In principle, an artificial retina should mimic as much as possible the spectral sensitivities of the real retina.
For technological reasons, building such an artificial device can lead to spectral approximations in comparison
with the real sensitivities. To understand if possible discrepancies can determine big differences in the final
perception, the whole visual system should be taken into consideration, not only the retinal input signal difference.
This paper aims at investigate how retinal sensitivity differences should affect the final perception. However,
answering to this question is a very complex problem related to the whole visual system, that we do not want
to extensively address in this paper. We only want to investigate the relationship between the spatial aspects
of color perception and the spectral differences among cone sensitivities. Moreover, a personal interdifference
has been observed in cone spatial distribution between human subjects, without any corresponding significant
difference in final color sensation. It is likely that spatial compensation, performed by human observers, strongly
decreases this subjectivity in color signal. We aim at address if a similar principle should be considered in
artificial vision. In this paper we analyze the interdifference among integrated values obtained using different
organic-based artificial sensors with different spectral sensitivities. Experiments show a significant decrease of
the effect of spectral sensitivity sensor differences when a spatial color correction is applied.