Gamut Evaluation of an n-Colour Printing Process with the Minimum Number of Measurements Kiran Deshpande, 1 * Phil Green, 2 Michael R Pointer 3 1 London College of Communication, University of the Arts London, London SE1 6SB, United Kingdom 2 Colour and Visual Computing Laboratory, Gjïvik University College, Gjïvik, Norway 3 Colour, Imaging and Design Centre, University of Leeds, Leeds LS2 9JT, United Kingdom Received 6 April 2014; revised 9 July 2014; accepted 20 July 2014 Abstract: The aim of this study was to use a minimum number of measured colour patches to evaluate the col- our gamut of an n-colour printing process. Traditionally, the colour gamut of a printing system has been derived by printing and then measuring a gamut target for exam- ple, a profiling chart. For an n-colour printing (printing with more than four process inks), it is desirable to know the colour gamut of the given set of inks without having to print a large number of test patches. Different spectral printer models were used to predict the gamut of a 7- colour printing process. The colorant space was divided into sectors each containing four inks. For each printer model, the colour gamut of the each four-ink sector was predicted. All sector-gamuts were then combined to pre- dict the overall colour gamut of the n-colour process. This predicted gamut was then compared with the gamut obtained by measurement using a gamut comparison index (GCI). The Yule–Nielsen modified spectral Neuge- bauer (YNSN) model gave the best accuracy, at the cost of a larger number of input measurements, than other models. A combination of the Kubelka–Munk (KM) and YNSN models performed well with the fewest input meas- urements. VC 2014 Wiley Periodicals, Inc. Col Res Appl, 40, 408–415, 2015; Published Online 8 August 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/col.21909 Key words: colour gamuts; n-colour printing; spectral printer models; colour reproduction; printing INTRODUCTION The main purpose of an n-colour printing process is to increase the colour gamut of a traditional four-colour printing process by the use of additional inks. There are many reasons to predict the gamut of an n-colour printing process for example, to select the optimum ink-set for n- colour printing and to compare the n-colour printing gamut with that of the proofing system. An n-colour printing system uses CMYK inks plus additional inks of intermediate hues, for example, orange, green, violet, etc. The increase in the attainable colour gamut however, is at the cost of added complexity, which creates a challenge for generating colour separations for these ink sets. Models for n-colour printing have been proposed in previous studies. 1–4 The gamut of a printing system is typically found by using a numerical model to predict coordinates on the gamut surface. This can be done by printing a set of col- our patches, obtaining their colorimetric co-ordinates from spectral measurements, deriving a model and using this model to calculate the 3D gamut boundary in a uni- form colour space such as CIELAB. Several methods of calculating gamut boundaries have been proposed. 5–8 For printing processes with more than four inks (n-colour printing), the number of colour patches required to deter- mine the gamut boundary using any of these approaches can become excessively complex because the sample combinations required for n-channels at the sampling den- sity d is d n . The gamut boundary can also be predicted from an ana- lytical model using a smaller sample of colours. 9,10 Classi- cal approaches include physical printer models such as that based on Neugebauer and Kubelka–Munk theory. Previous studies have dealt with the analysis and visualization of the *Correspondence to: Kiran Deshpande (e-mail: kiranudeshpande@ gmail.com) VC 2014 Wiley Periodicals, Inc. 408 COLOR research and application