Yes, I'm trying to estimate "native" gamut of the display, the maximum of the colors that the display is capable of. As I found with the default settings it is reduced.Basically the RGBCMY controls are applying an interpolation to map the incoming RGB to the native RGB. So if you are not trying to emulate some other color space in hardware (like sRGB or AdobeRGB), then you should be trying to set the RGBCMY controls to maximum gamut.
Looks like some kind of clipping occurs if saturate too much. For example stepped gradient image from black to blue with 32 steps. Steps that are close to max. of blue become indistinguishable if B-saturation is over 70 (range is 0-100). Also, for example, for the pure blue patch with spotread I noted that when decreasing B-saturation control from 100 to 0 the L (in Lab) or Y (in Yxy) decreases (actually it is about the same from 100 to 70 and from 70 to 0 it decreases). But for example for yellow, when I change saturation from 0 to 100 the L (in Lab) or Y (in Yxy) increases up to about 85 and then from 85 to 100 it decreases slightly. So, my overall impression is that hue/saturation controls of the display are far from ideal. Saturation control affects the lightness and I think that it may slightly affect hue also.Now if these controls have been done well, you won't be able to set them beyond the native gamut of the display. If they've been done badly, they will let you set values beyond the native gamut, and there will be clipping, which would be bad.
Is LCh is a kind of Munsell color coordinate system? In this case it will be very helpful to have LCh in spotread. My idea was by changing hue control setting to stick to a surface of the right hue (which is plane in the Munsell color coordinates), and then by changing saturation (and hue if necessary) trying to stay as close as possible to this surface.One idea for figuring out what's going on would be to setup dispwin -n -m and spotread -e, and adjust each color for maximumchrominance - a^2 + b^2 (Sorry about that, I'll add an LCh display mode to spotread to make such things easier). If you get to a pointwhere you increase the controls and nothing changes, you know it's clipping. If the controls are in xy coordinates, then you could try spotread -e -x, which will show Yxy. A shortcut would be to figure out the best (most saturated) RGB values, then compute the straight line xy values between these for the CMY values, since this will be the case for an additive device.
I see that you suggest "spotread -e". In my experiments I used "spotread -dw". Which one is better?
Thanks for the advices, Graeme. Best regards, Ivan
