2 Observations and data reduction

The observations were carried out with the 2.2-m MPIA telescope at the European Southern Observatory (ESO), La Silla, Chile, on the 19th of July 1993 using a 1024$\times$1024 pixels² Thomson CCD (ESO $\char93$ 19) detector attached to the ESO Faint Object Spectrograph and Camera (EFOSC II). A pixel of this CCD corresponds to 0.332 arcsec on the sky resulting in a field of view of $5.7' \times 5.7'$ at the Cassegrain focus of the telescope. ESO filters 584 (Bessell V) and 618 (Gunn i) were used for the observations. Table 1 gives a log of the observations. Figure 1 shows the identification chart for both cluster and field regions. The field region is located $\sim$ 4$\hbox{$^\prime$}$ south to the cluster and has an overlap of $\sim$ 1.5$\hbox{$^\prime$}$ with the cluster region. During the observations sky conditions were of good photometric quality and the seeing was better than 1$\hbox{$^{\prime\prime}$}$. All observations were carried out when targets were close to zenith.

  

Table 1: Observing log of CCD observations. Both cluster and field regions were observed on 1993 July 19

  

Figure 1: Identification map for both cluster and field regions of NGC 6553. East and north directions are marked. The (X,Y) coordinates are in pixel units and one pixel corresponds to 0.332$\hbox{$^{\prime\prime}$}$ on the sky. Filled circles of six sizes (each corresponding to one magnitude range in V) are used to represent brightness of the stars. Largest and smallest sizes denote stars in the brightest and faintest magnitude range of V = 13 to 14 and 18 to 19 respectively

Sixteen stars from the list of Landolt (1992) were repeatedly observed for the photometric calibration. They cover a sufficient range in brightness (13 < V < 16) as well as in colour (0.2 < (V-I) < 2.5).

Twilight flat fields were used for all flat fielding. All photometric reductions were done using the DAOPHOT profile fitting programme (Stetson 1990, 1992). Further processing and conversion of these raw instrumental magnitudes to the standard photometric system were performed using the procedure outlined by Stetson (1992).

In deriving the colour equation for the CCD system and evaluating the zeropoints for the data frames, nightly values of atmospheric extinction were used. Further details of the data reduction can be found in our earlier paper on NGC 6528 (cf. Richtler et al. 1998), which was observed during the same run. We have therefore used the same colour equations for transforming the CCD instrumental magnitudes to standards ones. For establishing the local standards in both cluster and field regions, we selected about 40 stars on their reference frames, which according to brightness and the absence of close neighbours could serve as "secondary standards'' in the sense of Stetson's calibration procedures. Using these local standards, the CCD profile magnitudes were transformed to the standard system. In this way, cluster and field regions are calibrated independently. The zeropoint accuracy of the colour equation is $\sim$ 0.02 mag in V and $\sim$ 0.01 mag in I. In addition to this, uncertainties in shutter timing and in correction between profile and aperture magnitudes, etc will also contribute to the final accuracy of our photometry. Also the accuracy of relative photometry is largely dominated by crowding and therefore it is a function of the relative position with respect to cluster centre. The internal errors estimated from the DAOPHOT programme are function of brightness. As expected, they increase with decreasing brightness. For brighter stars ($V \le$ 18), it is $\sim$ 0.02 mag but becomes $\ge$0.2 mag near the limiting magnitudes of our observations. So the measurements of faintest magnitude bin in a region are not very reliable.

For all the stars observed in both cluster and field regions, the X and Y pixel coordinates, V and (V-I) photometric magnitudes and their DAOPHOT errors and image parameters ($\chi$ and sharpness) along with number of measurements in V and I are given in Table 2, which is available only in electronic form at the CDS in Strasbourg. We have given only average V and (V-I) values for stars common to both cluster and field regions in the combined data table, as the stars present in the overlapping regions show excellent agreement (within 0.01 - 0.02 mag) between their VI values.