In this article, we define a generic gradient for color and spectral images, considering a proposed taxonomy of the state of the art. A full-vector gradient, taking into account the sensor's characteristics, is in compliance with the metrological properties of genericity, robustness, and reproducibility. Here, we construct a protocol to compare gradients from different sensors. The comparison is developed by simulating sensors using their spectral characteristics. We develop three experiments using this protocol. The first experiment shows the consistency of results for similar sensors; the second demonstrates the genericity of the approach, adapted to any kind of imaging sensors; and the third focuses on the channel inter-correlation considering sensors such as in the color vision deficiency case.