Doug Kerr
Well-known member
There has been (in another thread here) a recent flurry of discussion about the ColorRight white balance tool and various material presented by its manufacturer.
I thought it might be worthwhile to give a review of the evolution of this tool, and of the literature describing it, as seen from my vantage point.
************
The tool began as the Color Parrot, what might be called a white balance measurement diffuser. The manufacturer (Drew Strickland, proprietor of the ProPhoto Home forums) emphasized that it was well suited for measurement made from the camera position for the shot, although it was also suitable for the less convenient mode, measurement at the subject location.
The manufacturer emphasized that with this tool available, there was no reason to depend on the application of white balance color correction during raw development, which was derided as tedious and no longer necessary. Rather, color correction in camera (using the "custom white balance mode", with the reference frame made with the use of the Color Parrot) would fully meet most photographic needs.
************
I (publicly) asked the manufacturer if he could be so kind as to describe the (evidently new) concept by which the Color Parrot could provide reliable and consistent white balance measurement "from the camera position"; I was unaware of any general concept through which this could be attained. I referenced some matters of photometry and colorimetry that are pertinent to this issue
I was condescendingly told that, while all this "scientific" stuff might be fascinating to some, there was no need to have it clutter up our appreciation of the Color Parrot, which "worked" and was being ordered by users.
************
The manufacturer emphasized the "spectral neutrality" of the tool, and cited "results" showing it to be superior to other tools of similar application in that regard.
It seemed from the nature of the results presented that it was actually "spectral uniformity" that was being spoken of. There are many subtleties in measuring the spectral performance of a transmissive object, so I asked the manufacturer just how the measurement was made.
Imagine my surprise in learning that it had been done with a reflective (not transmissive) spectrophotometer. I commented that such results would be meaningless. Au contraire, assured the manufacturer, pointing out that the transmissive spectrum (in which we are interested) would be just the inverse of the reflective spectrum (measured by the instrument), since "all the light that doesn't pass through is reflected". Evidently, his background in the theory of optical devices didn't include a grasp of absorption.
************
During one exchange of correspondence with the manufacturer, in which I was trying to divine some "principle" upon which the device depended, I pointed out that the properties of a white balance measurement diffuser upon which we ordinarily rely didn't seem involved here. "So", I asked, "what does the diffuser to do for us here?"
"Well Doug", the manufacture replied, "it diffuses".
************
Soon, a new version of the Color Parrot appeared. In this one, all of the diffuser except for a small circular region at the center is blanked with an opaque mask.
The manufacturer emphasized the improved "targeting" of the new model. I asked:
* What do you mean, "targeting"? Are you referring to a narrow acceptance pattern of the diffuser?
* If that's what you mean, why would that be advantageous for white balance measurement? (I could imagine no way in which that would fit beneficially into the process.)
* If that's what you mean, how does reducing the diameter of the active diffuser bring about a "narrower" acceptance pattern? (I could imagine no application of the principles of optics and photometry that would explain that result.)
No answer to any of these came forth.
In any case, tests made here indicated that the use of the mask to reduce the active size of the diffuser had no significant effect on the "breadth" of the diffuser's "acceptance pattern".
************
Recently, a new form of the device (by then known as the ColorRight) appeared: The ColorRight MAX. Here, the opaque annular region surrounding the active diffuser holds a pattern of eight sectors with different reflective colors. This device is particularly intended to be used as a reflective target, placed in the scene (at least a test scene), illuminated by the pertinent incident light.
The shots would be taken with no attempt to perform white balance color correction in-camera, and raw output files would be taken. Then, during "development" of the raw data, an "eyedropper" measurement on a sector of the device would be used to guide white balance color correction on the suite of images.
We note a significant reversal in philosophy by the manufacture here. Only a few months ago, color correction during raw development had been derided as tedious and unnecessary. Now, is is the technique to which the product was born.
It is well recognized that "theoretically ideal" color correction does not always yield the result that, given the tastes of the "audience", the nature of the subject, and the context, is the most artistically pleasing.
Thus often the photographer, perhaps having first invoked a "theoretically ideal" color correction (likely based on "eyedropper" measurement of a chromatically-neutral target in a test frame), will then shift the correction (using control "sliders" in the raw development software) until the desired appearance is achieved.
The ColorRight MAX has a feature that allows this to be done in a different way. Among the eight reflective target sectors are six that are "a little on the bluish side of neutral", with differences in their specific hue and saturation.
By putting the "eyedropper" on one of these alternate sectors, the photographer can arrange for a color correction that will give image tones that are "a bit more red or yellow" than under theoretically-ideal correction. The photographer can presumably try all six until she finds the one whose result seems the best for the particular situation.
This then leads to a manifest paradox in the manufacturer's claims. For this device, he strongly emphasizes the close conformity to chromatic neutrality of the "neutral" sectors of the target (said to be even more accurate than most other devices), and how vital this is to color correction work.
Yet, in the same breath, he points out that attaining the theoretically-ideal color correction (which indeed depends on an accurately "neutral" target) is not ordinarily what the photographer wants, but (for, portrait work, at least) rather an arbitrarily more red or yellow rendering (which we get by measuring on target sectors of arbitrary departure from "neutrality").
************
Well, that's the comtes rendus up to this point. Stay tuned for more amazing revelations.
I thought it might be worthwhile to give a review of the evolution of this tool, and of the literature describing it, as seen from my vantage point.
************
The tool began as the Color Parrot, what might be called a white balance measurement diffuser. The manufacturer (Drew Strickland, proprietor of the ProPhoto Home forums) emphasized that it was well suited for measurement made from the camera position for the shot, although it was also suitable for the less convenient mode, measurement at the subject location.
The manufacturer emphasized that with this tool available, there was no reason to depend on the application of white balance color correction during raw development, which was derided as tedious and no longer necessary. Rather, color correction in camera (using the "custom white balance mode", with the reference frame made with the use of the Color Parrot) would fully meet most photographic needs.
************
I (publicly) asked the manufacturer if he could be so kind as to describe the (evidently new) concept by which the Color Parrot could provide reliable and consistent white balance measurement "from the camera position"; I was unaware of any general concept through which this could be attained. I referenced some matters of photometry and colorimetry that are pertinent to this issue
I was condescendingly told that, while all this "scientific" stuff might be fascinating to some, there was no need to have it clutter up our appreciation of the Color Parrot, which "worked" and was being ordered by users.
************
The manufacturer emphasized the "spectral neutrality" of the tool, and cited "results" showing it to be superior to other tools of similar application in that regard.
It seemed from the nature of the results presented that it was actually "spectral uniformity" that was being spoken of. There are many subtleties in measuring the spectral performance of a transmissive object, so I asked the manufacturer just how the measurement was made.
Imagine my surprise in learning that it had been done with a reflective (not transmissive) spectrophotometer. I commented that such results would be meaningless. Au contraire, assured the manufacturer, pointing out that the transmissive spectrum (in which we are interested) would be just the inverse of the reflective spectrum (measured by the instrument), since "all the light that doesn't pass through is reflected". Evidently, his background in the theory of optical devices didn't include a grasp of absorption.
************
During one exchange of correspondence with the manufacturer, in which I was trying to divine some "principle" upon which the device depended, I pointed out that the properties of a white balance measurement diffuser upon which we ordinarily rely didn't seem involved here. "So", I asked, "what does the diffuser to do for us here?"
"Well Doug", the manufacture replied, "it diffuses".
************
Soon, a new version of the Color Parrot appeared. In this one, all of the diffuser except for a small circular region at the center is blanked with an opaque mask.
The manufacturer emphasized the improved "targeting" of the new model. I asked:
* What do you mean, "targeting"? Are you referring to a narrow acceptance pattern of the diffuser?
* If that's what you mean, why would that be advantageous for white balance measurement? (I could imagine no way in which that would fit beneficially into the process.)
* If that's what you mean, how does reducing the diameter of the active diffuser bring about a "narrower" acceptance pattern? (I could imagine no application of the principles of optics and photometry that would explain that result.)
No answer to any of these came forth.
In any case, tests made here indicated that the use of the mask to reduce the active size of the diffuser had no significant effect on the "breadth" of the diffuser's "acceptance pattern".
************
Recently, a new form of the device (by then known as the ColorRight) appeared: The ColorRight MAX. Here, the opaque annular region surrounding the active diffuser holds a pattern of eight sectors with different reflective colors. This device is particularly intended to be used as a reflective target, placed in the scene (at least a test scene), illuminated by the pertinent incident light.
The shots would be taken with no attempt to perform white balance color correction in-camera, and raw output files would be taken. Then, during "development" of the raw data, an "eyedropper" measurement on a sector of the device would be used to guide white balance color correction on the suite of images.
We note a significant reversal in philosophy by the manufacture here. Only a few months ago, color correction during raw development had been derided as tedious and unnecessary. Now, is is the technique to which the product was born.
It is well recognized that "theoretically ideal" color correction does not always yield the result that, given the tastes of the "audience", the nature of the subject, and the context, is the most artistically pleasing.
Thus often the photographer, perhaps having first invoked a "theoretically ideal" color correction (likely based on "eyedropper" measurement of a chromatically-neutral target in a test frame), will then shift the correction (using control "sliders" in the raw development software) until the desired appearance is achieved.
The ColorRight MAX has a feature that allows this to be done in a different way. Among the eight reflective target sectors are six that are "a little on the bluish side of neutral", with differences in their specific hue and saturation.
By putting the "eyedropper" on one of these alternate sectors, the photographer can arrange for a color correction that will give image tones that are "a bit more red or yellow" than under theoretically-ideal correction. The photographer can presumably try all six until she finds the one whose result seems the best for the particular situation.
This then leads to a manifest paradox in the manufacturer's claims. For this device, he strongly emphasizes the close conformity to chromatic neutrality of the "neutral" sectors of the target (said to be even more accurate than most other devices), and how vital this is to color correction work.
Yet, in the same breath, he points out that attaining the theoretically-ideal color correction (which indeed depends on an accurately "neutral" target) is not ordinarily what the photographer wants, but (for, portrait work, at least) rather an arbitrarily more red or yellow rendering (which we get by measuring on target sectors of arbitrary departure from "neutrality").
************
Well, that's the comtes rendus up to this point. Stay tuned for more amazing revelations.