Color terminology question

On Feb 19, 1:01 am, "" wrote:

Hi,

Take an RGB color like 192,0,32 where one of the numbers is zero. That color has, in a sense, bottomed out for that space since any further decrease in the R and B values will change the ratios of the original hue and saturation. Similarly there are 64 discrete steps up that it can go, to 256,64,96 where one of the numbers hits a max. What are the terms used to describe the bottom value, the top value and the colors in the middle where the ratios can be perfectly maintained?
Thanks,
Ron

I wanted to add something to this question something I have been wondering about.
Will a spot color that fits inside a small gamut space like sRGB have the same RGB numbers in all wider gamut RGB spaces?

wrote in message

On Feb 19, 1:01 am, "" wrote:

Hi,

Take an RGB color like 192,0,32 where one of the numbers is zero. That color has, in a sense, bottomed out for that space since any further decrease in the R and B values will change the ratios of the original hue and saturation. Similarly there are 64 discrete steps up that it can go, to 256,64,96 where one of the numbers hits a max.

Not quite – initial color has no green at all, so it is a shade of magenta-red. As soon as you add any green to the mix, you will lose saturation. RGB(256,64,96) for example has 64 points of gray added to the original color. The easiest way to see this is to play with the values in the color picker. In gamma 1.0 space, multiplication by a constant will maintain the same hue and saturation at different brightness values. In 2.2 gamma space, you’d need to convert to linear gamma, do the math, and then convert back to 2.2 gamma. Or stay in RGB. Lab and HSB are easier to work in for this sort of calculation. HSB is itself a computational convenience, and not very accurate. sort of like using a compass and ignoring the magnetic declination.

What are the

terms used to describe the bottom value, the top value and the colors in the middle where the ratios can be perfectly maintained?

I would call it the available dynamic range for that particualr hue and saturation.

I wanted to add something to this question something I have been wondering about.
Will a spot color that fits inside a small gamut space like sRGB have the same RGB numbers in all wider gamut RGB spaces?

No, it wouldn’t, even in linear gamma space. In general it would have similar brightness, but it would be less saturated. Once again the color picker is the easiest way to verify this.
—
Mike Russell
www.curvemeister.com/forum/

On Feb 19, 1:38 am, "Mike Russell" <
MOVE> wrote:

Not quite – initial color has no green at all, so it is a shade of magenta-red. As soon as you add any green to the mix, you will lose saturation. RGB(256,64,96) for example has 64 points of gray added to the original color.
Mike Russellwww.curvemeister.com/forum/

Is adding gray, in this case, like adding green? Does going from 192,0,32 to 193,1,33 change the hue since green is being added? It seems like it must. Does going from 193,1,33 to 194,2,34 change the hue again or is that completely unlike going from 192,0,32 to 193,1,33 where there was no green at all.

wrote in message

On Feb 19, 1:38 am, "Mike Russell" <
MOVE> wrote:

Not quite – initial color has no green at all, so it is a shade of magenta-red. As soon as you add any green to the mix, you will lose saturation. RGB(256,64,96) for example has 64 points of gray added to the
original color.
Mike Russellwww.curvemeister.com/forum/

Is adding gray, in this case, like adding green?

Yes – there is a gray component that increases as you add green, and at the same time the hue moves slowly toward green. The numbers for your specific color are so similar though that the only actual change (looking at the color picker) is in the saturation value.

Does going from
192,0,32 to 193,1,33 change the hue since green is being added? It seems like it must.

In the picker, it doesn’t actually change the hue, just the saturation, from 100 to 99.

Does going from 193,1,33 to 194,2,34 change the
hue again or is that completely unlike going from 192,0,32 to 193,1,33 where there was no green at all.

I checked in the color picker, and these are very small differences. Keep in mind, also, that HSB has its limits. It is an ingenious abstraction, based on splitting the color wheel into 8 sections, and then using color number ratios. This is nice because no trig functions are needed, but on the downside HSB can be off by quite a bit for certain small values, particularly near black.

Lab is probably a better way to look at changes in hue angle and saturation, since this is reflected in the a and b channels.

You’re touching on an interesting area here, which is how do you approximate an out of gamut color? Is it better to minimize the distance between your target color and the achievable color in Lab space (minimizing deltaE), keep hue (in Lab space. HSB is not used for precise color calculations) constant, and sacrifice saturation and brightness?

The answers to these questions determine how certain colors are printed – one famous and rather wide spread place where the conventional calculations get the wrong answer is blue skies, and skies go purple. —
Mike Russell
www.curvemeister.com/forum/

wrote:

Take an RGB color like 192,0,32 where one of the numbers is zero. That color has, in a sense, bottomed out for that space

That color is at the edge of the particular gamut. It could still be a valid color or it could be the result of e.g. 192,-1,32 that has been just truncated to 192,0,32.

since any further decrease in the R and B values will change the ratios of the original hue and saturation.

The original hue and saturation is kept unchanged by multiplication even if one or two channels are at zero. Such calculation has to be performed in a linear (gamma 1.0) color space).

Similarly there are 64 discrete steps up that it can go, to 256,64,96 where one of the numbers hits a max. What are the terms used to describethe bottom value, the top value and the colors
in the middle where the ratios can be perfectly maintained?

Well, they are the in-gamut colors that have the same hue as the base color. In this axis (adding white) you are changing both saturation and luminance while keeping the hue unchanged.

Will a spot color that fits inside a small gamut space like sRGB have the same RGB numbers in all wider gamut RGB spaces?

Spot color is a special ink color, it is out of the current CMYK space and/or has some special characteristics like metallic apperance.

Between RGB spaces the whole grayrange (all the RGB codes where R=G=B including "black" and "white") is the same in all RGB spaces but all the other RGB codes represent different color apperance in different color spaces.

Timo Autiokari
http://www.aim-dtp.net

On Feb 19, 9:18 am, Timo Autiokari wrote:

Between RGB spaces the whole grayrange [snip] is the same in all RGB spaces

Timo Autiokarihttp://www.aim-dtp.net

That finally sank in. That’s a pretty important thing to know.

Thanks,
Ron