The data were obtained with the 2.2 m MPIA telescope at the European Southern Observatory (ESO) at La Silla, Chile. We carried out our observations during two nights, the 18th and 19th of July, 1993. We used the ESO Faint Object Spectrograph and Camera (EFOSCII) equipped with a 1 K Thomson CCD (ESO #19). The pixel scale of this CCD is resulting in a field of view of . ESO filters 584 and 618 were used, which are Bessell V and Gunn i. Table 2 (click here) gives a log of the observations. Figure 1 (click here) shows the identification chart for the imaged cluster region. The seeing was slightly below 1'' on the short exposure frames and about to arcsec in all of the 3 min exposures.
|Filter||exp.time||no. of frames||date|
|V||60 s||1||18 July 93|
|V||180 s||6||18 July 93|
|I||30 s||1||18 July 93|
|I||180 s||5||18 July 93|
|I||30 s||2||19 July 93|
|V||30 s||2||19 July 93|
|I||216 s||1||19 July 93|
|I||180 s||3||19 July 93|
|V||180 s||4||19 July 93|
Figure 1: A finding chart for our field centered on NGC 6528. Seven different symbol sizes are used to represent the magnitude range from V=13 to 19 mag. North is up, west is right. Identification of stars is feasible via a short list of the brightest stars (Table 3 (click here)). The field of view is . Some stars have large errors due to crowding or location at the frame edge. We left them in to facilitate the identification
In addition, 16 stars from the list of Landolt (1992) selected by M.\ Bessell for telescopes with small fields of view were repeatedly observed. Although our filters differ from the filters used by Landolt, we could reproduce the Landolt system quite well, at least below the limits which are implied by the photometric quality. Although the observations in the first night were obtained in clear conditions, the photometric quality was not as good as in the second night. Therefore we relied on the second night alone for the photometric transformation.
Twilight flat fields were used for all flat fielding. All photometric reductions were done using the DAOPHOT II profile fitting software (Stetson 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
zero-points for the data frames, nightly values of atmospheric extinction
were used, which were simultaneously calculated with colour equations and zero
points and were found to agree well with the standard values for La Silla, pointing
to the photometric quality of the run. The colour equations for the CCD system were determined by
performing aperture photometry on the Landolt's standards, which cover a
sufficiently wide range in colour.
A linear regression yields the following colour equations:
where V and I are standard magnitudes taken from Landolt (1992) and and are CCD aperture magnitudes. For establishing the local standards, we selected about 40 stars on the reference frame, which according to brightness and the absence of close neighbours could serve as "secondary standards'' in the sense of Stetson's calibration procedures. All other stars were subtracted from the frame, thus alleviating the difficulties caused by the severe crowding of stars.
For determining the difference between aperture and PSF-magnitudes, aperture "curves-of-growth'' were constructed using the DAOGROW programme (Stetson 1990). These differences and difference in exposure time and atmospheric extinction were used for determining the standard magnitudes of "local standards'', i.e. selected stars on the frame, which are suitable for computing the difference between PSF- and aperture magnitudes. Using these local standards, the profile magnitudes were transformed to the standard system. The zeropoints are uncertain by 0.03 mag in V and mag in I, which is the dispersion of the frame-to-frame scatter.
We can compare our photometry with that of OBB because S. Ortolani kindly provided us with a computer-readable table of their photometric results. The transformation equations relating the OBB (, ) coordinate system top ours (, ) are:
There are 1343 stars in the OBB data whose positions coincide within 1 pixel with the stars measured by us. The differences between the data are given in Table 4 (click here). Except for a few outliers, which appear to be mostly those treated as single in our measurements and as blended doubles in theirs, the distribution of the photometric differences indicates a constant zero-point offset of about 0.14 mag for stars brighter than mag, which becomes as large as 0.27 mag for stars with mag. There seems to be good agreement between the two sets of data in V-I for stars brighter than 18 mag.
Obviously, there is a zero-point difference both in V and I. While the details of the calibration of OBB are not assessable for us, we want to cite evidence that our zero-points are not erroneous. We observed 47 Tuc in the same night and applied the same calibration as for NGC 6528. The CMD for 47 Tuc matched the Hesser et al. (1987) data excellently. The difference both in the HB brightnesses and in the RGB colours at the HB level is of the order of only 0.01 mag. Notice also that the gain in limiting magnitude with the long exposures on NGC 6528 is not so dramatic. It is the strong crowding around NGC 6528 which determines photometric errors and limiting magnitudes. Furthermore, a tendency for a non-linearity might be seen in the growth of the zero-point differences with decreasing brightness. We suspect that this may be related to the non-linearity of the RCA chips used by OBB, which were unkown at the time of observations (Abbott & Sinclaire 1993) (however, the referee expressed doubts that a non-linearity may result in such a large difference).
Comparison with the photographic data of van den Bergh & Younger (1979)
is done for the V magnitudes. The deviations become
larger at the limit of the photographic data ( 17 mag), indicating
a strong non-linearity in the photographic calibrations.
The statistical results are given in Table 5 (click here). The referee pointed
out that the cross-check present values-OBB-van den Bergh is not
consistent, since OBB give -0.1 mag as the difference OBB-van den Berg.
We are unable to resolve that inconsistency.
Table 6: This table is available at CDS. It contains 24999 stars with x-y-coordinates (the system of Fig. 3 (click here)), V, V-I and the respective errors, as well as the chi-values returned from DAOPHOT