Digital color analysis (DCA) of colorimetric and luminescent assays is rapidly spreading in analytical routine. However, the inconsistency of the illumination systems and reading devices, as well as the imperfect choice of the color components for DCA, may result in significant reduction of signal amplitude and in overall response deterioration. An adequate comparison of existing approaches to detection and processing of colorimetric signals in terms of precision of the acquired data is necessary in order to develop an unequivocal criterion for the device and color space selection. We show here that the robustness parameter can serve as such a criterion regardless of the analytical task, the origin of the colorimetric signal/response and the available instrumentation, and propose a comprehensive approach to optimization of the light source, the reading device, and the color space for the analysis of colorimetric assays. Combinations of incandescent lamps with a digital camera as well as red, green and blue LEDs with a monochrome camera were used for signal acquisition. Normalized Red, Green, Blue (RGB) and Hue, Saturation, Value (HSV) color spaces were used for DCA. Each combination of acquisition technique and processing algorithm was tested with a set of chromoionophore-based polymeric optodes, and characterized in terms of robustness.