I got in contact with the author of the paper. He confirmed that the increased dynamic range does not increase the gamut size as expected. The increase in gamut size is attributed to "However, we could achieve a higher color gamut by utilizing high-quality Kodak transparent films, compared to the ordinary Gretagmacbeth target. So the gamut size is achieved by colorants." I also queried the results under different illumination and hope to hear back soon. On 21 October 2010 23:14, Graeme Gill <graeme@xxxxxxxxxxxxx> wrote: > Contrast that with the procedure described in the paper. A small number > of their test patches used real world spectra under a realistic illuminant > (48), and the vast majority (576) uses color film (that has only three > colorants, > and therefore produced colors with spectra that have only three degrees of > freedom), with some unnamed backlit table illumination (fluorescent ?). > > My understanding of the paper is that the backlit patches are the only ones used in the calibration process. The reflective Macbeth patches are used only as a test set to confirm the result of the calibration under a different illuminant. Cheers, Jeremy