ANALYSE/ROMAFOT

ANALYSE/ROMAFOT frame [cat_tab] [int_tab] [sigma,sat]

This command is the most interactive part in the reduction procedure. Because of its flexibility, its use may not seem straightforward. Therefore, after starting up the command, the user is advised to press the key ``H'': it will display the help information the screen. For the sake of clarity only those commands which are useful at this stage will be discussed below; other commands of the ANALYSE/ROMAFOT will be discussed later.

After having received the appropriate input from the MIDAS command line, the command waits for a command input. Following the present example, the user should type ``D'' and a number. The command then asks for a number, say n, and will read the data for the $n^{\rm th}$ group of components stored in the intermediate table. ANALYSE/ROMAFOT will display the data profiles and the fit superimposed on the upper part of the image display. On the lower part of the screen, three images are shown: the subframe containing the real data, the subframe containing the fit and the difference of the two, respectively.

The user can squeeze the interval over which the 256 colours are distributed through the command ``W'' or vary the minimum of such an interval with the command ``U''. In order to have a better check of the correspondence between the data and the fit, the command ``X'' displays isophotes at three different levels. The key ``Y'' has the same function but starts at a level given by the user (typical numbers are 0.01 to 0.001). Another command useful for display is ``6'' which allows the smoothing of the subframe containing the real data. This smoothing is performed via a $3 \times 3$ gaussian convolution mask on the array stored in memory and does not affect the original array.

By means of the displayed information one can decide if the fit is adequate or for example systematically wrong (for instance, all the images have overestimated $\sigma$'s or heights). In the latter case FIT/ROMAFOT can be run with different parameters. The procedure continues with ``D'', n+1, n+2, ...... etc.

A mean of all fits obtained will finally give the $\sigma$ of the PSF. Since $\beta$ had been already fixed, the PSF is now determined.