In 1992 I started observing regular eclipses of Jovian satellites with the same equipment. The Avia programs turned out to be useless for reduction of these events, mainly because the slightest motion in the recordings introduced vast amounts of the light of Jupiter to be included in the windows, yielding curves with large discontinuities. This called for a new approach, which would automatically correct for any movements.
Figure 4, [Bul93], describes the new approach in relation to the previous one. Determine in advance the fixed X and Y offsets (number of pixels) of the reference object in relation to the object of interest (M). The basic idea is to define a search window aroundthe reference object (R). Simply define the pixel with the highest light level found within this window as the centre of the reference object. This way, knowing the position of R, you know the position of M as well. Now define the new position of a set of small embracing windows to make the actual light measurements. Any light of Jupiter measured together with M will always be the same avoiding any discontinuities in the resulting lightcurves, except for a small incline or decline introduced by the proper motion of R.
This in general can be neglected because the proper motion is
very small during the 15 minute period of the event. Curves
obtained this way are shown in figure 5 and 6.
