Detector memory effects.

The transient behaviour in-orbit is essentially as described in the ISOCAM Observer's Manual, sections 2.2.6 and 2.3.6.

It is strongly recommended that observers follow the guidelines for choosing Nstab and Nexp given in the Observer's Manual and it is suggested that observers carefully double-check the values they have specified. The observer is reminded that source detection can often be achieved with only a few readouts since 60% of the flux is measured with the first frame. However, very weak sources can hardly be extracted from the background and glitches in less than 10 readouts and photometry strongly depends on further readouts. It is suggested that observations with less than 10 readouts per position will not give better photometry than a factor of about 2 for weak sources, and that a minimum of 20 readouts would be safer. However, it is still too soon to give a quantitative estimate of the photometry as a function of Nexp. The observer is encouraged to note the following points:

• The issue of stabilisation of the detectors is not new and observations don't show different transient behaviour than predicted from ground so the recommendations in the ISOCAM Observer's Manual are still valid. However, contrary to those recommendations, there are many observations submitted during Phase 2 which have a small number of readouts ("frames"). Observers are encouraged to carefully reconsider the number of readouts per sky position when good photometry is required.