emosaicproc is a task to perform coherently source detection on several more or less displaced exposures from different observations, or "pseudo-exposures" as created by the task emosaic_prep, designed specifically for treating XMM-Newton observations in emosaic mode. A combination of real and "pseudo-exposures" (evtl. from different emosaic observations) is of course also possible. Spectral bands can be defined for maximizing the source detection.
Given the nature of this SAS task, it is possible to run out of memory if a large sky area is covered by the mosaic or the different ODFs. In this case, it is recommended to split the process in several runs.
Due to memory limitations od this task, usersmay find it necessary to work with subsets of different pointings We encorage user to be conservative when using emosaicproc and users have to be aware that RAM memory can be an issue.
When co-adding information for source detection from displaced exposures which cover more or less the same area, several factors act in different way for the images of the same source, eg. vignetting and PSF changes within the FOV have to be taken into account. This is the reason why it is more accurate to treat the data basically pointing by pointing, instead of merging different pointings in the first place.
emosaicproc works on several individual (pseudo-)exposures. The reduced data (event files, as resulting from e[m][p]proc for pointing observations or pre-processed through emosaic_prep for EPIC mosaic observations, previously reduced) has to be prepared in a way, that each (pseudo)-exposure is contained in a separate directory, immediately above the working directory (eg. ../<my-working-dir>/prep_mosaic_001 ; ../<my-working-dir>/prep_mosaic_002 , etc). This is exactly the way data corresponding to different pointings in a mosaicking observation is left, after separating the data in pseudo-exposures by emosaic_prep. To avoid confusion and to keep the procedure simple, the names of the directories containing the (pseudo-)exposures are fixed (eg. "prep_mosaic_<xxx>", where <xxx> between 000 and 999 stands for each position. The numbering does not necessarily reflect any sequence - furthermore any combination of the (pseudo-)exposures can be taken for processing).
