For each of MOS and pn, it must be decided if the source is bright enough to give rise to photon pile-up that would degrade the calibration beyond the science goals when using the full frame imaging mode (see UHB section on EPIC pile-up ). In the case of a bright point source, pile-up is likely to be a concern. Then the user should choose the partial window mode with the largest FOV that minimises pile-up or, for the very brightest sources, use the timing mode.
In order to assess possible pile-up in individual emission lines in RGS data of bright targets, SciSim should be used for modeling the source spectrum; in case photon pile-up is not a problem, the RGS should (for standard spectroscopy) be left in the default SPECTROSCOPY mode. If high time resolution is required, the HIGH TIME RESOLUTION mode should be chosen (see UHB section on RGS modes ).
For a bright point source the user might consider using the OM's fast mode, if high time-resolution photometry is required. For each of the science instruments as prime there is one fast mode default configuration. The user should consider whether the time resolution of OM fast mode data is important for the observations and choose the time slice duration accordingly. In case of OM grism observations with a fast mode window on the zeroth order image of the target, OM should be declared prime.
If the observation is long compared with the visibility window (as reported by the online XMM Target Visibility Tool), the user must consider how best to split it up, e.g., into multiple observations that fit into continuous visibility periods. Other criteria for the split of an observation could be optimal filling of gaps between CCDs, or a calculation to fit in a certain sequence of OM filters observed with one of the OM default configurations (i.e., with exposure multiples of five per filter).
Users must check the visible magnitude of in-field or nearby optical sources and the science target itself. If the soft X-ray response is important, one should choose the thinnest filter compatible with the brightest visible objects, as described in the UHB section on EPIC filters .
In case of SPECTROSCOPY mode observations of a target with particularly strong emission lines, observers might want to read out individual CCDs more often than others (see § 5.3.3.4). In case of HIGH TIME RESOLUTION mode observations it must be decided whether one or all nine CCDs shall be read out.
For the OM it must be decided if specific filter coverage is necessary for the science or if the recommended filter sequence is adequate (see UHB section on OM default configurations ).
Telemetry and onboard memory limits place upper and lower boundaries on the duration of single OM exposures as listed in § 5.3.3.5.