High-Accuracy Digital Camera Color Transforms for Wide Gamut Workflows

For digital camera systems, transforming from the native camera RGB signals into an intermediate working space is often required, with common examples involving transformations into ACES or XYZ. For scene-linear camera signals, by far the most common approach utilizes 3times3 matrices (formed using regression methods), which are low-complexity approximations to the exact transformation that would be obtained using a full spectral analysis. For workflows designed for Rec709 displays, matrix-based input transforms are capable of producing reasonable accuracy in this domain. However, the 3times3 matrix colorimetric errors can become significant for saturated colors in workflows involving wide-gamut primary systems such as UHDTV or ACES. To address this shortfall, a novel input color transformation method has been developed that involves separate one-dimensional and two-dimensional operations. From the native camera RGB signals, chromaticity-like coordinates are computed and these are used to index into a two-dimensional lookup table (LUT); the output of the two-dimensional LUT is then scaled according to the input signal. Because the surfaces associated with the 2D LUTs possess many degrees of freedom, highly accurate colorimetric transformations can be achieved. For several cinematic and broadcast cameras tested, this new transformation method consistently shows a modest reduction of mean deltaE errors for colors within the Rec709 primaries. The improvement in accuracy becomes much more significant for saturated colors, for which the mean deltaE errors are reduced by more than a factor of three for colors that lie between Rec709 and Rec2020.