Any other ideas, folks?
Perennial problem: What is it - precisely - that we see when we see a colourant rendering on a digital colour device?
This has preoccupied people since the start of digitisation. In prepress as it was until 1995 when the International Color Consortium released the ICC Specification, you had a viewable graphic as a colour photograph on reflective paper or as a colour photograph on transmissive film (called a 'chrome').
The photographer handed the 'chrome' to the scanner operator and on the optical lathe of a drum scanner the light was split into an R, a G, and a B channel, converted into electric and then electronic data, corrected in the colour computer, and 'saved' directly onto separated film or into DCS EPS.
If there was disagreement about the colour reproduction, then the photographer had the original viewable graphic, and the scanner operator and press operator had to pick which one was responsible for not reproducing the colours. The idea in the ICC framework is to address the situation that arises in a world without a viewable graphic.
In a world without a viewable graphic, the colour photograph consists of TWO parts, not ONE part. The first part is the digitized colourant data, the Red component, the Green component, and the Blue component in the channel values and the second part consists of the colour definitions in the form of the ICC profile that determines the colours those colourants should form.
When you see digital colour devices that render colour differently, you need to begin by asking yourself what ICC device profiles determine how those colour devices render colour, and what rendering intents determine the conversions between those ICC device profiles. Software publishers should make this simple, since the ICC framework is itself simple enough.
/hh