How Can a Gray Card Give White Balance?

TomHeaven

Forum Enthusiast
Messages
310
Reaction score
1
Location
US
How can you set White Balance with a Gray Card? Gray is not white.

Canon's XTi Manual, under Custom Setting White Balance, advises to photograph something white for reference; but footnotes that shooting a gray card will give an even more accurate result.
 
Because a true gray card, I believe will be 17%, which is what the camera is going to try to balance everything in the photo to. So if it knows that this is a true 17% gray card, it will know how the light is affecting things, and balance accordingly. That is my understanding.
--
New to DSLR. Just got my D50, and I'll be a newbie until at least, hmmmmm 2010!
 
As far as your camera's meter is concerned, white is gray, too.

If you take a shot of a white sheet of paper with your camera on auto exposure and a shot of a gray card the same way, they will both look the same. The camera not only expects gray, but it tries to recreate it whenever possible.
--
Jim
 
Both are supposed to be neutral. Grey has the advantage that it is more likely to be truly neutral than white because it has more density. It is very difficult to discern or produce a neutral white. With say a 10% color error from true neutral it will still liook white. With grey you can see the color shift more easily.

Paul
 
And you can get certified neutral gray cards.

--
RDKirk
'TANSTAAFL: The only unbreakable rule in photography.'
 
Gray is not white.
actually white is gray, digital camera or video camera, after you set the WB, no matter it is pure white, light gray, dark gray or even black, the TTL system will expose everything to 18% gray.

WB don’t copy color

most people special video professionals, they think that when they WB on a white paper, the camera copy this color and apply to their image, so that a white paper under a hot yellow light will change back to white just like the white paper, so they never WB on gray and worry their image will change to gray, this is very wrong, WB don’t work like this.

so how does WB works?

WB work with RGB values, in a digital camera or video camera point of view, white is make up of [R=255 G=255 B=255], gray is [R=128 G=128 B=128] and black is [R=0 G=0 B=0] with this in mind, under a bright sunny day, a gray card may have a values of [R=128 G=128 B=128] all values equal but under any warm yellow Incandescent light it may have a values like [R=128 G=128 B=100] the blue values has decreased, so WB system know blue channel has decreased, and will try to balance it back to 3 equal numbers to color correct the yellow cast.

why WB on white is bad?

because WB need rooms to detech color cast, since a channel only has 255 step of grays, when you expose to either side white or black, the top or bottom information has clipped, and the WB system don’t have enough information to correct the cast, so a gray to you is a color, but in a WB system gray is information how much it can correct the color cast, so this is why gray is always better than white.

I have seem alot of video professionals that works in TV stations and wedding, they WB on a piece of paper, most of the time i take closer look at their screen, yellow cast is still there, and sometime WB on white paper will make everything a very light green, but most people don’ care about it.
 
(video pro here)

we white bal on white, or use warming filters (warm-cards) when nessecary.
It's the previous stated reason.

--
==
'My eminence was merely imminent.'

 
why WB on white is bad?
because WB need rooms to detech color cast, since a channel only
has 255 step of grays, when you expose to either side white or
black, the top or bottom information has clipped, and the WB system
don’t have enough information to correct the cast, so a gray to you
is a color, but in a WB system gray is information how much it can
correct the color cast, so this is why gray is always better than
white.
This is a new theory of exposure. If you use your auto-exposure system and photograph a black card, then a grey card, then a white card, so that the card image completely fills the image frame, I wonder what you will see on each image? I wonder if you can tell them apart? I wonder what the RGB values for each of the card images will be?

please explain...
I have seem alot of video professionals that works in TV stations
and wedding, they WB on a piece of paper, most of the time i take
closer look at their screen, yellow cast is still there, and
sometime WB on white paper will make everything a very light green,
but most people don’ care about it.
Humm, is this caused by the paper or is this caused by the lighting system? Is it possible that shooting under gas discharge systems will produce this effect? ... Particularily when multiple gas types are fired on different parts of the AC cycle?

Please explain....
--
tony
http://www.tphoto.ca
 
it doesn't matter if it is grey or white

white balance (really should be called something like COLOUR balance) just makes a certain thing neutral

and if you want to get the right balance for a scene then you need something neutral to take a white balance reading

as the grey cards are officially neutral (and white paper often isn't --- buy a few different types of paper, and you can easily see some are bluer others yellower) they are ideal

--
AJ
http://www.pbase.com/manjade
 
The first problem with using a white card is that there aren't many certified neutral white cards commonly available (perhaps the video industry uses them--I don't know). If the card is not neutral under any lighting condition, then it throws off the white balance. Commonly, white paper is made to look white by brighteners that floresce bluer under electronic flash, for instance.

The second problem, which has been mentioned, is that if any color channel is exposed at or near the response capabilities of the sensor, that will also throw off the white balance. Using a white card, you're already putting all the color channels on the extreme high end of the sensor capability--it's very easy to be just enough over in one channel to throw off the color balance. That's also why we don't use black cards for white balance.

In fact, there is some evidence that even 18 percent gray cards are a bit too dark for the best color balance, which is why WhiBal cards are a bit lighter.
--
RDKirk
'TANSTAAFL: The only unbreakable rule in photography.'
 
In fact, there is some evidence that even 18 percent gray cards are
a bit too dark for the best color balance...
--
Just curious- do you know why that is? I've heard that many times, but never with an explanation. If I habitually "expose to the right" (without clipping any channels,) will it make any difference to the CWB accuracy if the card is black, grey, or white, as long as it is neutral?
 
I wish the links would still work to the old Rob Galbraith forum.We had a good discussion about this and even Thomas Knowl, the co-author of Photoshop and the author of Adobe Camera Raw, participated. (I have the thread bookmarked, but it won't work.)

The software, whether a raw converter or incamera is optimized to look for values in the RGB range of approximately 190~210, which makes the lightest gray on the Macbeth Colorchecker and the WhiBal light gray ideal.

--
Charles Beasley
http://www.charlesphotography.com
 
In fact, there is some evidence that even 18 percent gray cards are
a bit too dark for the best color balance...
--
Just curious- do you know why that is? I've heard that many times,
but never with an explanation. If I habitually "expose to the
right" (without clipping any channels,) will it make any difference
to the CWB accuracy if the card is black, grey, or white, as long
as it is neutral?
See: http://www.bythom.com/graycards.htm
 
The very end of this thread....

http://photo.net/bboard/q-and-a-fetch-msg?msg_id=009dvP

Which has the most plausable historical explaination of the origin of the 18% gray card.

Adams favored 18% because perceptually that reflectance value appear to be halfway between white and black. The zones in Adams system are not f/stops differences in the scene as many erroneously think, but rather the arbitrary division of tones in the PRINT. He picked 10 zones because: 1) its a nice round number, 2) the #2 paper he based the system around could reproduce a 10-stop difference in negative density.

In Adams system if a gray card was included in a "normal" scene with a 10-stop range of brightness it would be reproduced exactly the same tone on the print. Back in those days everyone and their mother didn't make calibrated gray cards, only Kodak did. That's probably why he freaked when Kodak wanted to change the card to ANSI standard (12%).

There's a lot of unnecessary smoke, mirrors and snake oil in Adams system and the 18% gray card is part of that legacy. He was in many ways the Gary Fong of his day :-).

CG
 
... Commonly, white paper is made to look white by brighteners that
floresce bluer under electronic flash, for instance.
My post was obviously making a different point... Oh well ....

Your comment is certainly true. A number of blancophor added to paper (but by no means all) will floresce under UV light. Many papers have little or no blancophors added - among them are some varieties of inkjet photo paper. I have found that Kirkland glossy paper and a Museo matte - which I examined closely a year or two ago - resulted in no problems that my eyes could see on screen or print when shooting flash, tungsten, outdoors, etc.

Ink photo print paper would obviously have a serious viewing issue if it contained inappropriate blancophors. None the less, Agfa used to use blancophors in its Brovira series of glossy silver-print papers.
The second problem, which has been mentioned, is that if any color
channel is exposed at or near the response capabilities of the
sensor, that will also throw off the white balance. Using a white
card, you're already putting all the color channels on the extreme
high end of the sensor capability--it's very easy to be just enough
over in one channel to throw off the color balance. That's also why
we don't use black cards for white balance.
Yes, no matter what you do "near the response capabilities" of the sensor you will find some undesired results. The sensor works well only when used within it's specified limits.

To the best of my knowledge, the Canon and Nikon custom white balance system rely on an image taken (probably with auto-exposure) of a white, grey, or black colour-neutral object. This auto-exposure will ensure that the neutral card image will nominally result in approximately RGB 125 (i.e. 50%) histogram. This exposure is determined by aperture and shutter. The net effect is that the sensor never sees anything other than a 50% exposure (except as created by operator error such as leaving an EV correction in place). In other words, you are never "near the response capabilities".

I would suspect that the determined white balance is only "valid" at the ISO used to generate the WB image - at other ISO, the white balance may slightly vary because of electrical characteristics of amplifiers, etc. However, I have found that a manual WB set at ISO 100 is (by MY eyes) yields identical screen or print images to a manual WB set at ISO 1600. (however, my colour eyes are not too hot).
In fact, there is some evidence that even 18 percent gray cards are
a bit too dark for the best color balance, which is why WhiBal
cards are a bit lighter.
Seems to me that this MIGHT have some validity if you "follow the WhiBal workflow" that also uses the black point and white points. Of course, this might be considered a good idea by some folks.

None the less, it would take special shooting requirements or a less than competent photographer to force the black or white patches "near the response capabilities" with odd-ball exposure. In this case, the WhiBal work flow would not be appropriate.
--
tony
http://www.tphoto.ca
 
There's a lot of unnecessary smoke, mirrors and snake oil in Adams
system and the 18% gray card is part of that legacy. He was in
many ways the Gary Fong of his day :-).
That is quite a silly statement - I hope you smily indicates that you do not believe your own comment.

The zone system was nothing more than a visualization technique to

1. allow the photographers to visualize B&W tones based upon scene luminosity

2. how to ensure that a single developed film plate would be developed to render the tonal separation needed

3. select a paper grade that would result in the full tone range being printed.

This is (almost) a trivial operation - people who "don't know the zone system" usually follow most of these steps by intuition. Adams only codified his experiences and made a few bucks for himself , Minor White, and others by teaching and selling books.

There is no snake oil involved - unless you are looking at the Minor White mistique process and the various workshops that made copious amounts of money teaching the process to willing photographers.

In fact, there is nothing in the zone system that truly requires an 18% grey. If you actually follow the process, one of the things you do is determine the effective ASA of your film - which means you might provide an EV correction to your lightmeter. This makes the 18 / 12 percent reflection immaterial.
--
tony
http://www.tphoto.ca
 
...I thought you were asking for an explaination.
My post was obviously making a different point... Oh well ....
--
RDKirk
'TANSTAAFL: The only unbreakable rule in photography.'
 

Keyboard shortcuts

Back
Top