How well do the instrument team know the background of the GSO?
It depends on the observation, but again we have background systematic errors within roughly 3% in 50-100 keV band if the data have a few x 10 ks exposure. • The background spectra seem to have huge count rates. Is there something wrong? No. The normalizations of the background spectra were multiplied by 10 for the PIN in order to reduce statistical uncertainties. Note that the GSO background data have been created with the same count rate as the real data. • Could you explain the deadtime correction procedure? Do we do it also on the distributed background data? Deadtime correction of a (spectral) PI file can be performed using the hxddtcor tool in the HEADAS package. The details are available in the newest version of the ABC guide. If users want to correct the deadtime in their HXD light curves, they will need to perform the correction time bin by time bin. They should perform the deadtime correction for GSO background data, but need not do so for the PIN background data, which are g