It should be noted that the ISOCAM time estimator (v.2.0) gives the fallacious
impression that there is an apparent gain, in some cases, if a very short
integration time is used. However, the stabilisation time to reach a given
accuracy is essentially independent of Tint, while the noise varies as
. Further, the very short integration time of 0.28
seconds will not be used as often as the others, so calibrations will be
poorer. This very short integration time is not recommended, except where
necessary to avoid saturation of the detector.