Screen and Profile colours - a wee bit of History
Here is the coloured gradient, constructed in Photoshop
And as I constructed the sRGB gradient and saved it as a psd file, I knew what they should be and I knew there would be a difference. Here is the plot of differences when the file was shown by Photoshop [255 values on the left and 0 on the right].
Quite a difference in the Green and Blue values - but I smelt a rat - the top plot just did not seem right for neutrals. I thought it might be some scale mis-match with pixels and twips. Maybe it is my (Huey calibrated) old CRT monitor playing up - but I got similar results on a laptop LCD screen.
I had constructed the gradient 6 pixels deep (no idea why!), so I tried reading 3 lines from each strip and 9 times out of 10, the values were different (just by a value or two). I double checked using the Photoshop Info readout for RGB and watching the xy position in pixels and they were always the same. So what was happening between the file's values and the screens ones - and why the variation?
I now know what the problem is - Photoshop! It can not create a gradient properly - in fact PS gives different (Info palette) values along the gradients depending on the zoom factor of the image window and the physical position of the gradient on the screen (I had not thought about moving the image to see if I would get different reading!!). It only took me 2 weeks to find this out - makes one love software features/bugs.
I constructed another strip, this time manually intertwining the colours or with each change separated by a black pixel.
So I will use the new strip, but perceptually think of the first one!