In this paper, two strategies of spectral printing were implemented and compared: using the Interim Connection Space LabPQR and the Spectral Gamut Mapping Framework (SGMF). The LabPQR approach involved at first, converting spectral information to a lower-dimensional Interim Connection Space and then interpolation using a multi-dimensional lookup table. The SGMF approach involved at first, converting spectral information into CIELAB values for multiple illuminants and then performing direct searches of multiple histogram tables to find matches to these CIELAB values. Additionally, an optimization that used a CIEDE2000-based uniform color space was applied to both approaches. These approaches were tested and compared. The differences between these two methods involved the tradeoffs between interpolation (LabPQR) and indexing/searching of quantized predetermined tables (SGMF). LabP resulted in better colorimetric accuracy under a single illuminant with better spatial image quality. The SGMF approach resulted in slightly better spectral accuracy with a reduction in colorimetric accuracy and noticeable spatial artifacts in low frequency image areas.
Maxim W. Derhak, Roy S. Berns, "Comparing LabPQR and the Spectral Gamut Mapping Framework" in Proc. IS&T 18th Color and Imaging Conf., 2010, pp 206 - 212, https://doi.org/10.2352/CIC.2010.18.1.art00036