Why We Soft Proof and sRGB v Wide RGB monitor types

[Doug Gray]

I thought I'd quantify the value of soft proofing and also where and how it fails.

The set of saturated RGB colors in any given color space is all RGB triplets having values ranging up to 255 where at least one value is 0. These colors form the boundary of maximally saturated colors. The question then is how well does softproofing work on wide gamut monitors (I used Adobe RGB (1998) as a proxy for them) and regular gamut monitors (sRGB as their proxy).

I explored two questions.

• When working directly in the printer's RGB or in a color space like ProPhoto that is larger everywhere than the printer gamut, how well are the maximally saturated printer colors reproduced during soft proofing with a standard or wide gamut monitor?
• When using either sRGB or Adobe RGB workspace, how inaccurate is NOT using soft proofing?

The first situation clearly benefits the most from soft proofing since Adobe RGB overlaps a much larger part of most printer gamuts than does sRGB. The first attached graph titled "Viewable in Soft Proof ..." show the percentage (Y-axis) of saturated colors that are viewable at or below a delta E values (X-axis).  For instance, 10% of the saturated colors display with more than 5 delta Es from their true color on a wide gamut monitor.  For standard sRGB monitors, 10% of the saturated colors exceed 12 delta Es.


The second situation is something people don't often consider. Much of the colors that one can use are not printable. Even sRGB colors. Soft proofing is particularly valuable to avoid changing image colors in ways that are not printable without being aware this is occurring.

Interestingly, the problem of a mismatch flips in comparison to the first question.  Wide gamut colorspaces and monitors can more easily produce many colors that aren't printable. There differences aren't as stark as in the prior example. For instance, 10% of the sRGB displayed, saturated colors print with more than 9 delta Es.  For Adobe RGB, 10% of the saturated colors exceed 11 delta Es from that printed.  The second attached graph is titled "Viewable v Printable ..."

[digitaldog]

When working directly in the printer's RGB or in a color space like ProPhoto that is larger everywhere than the printer gamut...

You sure about that or am I misunderstanding the sentence? Larger everywhere than the printer gamut?

[Doug Gray]

You sure about that or am I misunderstanding the sentence? Larger everywhere than the printer gamut?

I could have worded it better. That chart shows the difference between what the printer is capable of printing and the soft proofing monitor is capable of showing. Unlike the second chart, it's unrelated to the working space so long as it is at least as large as the printer's gamut which ProPhoto is.