2 Observations and primary treatment of images

  In 1991-92 more than one hundred doubles were observed at Las Campanas observatory (Chile) by P. P.. The evening dates of nights of observations were I. 5 May; II. 17, 18 Sep.; III. 30 Nov., 10 Dec. 1991; IV. 16-18, 20-24 Feb.; V. 15-17, 19 Mar. 1992 (the missions are denoted by roman letters). The mean epoch is 1992.0.

A nitrogen cooled CCD camera was attached to the 61-cm Cassegrain telescope. The chip had the dimensions $512\times512$ with square pixels and low readout noise of 5 electrons. During the observations only the central $258\times258$ pixels window was read.

The camera was equipped with a set of wide band filters which approximately corresponded to the U , B and V -bands of the Johnson photometric system. Some objects were observed in B and V only. Flat fields of the dusk sky or dome in all filters and bias frames were taken each night before observations. For the targets about 3 frames were exposed in V , 4-6 in B and 8-10 in U bands. Exposure duration varied from 3 s in V to 30 s in U . More than half of the objects were observed twice in different nights.

To obtain the absolute photometry data on our objects observations were restricted to photometric nights. The information on extinction was planned to be taken afterwards from Geneva data (see [3, Burki et al. 1995]) obtained at nearby La Silla. To determine the transformations from instrumental to standard magnitudes one night in the end of November 1991 was devoted to standard star observations.

In order to calibrate the astrometric parameters of our objects some astrometric standard binaries with well established positional parameters were observed in the V -filter to give us the scale and orientation of the detector. They were selected from the list of [2, Brosche & Sinachopoulos (1988)] created in the frame of the Hipparcos project. Several stellar traces obtained with the stopped telescope clock drive were also made to determine independently the orientation of chip.

Observational conditions proved to be satisfactory; the seeing was mostly below $1\hbox{$^{\prime\prime}$}-1\hbox{$.\!\!^{\prime\prime}$}5$. The transparency variations were quite low with few exceptions (see below -- Table 2). The ambient temperature changed from one observational session to another in the range from $9^\circ$C to $18^\circ$C.

All in all more than 2600 frames were taken. The frames were reduced (removal of cosmic ray hits, bias subtraction and flat-field division) using the ESO-MIDAS software package. The absolute photometry measurements were made by integration of the total object flux in an $\approx 34\hbox{$^{\prime\prime}$}$round aperture. Differential parameters of double stars ($\rho$, $\theta$,  $\Delta\ {\rm Mag}$) were obtained using the Levenberg-Marquardt least square fit of the object profile [, (Press et al. 1987)]. The two-dimensional Moffat function with an ellipticity in an arbitrary direction was fitted simultaneously to both star images:

For wide astrometric standard measurements we used the DAOPHOT package which was checked to give the same results as the Moffat profile fitting.