Moreover, as the relational equations (the first relational equation and the second relational equation), equations represented using a power function may be adopted. In this case, for example, values of variables A and B in the following Equation (7) are obtained on the basis of the above-mentioned combination data (data represented as a plot on the relational graph). Then, an equation which reflects the values of the variables A and B is used as the relational equation. Note that, in a case in which the first relational equation is represented by the following Equation (7), y is a reflectance of the characteristics-acquired gradation value, and x is a reflectance of the maximum gradation value. Moreover, in a case in which the second relational equation is represented by the following Equation (7), y is a reflectance of the prediction target gradation value, and x is a reflectance of the maximum gradation value.
y=Ax^B??(7)
<1.5.4 Fourth Modified Example>

In the above-described first embodiment, the process for predicting the spectral reflectances as spectral characteristics is performed. However, the present invention is not limited to this, and a process for predicting spectral characteristics other than the spectral reflectances may be performed. Examples of the spectral characteristics other than the spectral reflectances include spectral absorption factors (values obtained by subtracting respective spectral reflectances from one) and spectral absorption coefficients α obtained from the following Equation (8). When R₀ is the reflectance of paper white (minimum gradation value) at a certain wavelength, R is the reflectance of the relevant Gradation value, and x is a thickness of the ink, the spectral absorption coefficient α is represented by the following Equation (8) when multiple reflection is not taken into account.