History of the Very Odd sRGB Color Space

1996: The color management problem

In the early 1990s, digital images were being produced on all kinds of hardware, using all kinds of software, with and without embedded color space information, that might or might not be interpretable by someone else's hardware and software.

It was only by the wildest of coincidences if the consumer of a digital image saw anything close to what the producer of that digital image actually intended. It was the wild, wild west of digital color management, and the need for a sheriff to move in and establish order was obvious to everyone.

In 1996, Hewlett-Packard and Microsoft took on the role of sheriff. They wrote a proposal entitled A Standard Default Color Space for the Internet - sRGB. The proposal said:

Currently, the ICC has one means of tracking and ensuring that a color is correctly mapped from the input to the output color space. This is done by attaching a profile for the input color space to the image in question. This is appropriate for high end users. [emphasis added] However, there are a broad range of users that do not require this level of flexibility and control.

In 1996, the "high-end users" to whom the proposal referred were people working in the big professional publishing houses. sRGB was aimed at the rest of us.

The solution: A universal color space

Quoting from the Hewlett-Packard/Microsoft proposal:

Hewlett-Packard and Microsoft propose the addition of support for a standard color space, sRGB, within the Microsoft operating systems, HP products, the Internet, and all other interested vendors. . . . Based on a calibrated colorimetric RGB color space well suited to Cathode Ray Tube (CRT) monitors, television, scanners, digital cameras, and printing systems, such a space can be supported with minimum cost to software and hardware vendors [emphasis added].

Minimum cost

In the early 1990s, many of the software and hardware vendors weren't incurring any color management costs at all, because they weren't yet addressing the issue of color management. So adopting sRGB as the universal color space did have its initial cost for some vendors.

But consumers wanted the pictures that they saw on the web or sent by email to look at least somewhat believable. Which meant there was a demand for some kind of color management solution. Supporting one universally used color space was a lot cheaper and easier for the hardware and software vendors, than supporting real color management with multiple color spaces:

. . . if operating system vendors provide support for a standard RGB color space, the input and output device vendors that support this standard color space could easily and confidently communicate color without further color management overhead [emphasis added] in the most common situations.

Minimum overhead

What's this "color management overhead" that the universal adoption of sRGB could avoid? There were actually three different types of overhead that could be avoided — hardware overhead, software overhead, and what might be called "useability" overhead:

Hardware overhead:

Not so long ago, every little bit of extra storage space you could save, and every CPU cycle you could avoid using, really counted.

One of the reasons why Hewlett-Packard and Microsoft suggested sRGB as the universal consumer color space was to save computer storage space by eliminating the need for embedded ICC profiles. Embedded profiles make images larger by a half-kilobyte and up to a megabyte or even more.

Another reason was to eliminate CPU cycles that would otherwise be needed to deal with color management. If every image is in the same color space, there is no reason to waste CPU cycles checking for embedded profiles or converting an image from one color space to another.

To give you an idea of why saving computer storage space and CPU cycles was so important in 1996, here are some personal computer specs for 1991 and 1998, and for comparison, 2012:

Software and useability overhead:

Eliminating color management made writing operating systems easier, because there was no need to color manage the desktop. And anyone who's ever developed editing software will tell you eliminating color management made writing image editing programs easier. The first commercial release of Adobe Photoshop was in 1989, but the first color-managed version of Photoshop wasn't released until 1998. Similarly, Gimp 1.0 was released in 1998, but Gimp wasn't fully color-managed until 2007, with the release of Gimp 2.4.

As far as useability goes, most people who use image editing programs today will tell you that color management isn't particularly easy to understand. If it were, there wouldn't be so many color management tutorials out there on the web and so many questions about color management being asked in user forums.

Personally, I like color management. It's fascinating and can be used creatively. But I have to admit that before all the editing programs started implementing color management, life was easier in the digital darkroom. Ignorance was bliss, and I was blissfully unaware that the colors I saw on my monitor weren't likely to match the colors that would be displayed on anyone else's monitor.

Why sRGB (and not some other color space)?

OK, Hewlett-Packard and Microsoft made a good case for the need for a universal color space to eliminate the vexing problem of color management. But why did they choose sRGB and not some other color space?

sRGB is a very odd color space because it isn't based on anything intrinsically meaninful to digital image editing:

• It isn't based on any rational selection of the most important and most commonly encountered real-world colors.
• It isn't based on the color gamut of color film or color slides, or the color gamut of a scanner profile that might be used with a high quality scanner to transform color film or color slides into digital files.
• It isn't based on the color gamut of a camera input profile that might be used with a decent digital camera.
• It isn't based on the color gamut of a printer profile that might be used with a good printer to make a fine art print.
• Its transfer curve ("gamma" curve, except it isn't really a true gamma curve) isn't based on anything related to human perception of brightness; nor is it based on how film or digital camera sensors respond to light.

No, sRGB is based on — are you ready for this? — sRGB is based on the display characteristics of consumer-grade CRT monitors manufactured in the early 1990s:

• The sRGB chromaticity coordinates come from the chromaticity coordinates of the phosphors that were used to make CRT monitors back in the 1990s.
• Remember the old CRT white point presets that ranged from the blindingly bright but bluish 9300K down to the somewhat dingy and yellow 5000K? sRGB's 6500K (D65) white point was chosen to be a good compromise.
• The sRGB gamma curve is based on a compromise between all the various gamma curves inherent in imaging devices (monitors, televisions) of the 1990s.

It might seem weird at first glance. But making sRGB match the display characteristics of CRT monitors was the cornerstone of the Hewlett-Packard/Microsoft solution to the problem of color management. Picking a universal color space that didn't match the display characteristics of consumer grade monitors would have required using color management to tell these self-same monitors how to display the image.

Oddly enough, the CRT monitor itself was the weakest link in the Hewlett-Packard/Microsoft proposal to eliminate color management overhead. Even back in the heyday of the CRT monitor, simply assuming that your monitor came from the factory calibrated to sRGB and ready to be used for image editing, was likely to be a big mistake. To enjoy the full benefit of having sRGB as the universal color space, the consumer needed to periodically calibrate his or her monitor to match sRGB.

And what about you, today, sitting in front of your LCD monitor? LCDs and CRTs use different display technology. Given how deeply entrenched sRGB is in the assumptions underlying both color-managed and non-color-managed software, it is unfortunate that sRGB is probably not a good monitor profile for your LCD monitor.

And the rest is history — everyone adopted sRGB as the universal color space

In 1996, Hewlett-Packard and Microsoft created sRGB for use as a universal color space. They proposed that everyone else agree to use it. And everyone else did in fact agree to use it. sRGB became the defacto universal, all-purpose, "one space to rule them all" color space for monitors, printers, scanners, and digital cameras; and also for image editing software, desktop computers, browsers, and the entire world wide web.

Who was "everyone"? The Microsoft operating system, all the manufacturers who made consumer-oriented digital cameras, printers, scanners, and monitors; all the imaging software vendors, and all the people who got together and agreed on how the internet should be run. And not to be left out in the cold, all the Linux softwares followed suit, too. (I'm pretty sure Apple was the only major player who didn't get on board — sticking with AppleRGB instead, but not being a Macintosh person, I don't know for sure.)

And there you have it, the history behind the very odd sRGB color space. For more information, read the original Hewlett-Packard/Microsoft proposal, A Standard Default Color Space for the Internet - sRGB. While you are diving into color management history, check out the WayBack Machine's archive of Hewlett-Packard's original sRGB website. Also see Wikipedia's article on sRGB, which is where I found the link to the WayBack Machine archive.