here. In principle, I want to maintain the color of the light and not filter/wb it so everything looks like shot at 6 500 K. So when I started shooting digital, I took a purist approach, set the camera WB to Daylight, and did not change that. However, images taken on a summer evening under an overcast sky, looked really TOO blue-green. So I fell back on AWB. It gives satisfactory results - without having a comparison. What I am really after is the  "theoretically correct" color.

Gabor, as to my not-so-uni WB: I don't think this can stem from a misunderstanding from where to pick the WB. I did not use an eyedropper in this connection. The fact that my Kodak White Card is a specimen which has seen better days could not possibly cause such a gross bias??
---
>The point here is not to use a raw clipped patch, as that would be misleading. ACR does not allow picking WB on a patch containing saturated pixels, but I think most cameras do not care for that

So this would indicate to expose the White Card for zone VI if one plans to use it for post-processing.

Good light!

[Jonathan Wienke]

What I am really after is the  "theoretically correct" color.

There is no such thing. You can use a white reference (Color Checker, WhiBal, etc.) to achieve a reasonably close color match between the original subject and a print, but that is of limited aesthetic use. Outside of catalog, documentary, and forensic photography, the color cast of the lighting is usually an important part of the visual appeal of the image. Portraits are commonly "warmed up" (intentionally given a slight yellow-orange color cast relative to "colorimetrically correct") to make the skin tones more attractive. The same is often true of landscapes shot at sunrise or sunset--compensating for the lighting so that whites are R=G=B neutral usually neuters the character of the lighting and makes the image appear flat and dull. Deciding how much to allow the coloration of the lighting to skew whites & grays from neutrality is a judgment call made for creative reasons. If it could be consistently calculated by a formula, camera AWB would have been perfected decades ago.

[Wayne Fox]

Being under a forest canopy is a problem, too.  Where do you point it?

This reminds me, I need to buy a whibal.

I assume you mean an expodisk?

A common mistake with the expodisk is to just slap it on your camera and point it at the scene.  to be used correctly the camera must be facing the same light source that the scene is.  For example, if shooting a portrait sitting or still life, you place the camera where your subject is and point it to where your camera will be.  If shooting a scenic, you normally turn around and shoot it in the opposite direction.

The challenge in a landscape is often where you are standing the light is nothing like the light of your scene.  Of course, this is challenging for both an exposdisk or a whibal chard.

Personally I find using a card much simpler.

[Panopeeper]

[quote name='Wayne Fox' date='Jan 1 2009, 02:56 PM' post='248617']
For example, if shooting a portrait sitting or still life, you place the camera where your subject is and point it to where your camera will be.  If shooting a scenic, you normally turn around and shoot it in the opposite direction

I don't have any expodisk, but my understanding is, that you need to shoot the light source; that may come from the direction of the camera's position, or it may not.

[Guillermo Luijk]

I know what WB is and I know how to adjust it to obtain the desired results. But still there is a question I ask to myself from time to time: what's the reason for needing different white balances depending on the colours present in the scene to get a natural result, if those colours also changed in-place when our eyes where looking at the scenes?

For example, if I develop a landscape with a 'Daylight' WB preset, the result looks good. If now I go indoor, switch a light on and shoot again, why do I need to do a proper tungsten WB to obtain natural colours? why the Daylight preset produces an orange unreallistic appearence on my image? the camera doesn't lie, it just captures light on a tri-band RGB basis. Why the channel alignment (WB) that was fine for the outdoor landscape produces a natural result to my eyes, but need to change to a tungsten channel alignment (WB) to get the same feeling my eyes had being there?

The only answer to me is the way our brain works, performing by itself some kind of WB or temperature compensation. In the tungsten scene the light was really very orange, like the camera captured it, but our visual system involuntarily corrects it to make it appear more 'natural'. If we develop the indoor scene with Daylight white balance and look at it, our visual system will not correct it since our environment prevents us from doing it.

What do you think?

BR

[Hening Bettermann]

I think your description is right. We might add that the WB of our brain does not work 100%. You DO see the sunset (or the rise, if you're up that early;-) ) as red. An empirically based definition of the theoretically correct color would have to measure the distance of the human-brain-WB from physical white throughout the spectrum, and for a lot of people. This is for the landscape photographer, for  whom the color of the light is part of the subject, unlike the catalog photographer, who may achieve his goal with a white card.

Individual differences may be considerable. I would suspect that the difference is less for a trained (naturalistic) painter than for a person whose memory is leveled down by only seeing images which are tailored to cater for this memory.

Good light - of any color!

[Panopeeper]

The only answer to me is the way our brain works, performing by itself some kind of WB or temperature compensation. In the tungsten scene the light was really very orange, like the camera captured it, but our visual system involuntarily corrects it to make it appear more 'natural'

This is a well-known fact, though I would not call it "more natural" but "more like expected".

It is the same issue as with the perspectivic distortion. When standing at the front of a tall building, look upwards: you don't see the narrowing upwards as distortion, for you know that this has to be seen that way (for you have learnt it). However, make a shot of that and look at it on paper: it is horrendeous (except for those photographers, who can not make it better).

[David Sutton]

This is why I enjoy sites like http://web.mit.edu/persci/index.html to see how our perception of WB can break down. David

[joedecker]

The only answer to me is the way our brain works, performing by itself some kind of WB or temperature compensation. In the tungsten scene the light was really very orange, like the camera captured it, but our visual system involuntarily corrects it to make it appear more 'natural'. If we develop the indoor scene with Daylight white balance and look at it, our visual system will not correct it since our environment prevents us from doing it.

What do you think?

I've studied visual perception quite a bit, and this is basically it.  Our brain and eyes do perform a sort of white balance as light changes.   For example, during most of the day, on a typical sunny day, we don't really notice the color of the light change, even though it does (and I'm excluding magic hour here).   The brain's WB is working pretty well here..  but there are limits to how well it works.  I'm under Tungsten lighting here at home, and it's pretty bright, but it's still warm enough that I can see the yellow, my eye doesn't completely white-balance the yellow away.  Similarly with sunsets, as "magic hour" approaches, as the sunset gets yellower, oranger, or redder, the white balance our eyes does is incomplete, and we see the beautiful orange sunset even if it's all we're looking at, maybe not so colorfully as it would be with daylight film, but more colorfully than if our eyes had completely neutralized the colors we were looking at.

That's all pretty messy, what''s important to take away is this:  Using manual white balance tends to produce very subjectively accurate results in "normal" natural light (midday sunlight or cloudy light, shade), but tends to take the ooomph out of some kinds of more interesting light that our brain doesn't adjust to as easily (sunsets, firelight, incandescent lights...), making those situations better suited to white-balance adjustment at post-processing time.

[Hening Bettermann]

Hi Joe

are there any empirical data, which describe the curve, by which our in-brain white balance deviates from true white as color temperature and tint change? If something like that existed, the post-processing white balance could be guided by that and would not need to be purely arbitrary, or based on a memory which is a rubber measure. One could take a white card reading and deviate from it in a defined way.

Curious - Hening.

[Panopeeper]

Hening,

I doubt anything can be gained there. I don't think that the trick of human perception works on a photo. As I see it (literally), the adoptation is relative to the surrounding, not to a piece of paper.

[Guillermo Luijk]

What white balance would be the 'genuine' one? meaning by 'genuine' the adjustment that would provide an image that when displayed on an ideal device (a perfectly calibrated monitor with no gamut limitations for example), would be as close to the real scene (I mean physically measuring with an spectrum meter, not through our eyes) as possible?

Maybe something close to the typical Daylight preset found on any camera?

BR

[Hening Bettermann]

@ Gabor

Hm... Of course the adaptation is relative to the surrounding - but that is: to light. I mean color models and color spaces are based on comparing physical measurements to human vision. Why should it not be possible to make such comparison for WB in changing light?

Furthermore: maybe these data do exist - and we/I just don't know 'em. And maybe the camera maker engineers have such data, and maybe the AWB is built on them. After all, I am so far happy with AWB. But that is without comparison - and I would just like to know what is going on under the hood.

So for now, I settle with AWB, and have learned, that there is no reason to fiddle around with a white card or invest in an ExpoDisk.

@ Guillermo:
> Maybe something close to the typical Daylight preset found on any camera?

I would think so. -

Good light - of any color! - Hening.