In recent years, organic semiconductors have been used to develop a new generation of photodetectors; in some cases their outstanding properties, especially in terms of spectral tuning, have been exploited in order to reproduce human cone sensitivities.To date, however, it is still not clear if the spectral differences between real and artificial cone responses, unavoidable at a certain extent, may lead to real, corresponding differences in the final color perception. As a matter of fact, one should note that perception is the final result of a complex analysis and elaboration made by our visual system at a superior level respect to the color sensation, as detected in the retinal photoreceptors layer. Therefore, aiming at the development of an artificial retina, the way how human perception actually works can not be disregarded.In this paper, we focus in detail on the role and effect of spatial normalization, when applied to a set of tristimulus values obtained using different integration curves, derived by different organic semiconducting materials.In a recent work, we proposed an experimental setup to investigate this issue. We used a multispectral rendering of a virtual scene as a simulation of incoming light spectra, and a set of artificial cone sensitivities to obtain different tristimulus values for each combination of integrating curves. Finally, we applied different computational models with and without spatial color computation to partially simulate human perception. A preliminary analysis of the values showed that the application of a spatial color algorithm leads to a normalization of the differences in artificial cones spectral sensitivities.In this paper we present the results of a new session of experiments, based on the same experimental setup, but using new multispectral test images of real scenes, and a different selection of organic active materials. We analyze the values obtained after the application of the processing methods, trying to define some latex in the selection, among the many available organic semiconductors, of their most effective combination. Moreover, we introduce some hypothesis regarding the effect of different frequency cut points and overlapping areas between the photoresponsivity curves.