Mac Pro: The mystery of the 10-bit color & jumping mouse phenomenon solved
After more than a year of finger pointing between high-end color monitor developers, Apple, and graphic card and DisplayPort adapter vendors, the problem of jumping cursors on 10-bit-color systems appears to be resolved.
Here, I reveal my past as a color geek. Back in the days, I once worked for a company that sold a color-calibrated monitor in the professional color markets, such as prepress and digital photography. I've retained an interest in advanced color workflows.
At the Luminous Landscape professional photography blog, Pete Myers described some unusual problems connecting a Mac with Eizo Nanao's ColorEdge CG 243w monitor. When using a DisplayPort connection, the monitor will support 10-bit color with a compatible graphics card. And this is a good thing.
Myers used a Mac Pro with the ATI Radeon HD 4870 graphics card, which offers the 10-bit color Displayport interface. DVI connectors can handle 8–bits, while DisplayPort can support higher bit depths. The 10-bit difference is the potential display of 1 billion simultaneous colors over the more familiar 8-bit's 16.7 million colors.
Of course, the Mac provides a Mini-DisplayPort connector, while the LCD monitor has a full DisplayPort connector. This requires a third-party adapter. Then came the problems with the cursor. The mouse jumped around and didn't work properly. But where was the problem?
Myers offers the convoluted tale of the solution. It's a good story.
Thankfully, there's a solution now. Here's his observation of the differences, which he's still incorporating into his workflow.
The most obvious difference shows up during calibration, where it can be readily seen that the Delta-D values for the calibration of the monitor are much more on target, not tending to “skip” in the vicinity of a calibration point.
But the most obvious user observation would be that the lower zones of the image are depicted far more accurately in both color and gradation. While the human eye is capable of resolving around two hundred just noticeable differences of gradation based on human factors perception, those values have to be accurately placed along a non-linear curve to be perceived properly. On a practical basis, 256 steps of gradation just does not seem to do nearly as well as 1,024.
Thanks to Patrick Herold at the Chromix blog for pointing out this post. http://blog.chromix.com/2010/07/10-bit-color-resolution-for-display.html