When the number of sources in the FoV are larger than a few (3-4), the analysis of SPI data becomes increasingly difficult, especially when sources are separated with angles comparable to the SPI resolution of about 2 degrees. spiros can get confused and attribute the counts due to one source to another. A good test is to add a mock source (or a few) in the middle of the field and to compare the flux extracted by spiros with the input one.
It is possible to derive the contribution of a mock source with the program spi_add_sim. The detector counts expected from a source at given position and intensity are computed using the instrumental response, and added into the binned event data set. This step should be disabled for most standard analysis. The program only works for single energy bin analysis. The user should check the documents provided by the SPI team for further reference [11].