Idaho Airships, Inc.

In the past several months there has been a marked increase of interest and noise concerning camera profiling. This is a very positive trend-it demonstrates the aggressively expanding engagement of technical matters that only a few years ago were reserved for elite color labs and studios.

The Luminous Landscape recently (9/07) posted an interesting presentation on camera profiling. You should be aware that the product they are featuring is not the only such process available.

Like many other topics on this site, this discussion could rapidly elevate into a discouragingly technical dissertation and do most practitioners absolutely no good. Let’s just cover a few of the cardinal points concerning how camera profiling would be completely frustrated by handheld aerial applications (there…we’ve already presented our conclusion):

1. The typical subject of an aerial portfolio is completely randomized in terms of color temperature, contrast, and other factors profoundly affecting image quality. In a studio, lighting is constant in intensity, direction, and color-not so in the great outdoors. This alone disqualifies the advantages a camera profile, produced under controlled conditions, might produce.
2. A camera is dependent upon the lens for color rendition. I am convinced that image quality, including a variety of color factors, change even within the same zoom lens at varied focal lengths, and certainly in various lens zones. Thus, each prime lens would need to be profiled, and each zoom lens would need to be profiled at varied focal lengths.
3. Aperture will affect color rendition, particularly in the periphery (Transverse Chromatic Aberration, Optical Vignetting).
4. I believe that UV filters, commonly and reasonably used to protect lens elements in field applications, respond to various color temperatures in a variable manner-obviously with their effect restrained to the violet end of spectrum as the Blue and Red Histogram Curves shift (such is the effect of Color Temperature on the spectrum).
5. The manufacturers of camera profiling systems seem to be quite intent on describing the random issues that can be controlled by profiling, but-as is reasonable-they don’t explore the remnant variabilities that can produce considerable profiling problems. For example, when shooting a target of patches during the profiling process, each color patch represents a a particular color performance of an area of the lens and an area of the sensor. It is reasonable to presume that if performance varies from lens to lens and from sensor to sensor, that even individual photosites and their microlenses may have their own character. Contemporary profiling methodologies cannot compensate for this. Additionally, the obliquity of light on sensor sights in the periphery may produce nonstandardized collection results-another issue capable (but not proven, to my knowledge) of producing refractive and even spatially-dependent effects on color, which too cannot be resolved by profiling (since spatial matters will vary by subject and camera position).
6. There is little to no call for absolute color accuracy in aerial photography (such as matching Pantone-prescribed logo colors).
7. The requirement for absolute color accuracy obviates many of Photoshop’s manipulation tools, particularly RGB Composite Curves, to say nothing of mandating supergrade output calibration.
8. Ambient air temperature may significantly affect color performance by influencing electronic noise levels. ‘Nuff said.

This doesn’t mean that camera profiling can’t be used effectively in controlled conditions, such as a studio. It can. It does. And it is certainly something to be adopted in elite production processes.

But it has no real home in handheld aerial photography.

L