Picking the right colorspace based on image content
An often overlooked aspect of color-spaces is the ability to use them to affect the overall “look” of the image. This 3D model represents 4 color spaces:
Pro RGB (in red), Adobe1998 (translucent white), sRGB (white wire-frame) and in yellow; a professional giclee printer – the Epson 9900 on Ultra Smooth Fine Art Paper.
The top view shows the saturation boundaries of the colorspaces. The larger the space appears here, the more saturation the color space will support.
The side view shows the brightness levels available in the various color spaces. White is represented at the top and black at the bottom.
Color-spaces with larger hulls allow for greater saturation limits. This means a red with an RGB build of 255, 120, 120 will appear more saturated in ProPhotoRGB than it does in Adobe 1998. Neutral colors will appear identical for hue across the color spaces, though the density (brightness) of those neutrals may differ.
How this affects the look of the image is quite dramatic. A side effect of saturation limits is it’s affect on the visual difference between two neighboring color values. The examples below are screen grabs of the same color build across the three most popular working color-spaces. The left side of the boxes are a build of 255R 255G 126B, and the right sides are 255R 255G 112B
The color variation is barely perceptible in sRGB, but noticeable in the slightly larger Adobe1998 and more so in the much larger ProPhotoRGB. You can also see that as the size of the colorspace increased, the saturation increased.
Images with subtle variations in tone may be adversely affected from the use of a larger colorspace such as ProPhotoRGB, however if an increased separation is what you are looking for, tagging your file as ProPhotoRGB may benefit.
These samples, like the ones above contained all identical Photoshop builds, but were assigned different spaces. As the size of the color space increased you can see that the color separation also increases resulting in a loss of subtlety. This loss means in increase of color noise, and in 8 bit files: a potential for banding. Real 16 bit files (not files converted from 8bit to 16) have a small likelihood of banding as long as they remain in 16bit. However, large portion of professional printing devices will eventually convert your 16bit file to 8bit for printing, and this could result in banding issues. Regardless of bit-depth, saturation will be higher in the larger spaces, so it’s something to be aware of and use to your advantage when needed.
You will notice in the examples that as file is moved into spaces of increasing size, subtleties in the colors can be lost. You can see larger views by clicking on the sample images.
As saturation increases, the the visible difference between neighboring colors increases. One artifact of this is an increase in color noise. This becomes quite apparent when comparing the reflections between the sRGB and the ProPhotoRGB files. Also worth noting is how the “sky” colors in the reflection actually lose saturation with the larger ProPhotoRGB space. This is due to Adobe1998 and sRGB having greater saturation in a significant range of values in this region of color. So if sky saturation is of critical importance in your print, do a bit of testing before you commit to ProPhotoRGB and compensate when possible. Sometimes we get to accept some benefits at the expense of others, and working color-spaces are no exception.
So use your colorspace selection as a tool to further optimize your print results. Be conscious that you aren’t losing or gaining numbers of colors by using a different space, you are merely matching image type to saturation limits and distance between colors. And as always, should you have any questions, reach out in the comment section below!