I present a spectrometer for a dispersed-fringe optical astrometric interferometer. The two primary requirements, that integration time and systematic error be minimized, do not interact, and both are met. The integration time at the minimum is only a weak function of the parameters, allowing adjustment for secondary criteria such as cross-dispersion resolution without significantly increasing integration time.
M. BisiL. BoninoM. CecconiStefano CesareF. BertinettoG. ManaD. CarolloM. GaiM. G. LattanziEnrico CanutoF. Donati