I think it prudent to again publicly explain my scoring of the Professional Aerial Photographers’ Association, International 2009 Iron Photoshop contest. It was incredibly close and distilled down to some very esoteric color science issues.
The two ultimate finalists were Steven Holtzman, VP of West Coast Aerial Photography in Sherman Oaks, CA, and Allen MacBean of Photographic Solutions in Orem, Utah.
An outline of the contest is available here.
In short, both Mr. Holtzman’s and Mr. MacBean’s preparation of the sunrise aerial (accessible in the above link) were excellent. Mr. Holtzman had neglected some dust spots in the image yet had produced what I felt was a superior visual presentation-but this is a Photoshop contest, and the provided output profile (a Canon iPF5100 using Ilford Smooth Gloss media) was the target.
In order to distinguish which image was prepared most appropriately for the output profile I opened the submitted files (after flattening) in Norman Koren’s Gamutvision application and critiqued them.
Mr. Holtzman used an adjustment of Global Contrast with Levels, producing a very Contrasted and Saturated image. As I’ve mentioned, from a purely aesthetic point of view I liked looking at his preparation a bit better. Unfortunately, the Contrast and Saturation of the prepared image would not survive the conversion to the output profile and resulted in a relatively drastic Δ-E(94) signature.
Mr. MacBeans’s preparation involved the innovative use of an inverted Curve and avoidance of pinching the Histogram Heel. Interestingly, Mr. MacBean moderated his use of Saturative controls to a point where they were barely noticeable, and while this produced a (perhaps) less visually pleasing monitor version of the image, the print-ready version was incredibly compliant with the profile conversion (it converted with a low Δ-E(94) in some very critical areas). Mr. MacBean completely synthesized his blue sky, allowing his moderated Saturative control on the rest of the image, while Mr. Holtzman chose to use a Global Saturative control.
However, the “clincher” was the remapping of Mr. Holtzman’s image in the low key. Mr. MacBean’s image required little remapping in that quadrant due to his moderated Contrast, while Mr. Holtzman’s more drastic Contrast signature mandated some rather extreme remapping-along with associated penalties (I did not feel it necessary to query participants to ensure their use of Black Point Compensation).
Here is a 3D version of Mr. Holtzman’s remapping to the contest output profile:
(Click the thumbnail for a larger image).
Notice the remapping of constituent colors at the top and bottom of the color cluster.
Here is a 3D version of Mr. MacBeans’ remapping to the contest output profile:
(Click the thumbnail for a larger image).
Notice that the high and low key remapping is much less pronounced compared to Mr. Holtzman’s image diagnostic above.
After reviewing Gamut compliance, Softproofing, and other associated image virtue checks, I arrived at the conclusion that in this case, Mr. MacBean’s image preparation for the given output profile was superior.
L