Stamp Colour And Colour Keys

This grew out of a thread about the "straw" colour in the Canadian Edward VII 7c definitive that moved toward a more general discussion about identifying colour in stamps. Several colour keys have been suggested. The Scott US key, if I have read correctly, refers only to four stamps, so is of limited use. It makes sense that those using actual paint chips are the best. I only have a Gibbons, and it's that type. It's clear that trying to transmit accurate colour digitally is difficult because of variations in software, monitors, etc, and the inherent colour gamut limits of monitors. For example, when I initially posted a question about the "straw" stamp, I included a scan that I prepared with a couple of colour chips for reference. What appears in the list posting, though, is not the same colours even as my final scan - both viewed on my monitor. The bigger problem seems to be a lack of standardization in naming colours. There has been some effort to standardize outside philately, but Gibbons for example states that some colour names remain because that's what collectors are used to. Finally, I'm not a colour expert but I wonder why there is not a move to standardize based on the way colours are identified in image software. I've been playing with some that I have in which, as you mouse over individual pixels, the colour is broken down into it's components. If I put the cursor on the most dense spot of colour I can see on the scan of this stamp, it reads "red, 102, green 56, blue 1." It varies of course and you could do several and average them. Then you can compare those readings with readings from scans (done at the same time) of colour chips. In this case I found that the saturation (I think) in the chips is much higher as the numbers are higher. The low blue reading to me suggests a "browner" colour but . . .? My approach is imperfect for sure but maybe a system could come out of it. And of course, it doesn't work if a stamp is faded, or if your cursor is over postmark ink or dirt. But it may be useful in a comparative way at least. So much to learn . . .

The problem with using pixels to define shades and colors is where do you draw the line? For example, if a shade has 106 pixels of red, is 107 another shade? The difference will not be visible to the naked eye, so it will be of no practical use. Plus every PC is a bit different, so there would be no way to create an accurate number chart without a spectrometer. Which would not change the first problem.
Colors and shades are learned through the experience of seeing a LOT of stamps, and having people with even more experience show them in exhibits and in person. It will never be an exact science because we all see a bit differently, and in many cases there is often money and egos involved.

I've acknowledged that there are variations in computer peripherals etc. My point was that such a system could be used comparatively, to examine two stamps scanned at the same time, or stamps and colour chips scanned at the same time. I agree that experience counts for a lot, but it can't be everything. There still may be people using different colour keys, keys themselves can become faded, reference copies may change, peoples' eyes change over time, etc. Like most things, it doesn't hurt to have several tools in your tool box.

I think you have certainly outlined a potential method of comparing colors of two items scanned under the same conditions, one stamp or color sample the standard for comparison and the stamp to be identified. You may have to compare like regions of the two items, e.g., highlight a solid colored region of the stamp and compare its average color components to those of the standard sample. If you have the inclination it may be worth pursuing. I think it could work. The toughest part might be determining when you have a match. To make this easier you simply compare the stamp to a standard reference set of samples and choose the one that matches closest.

But frankly, I would probably never use such a system, rather trusting to my eyesight instead. That is, I would prefer to create a collection of colored stamps for the country, and period in question, using the most common (least expensive) stamps, and use those as a standard reference for others from the same period.

One consideration of many collectors is the cost factor that their budget allows them for such nice and useful items.

A very nice one would be appreciated but also a cheaper, more available to many collectors, would be of great benefit.