Multi-primary displays must be able to adhere to current video standards and be compatible with existing content. This implies that the multi-primary display must accept video data in a three-component format, such as RGB or YCbCr. Furthermore, the multi-primary display should present a superior image quality as compared to its RGB counterparts when using that data. Therefore, a real time video processing system is required that will accept a three-component input and convert it to multi-primary format suitable for the display.
This conversion raises several challenges, which are unique to multi-primary displays. First, unlike in the case of multi-primary device, there is no direct relationship between the RGB inputs and physical primaries. Second, the transformation from the three-dimensional RGB or YCbCr input data to more than three primaries signals is an underdetermined problem, and thus does not have a unique solution. The specific solution to be chosen is therefore dependent on other image quality parameters, such as its robustness to noise and physical tolerances. Finally, the shape and the structure of the input RGB color gamut are different from that of the multi-primary display. Color gamut mapping is therefore essential.
In addition to those conceptual challenges, other practical requirements of great importance should also be taken into account. The system must allow predictable color reproduction, which may be achieved by adhering to certain color standards, including wide gamut color standards. On the other hand, when an accurate reproduction of color is not required, the wide color gamut of the multi-primary display must be used to improve image quality, in order to obtain the most pleasing image. Using a flexible design, those two different aspects of color reproduction may coexist, by creating a different setting for each. This calls for complex gamut mapping processing, the algorithmic infrastructure of which must support the type of flexibility discussed. Finally, it must be noted that this gamut mapping and the complicated three to many conversion should be performed in real time at fast video rates.
In order to answer all challenges discussed above, Genoa has developed the ColorPeak(tm) concept. The system is shown below. The video path is composed from several basic building blocks. The first block in the chain accepts the stream of three-dimensional RGB or YCbCr pixels and translates them to absolute color space coordinates. This is followed by a gamut mapping and conversion of the absolute color data into the multi-primary signals. Finally, the multi-primary data enters into a multi-primary device dependent processing block that allow adjustments for each specific display. This structure of the video processing chain allows us to answer the requirements presented above. The processing also involves negative RGB values as required by some of the wide gamut video standards.