Here the capital letters represent distances from P whereby O and R stand for original and reproduction colour and Go and Gr are the gamut boundaries for the original and reproduction respectively. Gamut mapped colours are calculated in the above way for a number of hue angles around the originalís hue angle and that colour is chosen as the final reproduction colour, which has the smallest *D colour difference calculated as follows:
 

Note, that (l:c:h) = (3:1:2) is suggested to be the best choice by a previous study of the same authors and this is indeed in-line with the findings of Katoh and Ito (1996). In addition to this method, a second algorithm was also proposed where lightness was compressed via a soft-clipping function and chroma was compressed overall in the way suggested by Viggiano and Wang (1992). This was again followed by finding the gamut mapped colour with the minimum *D which again had the same parameters as above.

These two models were used for mapping between a dye sublimation and an inkjet printer and the resulting reproductions were evaluated using the pair comparison method. The experimental results suggested that the first model using image gamuts performed better than both the first and second models using media gamuts.