The colour gamuts of colour reproduction media are important properties of them and can play a decisive role in their use in colour reproduction applications as well as the improvement of their capabilities. While this topic has frequently been studied and is well understood and for output colour imaging media, a solution for input media is not to be had in a simple way. To this end, the present paper proposes a method for obtaining the theoretical gamut across which a device can capture colour differences and this method is based on simulating the responses of an input medium to given spectral power distributions. The gamut of an input medium is then determined on the basis of having a set of spectra that cover the majority of all possible spectra, knowing a medium's responses to them and then determining a boundary beyond which the medium does not produce variation in its responses. The present paper is an abridged version of a more extensive treatment of the topic submitted for journal publication. 1 access to the entire range of inputs to the medium - for output media this means access to the entire range of digital data that can be input to them and for input media one needs access to the entire range of colour stimuli that can be presented to them for capture. Once one has access to the entire range of inputs, the gamut is then determined on the basis of a medium's corresponding outputs. Note also that it is only meaningful to determine a medium's gamut in the space of colour stimuli as in the space of digital data it is always a cube (or hypercube) or some trivial subset of it (e.g. in printers due to setting a maximum total ink amount). Hence calculating the colour gamut of an output medium consists of sending every possible digital input (or a sample thereof) to it, measuring the colour of each corresponding output and then calculating a boundary enclosing these colours in a colour space. The generation of inputs to these media is a trivial matter as one has access to their entirety. Furthermore if the relationship between the medium's inputs and the colours of resulting stimuli (e.g. between RGB and CIELAB for a CRT) is monotonic it is sufficient to sample the extremes of the range of digital inputs (e.g. the faces of the RGB cube for a CRT) as these will correspond to the extremes of the resulting stimuli's colours - the gamut boundary. The boundary enclosing the medium's extremes can then be described using a range of methods and data structures. 3-5 The reason why complexity arises for the gamuts of input media is that sampling the entire range of possible inputs to them means sampling the entire range of possible colour stimuli. This is the case because, to determine the range across which differences in stimuli can be sensed, a set of stimuli with a gamut greater than or equal to the gamut of the given input m edium needs to be available. Practically this also means that the gamuts of input imaging media can only be determined theoretically as having such a set of actual surfaces is extremely difficult to achieve. Once a set of computationally generated samples from the entire possible gamut of stimuli is available, it is necessary to know the medium's responses to each of them and then based on this data to determine the medium's gamut boundaries. The approach suggested in this paper is based on simulation whereby samples will be generated numerically in terms of their spectra and a medium's responses to them will be simulated computationally as well. Hence, the present paper will consist of three principal parts: generation of a set of stimuli for determining the gamut boundaries of input media, modelling of the responses of input media and calculation of gamut boundaries of input media.
[1]
Ján Morovic,et al.
Gamut Mapping Algorithms Based On Psychophysical Experiment
,
1997,
Color Imaging Conference.
[2]
S. J.P..
Characteristic spectra of Munsell colors
,
2002
.
[3]
J. Parkkinen,et al.
Characteristic spectra of Munsell colors
,
1989
.
[4]
Mark D. Fairchild,et al.
Techniques for Gamut Surface Definition and Visualization
,
1997,
Color Imaging Conference.
[5]
Jan Morovic,et al.
To develop a universal gamut mapping algorithm
,
1998
.
[6]
Jan Morovic,et al.
Determining Colour Gamuts of Digital Cameras and Scanners
,
2003
.
[7]
Graham D. Finlayson,et al.
Recovering Device Sensitivities with Quadratic Programming
,
1998,
Color Imaging Conference.
[8]
Ron Gershon,et al.
Measurement and Analysis of Object Reflectance Spectra
,
1994
.
[9]
Graham D. Finlayson,et al.
Metamer Constrained Colour Correction
,
1999,
Color Imaging Conference.