Elboertie2 wrote:.

How can I find out this value from the predefined white balance setting?.

Chances are very high that it's not one of the "K-named" colors. A white-balance adjustment is one that adjusts for a non-white hue of the ambient light, so a white-balance preset is saying "the ambient light is *this* color.".

There are a bunch of colors with names like "5500K", "2800K", etc.. The reason these colors have these names is really of no use to the photographer, except to note that they have to do with the color put out by the temperature of some glowing objects, and so that's where the phrase "color temperature" comes from. But make no mistake, these are *names* of colors. (They're very specific names that refer unambiguously to their spectral composition, unlike vague names like "red" and "blue", which vary depending on the person, culture, etc.).

These "K-named" colors are technologically convenient because they are specific and can be computed, but chances are very small that you'll run into exactly one of these colors as the ambient light except in very controlled circumstances..

Thus, when doing a white-balance adjustment, you need to be able to specify the color of the ambient light more specifically than from among the "K-named" colors. The Nikon D200 (and likely your D80) measures the relative excess or lack of red and blue, relative to green. This is done without regard to the "K-named" colors, although Adobe Lightroom (for example) displays this color as a "temp/tint" combination..

So, to answer your question, if it's a raw file, ACR or Lightroom can tell you the closest K-named color ("temp") along with a delta from it ("tint"). For JPGs, you can look at the exif data (e.g. with http://regex.info/exif.cgi ) for the "ColorBalance1" item in the MakerNotes. It'll be four numbers like "1.796875 1.40625 1 1", the first two of which are the red/blue (or blue/red, I don't remember off hand) weights. They incorporate not only the white-balance setting, but also the adjustments needed to compensate for the properties of the camera's specific sensor and sensor-covering filters..

This is perhaps more than you wanted in an answer, and less than you hoped for, but there it is..

Jeffrey.

Jeffrey Friedl Kyoto, Japan http://regex.info/blog/..

Hi,.

I'm not quite sure what you mean. Those things are intended to set the WB exactly when the auto won't work or else you distrust it. So it ought to be slightly different for each shot. And very different over the course of filling a card with pictures..

Anyway, if it isn't shown in the camera's menu when being set then it might just be in the EXIF info somewhere. Sorry I can't say more than this..

Regards, David..

There is no visible or simple conversion from "custom white balance" (a sample shot) to "Kelvin white point"..

Internally, the camera uses a third form that is not easy to express, which defines the independent red gamma and blue gamma. When you take a custom white balance sample shot, it does not compute a Kelvin number, it computes the red gamma and blue gamma values. Similarly, when you dial in a Kelvin white point (or any of the other fixed presets), it also computes the red and blue gamma values..

If you have a program which can show histograms in RGB (not just luminance), and you know how to read a histogram, then take a RAW picture of a large neutral subject like a white sheet. The red gamma figure is an attempt to move the big red peak to match green, and the blue gamma figure is an attempt to move the big blue peak to match green. If all three peaks match, the color in the image will appear entirely unsaturated gray, by definition..

[ e d @ h a l l e yc c ] http://www.halley.cc/pix/..

Jfriedl wrote:.

This is perhaps more than you wanted in an answer, and less than youhoped for, but there it is..

This is much more than I asked for or expected, but exactly what I needed..

Thank you very much for sharing your knowledge Jeffrey, I have now a much better understanding of this..

RegardsJacques.

'A camera is a device that helps one appreciate the world without it.' - Jacques.

Image control:Zoom outZoom 100%Zoom inExpand AllOpen in new window.

Brian wrote:.

I hate to admit it, but your signature looks better in colour..

View DPR with wider columns: http://friend.smugmug.com/gallery/3106909...

Your explanation of the colour histogram peaks helped to make it more clear to me..

Thank you for sharing your knowledge..

Jacques.

'A camera is a device that helps one appreciate the world without it.' - Jacques.

Image control:Zoom outZoom 100%Zoom inExpand AllOpen in new window.

Brian wrote:.

I hate to admit it, but your signature looks better in colour..

View DPR with wider columns: http://friend.smugmug.com/gallery/3106909...

Ed is right. Colour temperature (measured in K) is all about correcting the relative strengths of the red and blue channels, in order to compensate for emission characteristics of the light source..

The earlier poster who said that Kelvin defines specific colours and that there may not be a Kelvin value for your white balance is not quite correct..

The Kelvin scale is a scale and it pretty accurately represents the emission characteristics of light sources that produce light through being heated. Such as candle flames, conventional incandescent lightbulbs, the sun, that sort of thing..

However, things like fluorescent strip lights do not produce light by heating things and do not behave in the same way. Fluorescent lights tend to emit a lot more light around the green part of the spectrum than other light sources, so balancing the red and blue using the Kelvin slider leaves a very strong green cast. The second value adjusts the gamma of the green channel, to compensate for these spikes..

This is why, although the D80 lets you choose a colour temperature, the manual warns you to use the flash or fluorescent setting for shooting under certain types of artificial lights (because manually selecting a colour temperature won't compensate for the green, but those presets will). This is also why there is a separate 'flash' preset even though, in Kelvin terms, it's essentially the same as the 'direct sunshine' setting (page 140 of the manual)..

In terms of how you can tell what a manually set 'Preset' value is, if it isn't shown on the second information menu, in playback mode, under WB, then I don't think there's any way of you being able to tell, in-camera. That said, I'm not sure how useful that information would be. You can find out, after the event, because many RAW converters will tell you what settings the camera applied. (Lightroom uses the camera's setting as it's starting point, unless you load a JPEG, in which case it assumes 5000K - ie that a white balance has already been applied)..

Does that help?..

The earlier poster who said that Kelvin defines specific colours andthat there may not be a Kelvin value for your white balance is notquite correct..

Actually, he is, and I'll second what he said..

(When counting votes, please ignore the fact that I am him and he is me; give our word the weight of three: me, myself, and I .

The Kelvin scale is a scale and it pretty accurately represents theemission characteristics of light sources that produce light throughbeing heated..

Kelvin is a scale of temperature, like Fahrenheit and Celsius. Human body temperature is about 310K. Water freezes at 273.15K Cast Iron melts at about 1,643K. Carbon (as in a diamond) melts at about 3,800K..

Theoretical physics can calculate the chromaticity and intensity of light given off by something heated to a specific temperature. "Chromaticity" refers to the relative balance among the wavelengths of the visible spectrum. One way to look at it as being "color without regard to brightness," and as such, pure gray and pure white are identical (because they both have an even balance of all the visible wavelengths). Orange and brown are also the same, differing only in brightness..

So, theoretical physics can calculate exactly the light given off by something heated. If you heat a diamond to just short of melting it... say, 3,600 degrees Kelvin, it will glow with a chromaticity that can be exactly computed, and that chromaticity is called "3600K". It's not a color until a specific brightness of it is perceived or sensed by something within some context, but it's fair to call it a "color" in casual conversation..

Here's an illustration:.

Image control:Zoom outZoom 100%Zoom inExpand AllOpen in new window.

(ignore the part that says "mouseover" this single image is from a set for a writeup I'm working on).

Anyway, it's rare that this kind of light is seen in nature because most objects that might be heated also reflect light, or transmit light. Also, that light then mingles with any other light (such as that light itself reflecting off of things, like the ground, trees, your shirt, etc.)..

I'm not an expert in theoretical physics, but I don't believe that light from the sun at all falls into this category, since that comes from atomic reactions and not heating..

In any case, even if it were, by the time the light gets to us here on earth, it's had to pass through all kinds of things that change it's nature. A lot of the blue wavelengths get scattered as it passes through the atmosphere.... that scattered blue then gets reflected and refracted through thousands of square miles of air until it finally finds it's way to your eyes as a "blue sky"..

More "direct" sunlight is what's left over, and as such is missing a fair amount of it's original blue. However, when you combine back in the blue from the hundreds of square miles of atmosphere visible from earth, and combine in all the many different colors of light reflected from things (the green reflected from the grass, etc.), you often end up with something fairly close to "whiteish" at midday..

Coincidentally, something heated to glow at about 6,000 degrees K (putting out light we call "6000K") is also fairly whiteish. The way each got to their own whiteish are totally unrelated, but since they're often sort of close, it's convenient to call direct sunlight "6000K" (or 5500K, or 6500K, depending on whom you ask)..

The closest on earth you'll get to true "K" light is that from an incandescent light bulb in a totally color-neutral (black/gray/white) room. In reality, the chromaticity of light in real-world circumstances is all over the map. Sometimes it all adds up to something close to one of the K-named colors, and sometimes it's not even close..

You might be wondering, then, why even bother with the K-named colors, at least with respect to photography. I wonder that myself, but I suspect it has to do with the convenience of being able to calculate the chromaticity easily and exactly. Perhaps someone found it convenient in the early history of film, and the practice stuck. I don't know. Frankly, it's only mildly useful in real life..

However, things like fluorescent strip lights do not produce light by.

The issue with fluorescent lights is that their spectrum has a lot of holes or spikes in it. There might be a spike at, say, the 575nm wavelength (yellow/orangeish) that is much stronger than the other wavelengths. This spike would have no effect on something that doesn't happen to reflect that wavelength (something blue, green, or red, for example), but a huge effect on something that does. Thus, it makes it all the harder to pick one white-balance adjustment for the whole scene when the light is so "bursty" in ways that effect each color differently..

(The same problems happen when a scene has mixed lighting. Consider someone whose face is half in the sunlight and half in the shade. The color of those lights are usually *vastly* different, so if you adjust for the color of one, the other is left looking odd. Hopefully in the future there will be raw converters which allow us to "paint" areas with a white-balance correction, so that we can correct each part of the scene as appropriate.).

Hope this helps...Jeffrey.

Jeffrey Friedl Kyoto, Japan http://regex.info/blog/..

It's not a color until a specific brightness of it is perceived or sensed by something >within some context, but it's fair to call it a "color" in casual conversation.

I wasn't entirely disagreeing with you, I merely intended to suggest that your answer wasn't quite clear or, at least, that the way you simplified it might be slightly misleading. I wasn't questioning that you knew the answer, more that the way you expressed it might not be as clear as you intended..

And yes, although the sun isn't a black body radiator, per se, it's output is similar enough that you can compensate for it as if it were, which is the important thing...

Marwood wrote:.

I merely intended to suggest that your answer wasn't quite clear or,at least, that the way you simplified it might be slightly misleading..

That's always a danger in technical writing, a bad balance between snowing a reader under, and being correct/complete. It's good to be called on things from time to time .

I have a lot of experience writing about technical subjects, but I still haven't been able to develop a really good presentation on white balance. I spent much of last summer researching ways to present things (including using my D200 to take thousands of white-balance presets in a wide variety of lighting situations, then extracting the color information just as the OP was asking about, and plotting the results on a graph along with all the K-numbered colors). I then spent the rest of the intervening year writing up presentations, but never feeling that it was ever good enough to publish, so it waits..

The biggest problem, I think, is figuring out the proper order to present things in. One really needs to present the basics first ("chromaticity" vs. "color", "brightness" vs. "intensity", how lights mix to form overall illumination, etc.) but presenting those first makes for a better sleep aid than a writeup on white balance..

And yes, although the sun isn't a black body radiator, per se, itsoutput is similar enough that you can compensate for it as if itwere, which is the important thing..

To my thinking, the whole talk about "black body radiators" and such just gets in the way, other than as a side discussion of why such odd nomenclature is used. You could have an innate understanding of how to deal with white-balance issues in the real world without ever knowing what a "black body radiator" is..

Anyway, until I figure out a good presentation, I'll just keep practicing in threads like this. The back and forth discussion is good for all.....

Jeffrey.

Jeffrey Friedl Kyoto, Japan http://regex.info/blog/..

Why not go back to Kelvin's idea and bring it up-to-date? I mean most people can see a candle flame as low temperature and absolute temperature isn't all that complex to explain....

BTW, we left out "Mirads" or am I hopelessly old fashioned?.

Regards, David..

David Hughes wrote:.

Why not go back to Kelvin's idea and bring it up-to-date? I mean mostpeople can see a candle flame as low temperature and absolutetemperature isn't all that complex to explain....

Sure, but what does that have to do with photography?.

BTW, we left out "Mirads" or am I hopelessly old fashioned?.

Well, it's mentioned on the image embedded a few posts back, so it's not totally forgotten (although Mirad colors are as relevant or unrelevant as the Kelvin colors, to photography)..

Jeffrey.

Jeffrey Friedl Kyoto, Japan http://regex.info/blog/..

I've deleted what I had written, having mis-read a previous post...

Jfriedl wrote:.

David Hughes wrote:.

Why not go back to Kelvin's idea and bring it up-to-date? I mean mostpeople can see a candle flame as low temperature and absolutetemperature isn't all that complex to explain....

Sure, but what does that have to do with photography?.

It links temperature to colour in a way people can see/imagine..

BTW, we left out "Mirads" or am I hopelessly old fashioned?.

Well, it's mentioned on the image embedded a few posts back, so it'snot totally forgotten (although Mirad colors are as relevant orunrelevant as the Kelvin colors, to photography)..

So you don't see a link between white balance and photography....

Regards, David..

To my thinking, the whole talk about "black body radiators" and suchjust gets in the way, other than as a side discussion of why such oddnomenclature is used..

I couldn't agree more. I've been, at a much simpler level, putting together an article about White Balance myself. I hope you don't mind, but I've emailed it through to you. As you say, sometimes discussing these things can provide some really useful input. If you have time, I'd really appreciate any feedback you have..

It's aimed at digital photographers who probably don't know or care what a black body radiator is, but would like to know whether it's worth them buying a WhiBal want to understand how (and when), to preset their white balance...

This was an excellent post - I will probably print it and save it with my other photographic reference materials. One minor point - I was under the impression that the term in this temperature scale is just "Kelvin", not "degrees Kelvin". But my memory is fading rapidly, and I'm probably wrong about this..

Congratulations again to you and several other posters for clarifying a rather complex subject.Art Caputi..

It's degrees Kelvin or degrees above absolute zero or Centigrade plus 273 in my little world but I'm begining to have doubts....

Regards, David..

Jfriedl wrote:.

For JPGs, you can look at the exif data (e.g. withhttp://regex.info/exif.cgi ) for the "ColorBalance1" item in theMakerNotes. It'll be four numbers like "1.796875 1.40625 1 1", thefirst two of which are the red/blue (or blue/red, I don't rememberoff hand) weights. They incorporate not only the white-balancesetting, but also the adjustments needed to compensate for theproperties of the camera's specific sensor and sensor-coveringfilters..

Here is some EXIF information (4 seperate lines from the Opanda viewer) from a JPG shot that I took in a bright sunny day, and also measured (Expodisc) in a bright sunny day:.

White Balance Bias Value = 0White Balance Red, Blue Coefficients = 575/256, 279/256, 256/256, 256/256000D = 00, 01, 06, 00Exposure Diff = 00, 01, 0C, 00.

The second line is what I suspect you are referring to regarding balancing of the colours. I do not, however, understand this '256' value system as the first two values are 575 and 279 out of 256, thus exceeding it..

Does this mean that the first value of 575/256 stands for the colour red, and that the intensity is more than twice the expected maximum intensity?.

Jacques.

'A camera is a device that helps one appreciate the world without it.' - Jacques.

Brian wrote:.

I hate to admit it, but your signature looks better in colour..

View DPR with wider columns: http://friend.smugmug.com/gallery/3106909...

Does this mean that the first value of 575/256 stands for the colourred, and that the intensity is more than twice the expected maximumintensity?.

They're adjustment coefficients, used to adjust for various things, including white balance. They also adjust for any unevenness in the sensor's sensitivity across the visible spectrum, as well as unevenness in the transmission properties of the filters across the visible spectrum..

They also help adjust for the fact that there's only one photosite for each of the blue/green channel, but two for the green channel, in each pixel..

You really can't read anything into them as absolute numbers (it's not meaningful to compare them against those from other makes or models of cameras). Whatever the numbers, they make sense only relative to themselves..

For your info, appended is a (HUGE) graph of their reciprocals, from thousands of tests in various lighting conditiosn with my D200, along with the "K-named" colors and the named whitebalance presets ("flash", etc.)..

Jeffrey.

Image control:Zoom outZoom 100%Zoom inExpand AllOpen in new window.

Jeffrey Friedl Kyoto, Japan http://regex.info/blog/..

Mammia mia Jeffrey, you know a lot!.

It will take me some time to come to grips with all this information and truly understand, but I am sure that this will answer a lot, and it already has..

That is also quite an impressive chart you've got there. Interesting how the results ends up us a diagonal line between top values of green and blue. This chart must have taken you a very long time to make, or?.

Tell me, did you study light or some science, because you know so much about this subject?.

Thanks againJacques.

'A camera is a device that helps one appreciate the world without it.' - Jacques.

Brian wrote:.

I hate to admit it, but your signature looks better in colour..

View DPR with wider columns: http://friend.smugmug.com/gallery/3106909...

Elboertie2 wrote:.

This chart must have taken you a very long time to make, or?.

The hard part was taking all the measurements, and then understanding how to plot it. David Coffin (of dcraw fame) helped me quite a bit in the understanding part..

Tell me, did you study light or some science, because you know somuch about this subject?.

I didn't know a thing about white balance or any of this 18 months ago, but did a little reading here and there. Much of it is just common sense, once you get a few basics out of the way..

The problem is that no one really knows how to present it well. I'm still trying to figure out a way to do so myself..

Jeffrey.

Jeffrey Friedl Kyoto, Japan http://regex.info/blog/..