Doug Kerr
Well-known member
The Color Parrot white balance diffuser (now known as the ColorRight) provides a relatively-narrow "acceptance pattern" - the plot of how much "weight" is given light arriving from differetn directions in the overall "mix" presented to the camera for measurement of its chromaticity.
One concept is that a desirable pattern would be the so-called "cosine" pattern, with which the diffuser would essentially mimic the behavior of a subject surface (assuming that the surface were essentially "Lambertian").
The manufacturer of the ColorRight demurs, saying that it is more advantageous (apparently both for measurements taken "at the subject" and those taken "from the camera position") to have a narrower acceptance pattern. And the ColorRight gives just that. We haven't heard just why that should be better - how that is believed to work.
The manufacturer has intimated that in fact the small diameter of the diffuser proper in the ColorRight was chosen just to make the acceptance pattern narrower than it would have been otherwise.
Suppose that one has a Color Parrot (v1.2 or later) or ColorRight and for some silly reason thinks that a broader acceptance pattern would be better. (Perhaps they've heard about cosines, and think they are real.) Could one then "widen" the pattern some by removal of the "masking disk", thus exposing the full available diameter of the diffuser material?
Well, notwithstanding the mention by the manufacturer that the masking of the diffuser proper to a modest diameter was in the interest of attaining a narrow acceptance pattern, we were unable to think of any reason why that might be so. But we are often surprised by such things.
So we in fact ran acceptance pattern tests on our Color Parrot 1.2 both in "stock" condition and with the masking disk removed. There was essentially no significant difference in the acceptance pattern.
Now, there is one "improvement" to be had by removing the masking disk. The unit as provided does not give (for typical lens setups) a uniform illuminance across the entire sensor. We don't know that this has any actual disadvantages - it certainly shouldn't in a Canon EOS camera, where only a central portion of the "pattern" is regarded for chromaticity measurement.
But if, for whatever reason, one would like the entire sensor illuminated by the "sample" light to be measured, one can more closely attain that by removal of the masking disk.
Certainly, we do not advocate any modification of the product without prior consultation with the manufacturer, as doing so might in some way jeopardize the warranty situation.
One concept is that a desirable pattern would be the so-called "cosine" pattern, with which the diffuser would essentially mimic the behavior of a subject surface (assuming that the surface were essentially "Lambertian").
The manufacturer of the ColorRight demurs, saying that it is more advantageous (apparently both for measurements taken "at the subject" and those taken "from the camera position") to have a narrower acceptance pattern. And the ColorRight gives just that. We haven't heard just why that should be better - how that is believed to work.
The manufacturer has intimated that in fact the small diameter of the diffuser proper in the ColorRight was chosen just to make the acceptance pattern narrower than it would have been otherwise.
Suppose that one has a Color Parrot (v1.2 or later) or ColorRight and for some silly reason thinks that a broader acceptance pattern would be better. (Perhaps they've heard about cosines, and think they are real.) Could one then "widen" the pattern some by removal of the "masking disk", thus exposing the full available diameter of the diffuser material?
Well, notwithstanding the mention by the manufacturer that the masking of the diffuser proper to a modest diameter was in the interest of attaining a narrow acceptance pattern, we were unable to think of any reason why that might be so. But we are often surprised by such things.
So we in fact ran acceptance pattern tests on our Color Parrot 1.2 both in "stock" condition and with the masking disk removed. There was essentially no significant difference in the acceptance pattern.
Now, there is one "improvement" to be had by removing the masking disk. The unit as provided does not give (for typical lens setups) a uniform illuminance across the entire sensor. We don't know that this has any actual disadvantages - it certainly shouldn't in a Canon EOS camera, where only a central portion of the "pattern" is regarded for chromaticity measurement.
But if, for whatever reason, one would like the entire sensor illuminated by the "sample" light to be measured, one can more closely attain that by removal of the masking disk.
Certainly, we do not advocate any modification of the product without prior consultation with the manufacturer, as doing so might in some way jeopardize the warranty situation.