To support 10-bit color the following are needed:

  • A monitor supporting it.
  • A GPU supporting it (only AMD FirePro and NVIDIA Quadro support this?).
  • Compatible software. Unless I am mistaken there are very few programs out there supporting 10-bit color. Photoshop is a notable example.

The questions are about how 10-bit monitors perform in comparison with 8-bit monitors:

  • In which situations would a 10-bit monitor give a noticeable advantage over an 8-bit monitor (say, for professional photography)?
  • Have 10-bit monitors been compared against 8-bit monitors based on subjective or objective tests? What were the results?
  • Human eyes can see only 10m colors, so would using a monitor with 1b colors make a difference?
  • 4
    I tried to reword your question a little so it wouldn't fall under the not constructive clause. I think it's an interesting question about niche hardware, so let's see if there are some experts who can share their experience here. – slhck Feb 2 '13 at 16:02
  • Recent 2016+ 27" iMacs support 10-bit color (aka "deep color") throughout, from the GPU to the OS to the screen. All Apple devices as of 2017 also support "wide color", specifically a variant of the DCI-P3 gamut. Source: iMac specs page – STO Nov 24 '17 at 2:41

I think the biggest factor in this is not the high fidelity output, but the possibility to more accurately match a given target color.

Especially when working in print, you want to take care that what you're seeing on screen matches the printed result to a tee. That is much harder if you only have a small amount of colors to chose from. If you have a billion colors, it's much easier to produce a match.

The Need for 10-bit Displays

Conventional display devices use 8-bits per color channel (or 24-bits per pixel) to display images and video. Although this amounts to more than 16 million colors, it still corresponds to a fraction of the colors we perceive in the real-world. This is illustrated in Figure 1, where the green triangle shows the boundaries of the sRGB color space on the CIE-xy chromaticity diagram.

enter image description here

Conventional sRGB compliant 8-bits monitors can only represent the colors that lie in this triangle, while the human eye is capable of perceiving all the colors in the entire chromaticity diagram. This discrepancy is further emphasized by the fact that today’s most professional cameras and printers have a color gamut larger than that of sRGB (such as Adobe RGB shown by the red triangle in Figure 1), creating a bottleneck on the display side.

HP also brings up what they call "banding", an effect that can be seen when very similar colors are displayed close together and become too distinguishable from each other.

The Benefits of 30-bit

It might seem that a 24-bit panel, which offers 16.7 million colors, would be sufficient. For most purposes, that’s true. However, there are cases where 8-bits per sub-pixel is not enough.

Consider a grayscale image. Gray (including white and black) is produced when the three sub-pixels (red, green, and blue) are equally bright. This means that the values for the three sub-pixels are the same: 35/35/35, for example. With 8-bits per sub-pixel, gray can go from 0/0/0 (black) to 255/255/255 (white). Therefore, there are only 256 levels of gray possible.

This can lead to "banding", which is an effect that arises because the step between adjacent levels of gray is big enough for the eye to detect. It can be a problem in certain kinds of visualization, such as 3D Understanding the HP DreamColor LP2480zx’s 30-bit Panel 2 Gray banding (left, exaggerated) is eliminated by the 30-bit panel (right) rendering for automotive styling. With a 30-bit panel, there are 1024 levels of gray, and it’s almost impossible for the eye to detect the step between adjacent levels.


Additional Information

Photoshop can manipulate and display images that use more than 8 bits per color channel. That does not imply direct support for 10 bit per color channel displays.

That was at least the case in 2010.

| improve this answer | |
  • 2
    With regard to 8 bit vs. 10 bit, only your second quote is relevant: the number of bits only affects how "dense" the (itself analogue) colour space is covered by digital representations, not the size of the space. 10-bit sRGB still doesn't have any of the extra colours in it which e.g. Adobe RGB includes. Conversely, you can use 8-bits or less for Adobe RGB or even Lab, the only caveat being that you loose fine-resolution of intermediate steps (which can however be compensated by dithered upsampling). – leftaroundabout Feb 2 '13 at 23:23
  • @leftaroundabout: Thanks. Please feel free to edit my post if you feel that it could be improved :) – Der Hochstapler Feb 3 '13 at 12:04

The basic problem is that the steps between the pixels is fixed while our eyes perceive the ratio. At the bright end of the spectrum the steps are plenty close enough, color #254 blends seamlessly with the #255 next to it and extra bits do you no good.

On the low end, though, while the steps are the same size in light intensity. The gap between #1 and #2 is huge.

| improve this answer | |
  • That's why the gamma-correction exists. Different thing. Bit-depth is about how many steps there are. – DanMan Jan 8 '16 at 5:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.