Doug Kerr
Well-known member
Of course, much of today's serous digital camera practice emphasizes the use of the raw output, with white balance color correction during image development perhaps based on a (potentially small) neutral target planted in a representative scene.
Nevertheless, there is still interest in the use of in-camera WB color correction, using such mechanisms as the Custom White Balance feature of Canon EOS cameras. Here, we have numerous opinions as to how the needed measurement of the chromaticity of the incident illumination is best made.
When the illumination involves flash (and, typically, significant ambient illumination from other sources as well), the matter becomes particularly challenging.
Some believe that this can be done by observation of the entire "scene" from the shooting position through a white balance measurement diffuser while a WB measurement (Nikon) or a WB reference frame (Nikon or Canon) is taken. This often produces pretty good results as a result of a fortuitous context; there are no technical principles that suggest this technique should consistently yield the needed data over a wide range of contexts.
I am suspicious of diffusers whose manufacturers claim that their unique technical properties affirmatively overcome this situation, but who am I to say that they aren't useful.
Another technique (not of course always practical, just as for the secreting of a white reference target in an illustrative scene for post-processing use), specifically attractive with the Canon Custom WB process, involves the use of a fairly large neutral reference target ("gray card").
The technique, of course, is to have this card temporarily placed in a "test scene" at the subject position, parallel to the subject surface (or for a non-flat subject, perhaps parallel to the surface we wish to have the most accurate color correction).
Working from the camera position to be used for the shot, we zoom the lens (if necessary) so the gray card fills the "partial metering circle" in the viewfinder and take a WB reference frame (with all metering - flash and ambient - set as it will be for the shot).
(It is perhaps best to "freeze" automatic head zoom of the flash unit to the head position to be used for the actual shot while doing this.)
Of course, we then invoke that frame as the reference for Custom White Balance correction for the ensuing actual shots.
I'm not familiar enough with the details of the Nikon white balance measurement process to know whether or not this is attractive there.
Note that there is little profit in paranoia about precise neutrality of the gray card. Modest departures from "theoretically ideal" color correction are not likely noticed in most situations (and of course we might not even find "ideal" correction desirable at all, wholly another matter).
There are different theories about its reflectance. Especially in a flash plus ambient setting, it is desirable that the two metering aspects work essentially as they will in the actual shot, which might suggest a card with a reflectance that we might think would be about that of the subject.
Commonly-available gray cards have reflectance in the range of perhaps 0.18-0.40, and any of those is as likely to do well as another.
Nevertheless, there is still interest in the use of in-camera WB color correction, using such mechanisms as the Custom White Balance feature of Canon EOS cameras. Here, we have numerous opinions as to how the needed measurement of the chromaticity of the incident illumination is best made.
When the illumination involves flash (and, typically, significant ambient illumination from other sources as well), the matter becomes particularly challenging.
Some believe that this can be done by observation of the entire "scene" from the shooting position through a white balance measurement diffuser while a WB measurement (Nikon) or a WB reference frame (Nikon or Canon) is taken. This often produces pretty good results as a result of a fortuitous context; there are no technical principles that suggest this technique should consistently yield the needed data over a wide range of contexts.
I am suspicious of diffusers whose manufacturers claim that their unique technical properties affirmatively overcome this situation, but who am I to say that they aren't useful.
Another technique (not of course always practical, just as for the secreting of a white reference target in an illustrative scene for post-processing use), specifically attractive with the Canon Custom WB process, involves the use of a fairly large neutral reference target ("gray card").
The technique, of course, is to have this card temporarily placed in a "test scene" at the subject position, parallel to the subject surface (or for a non-flat subject, perhaps parallel to the surface we wish to have the most accurate color correction).
Working from the camera position to be used for the shot, we zoom the lens (if necessary) so the gray card fills the "partial metering circle" in the viewfinder and take a WB reference frame (with all metering - flash and ambient - set as it will be for the shot).
(It is perhaps best to "freeze" automatic head zoom of the flash unit to the head position to be used for the actual shot while doing this.)
Of course, we then invoke that frame as the reference for Custom White Balance correction for the ensuing actual shots.
I'm not familiar enough with the details of the Nikon white balance measurement process to know whether or not this is attractive there.
Note that there is little profit in paranoia about precise neutrality of the gray card. Modest departures from "theoretically ideal" color correction are not likely noticed in most situations (and of course we might not even find "ideal" correction desirable at all, wholly another matter).
There are different theories about its reflectance. Especially in a flash plus ambient setting, it is desirable that the two metering aspects work essentially as they will in the actual shot, which might suggest a card with a reflectance that we might think would be about that of the subject.
Commonly-available gray cards have reflectance in the range of perhaps 0.18-0.40, and any of those is as likely to do well as another.