3 Data reduction and calibration

The data were reduced following the same procedure described in Paper I.

The absolute calibration of the observations is based on a set of standard stars of the catalog of Landolt (1992). Eighteen standard stars were observed; specifically, the observed fields were PG1525, PG1633, Mark-A and PG2213. At least 4 exposures at different airmasses were taken during the night for each standard field, making a total of $\sim 80$ individual measures per filter.

Thanks to the relatively stable seeing conditions, for the aperture photometry we used for all the standards (and the cluster fields) a 12 pixel aperture (4 arcsec). The aperture magnitudes were normalized by correcting for the exposure time and airmass. Using the standards observed at different airmasses, we estimated for the extinction coefficients in the two filters: A_V=0.11 and A_I=0.06. For the calibration curves we adopted a linear relation. The best fitting straight lines are:
$V-v({\rm ap})=(0.017\pm 0.003) \times (V-I)-(2.232\pm 0.002)$;
$I-i({\rm ap})=(0.053\pm 0.003) \times (V-I)-(2.679\pm 0.002)$.
The magnitude differences vs. the standard color are plotted in Fig. 2, where the solid lines represent the above equations.

Figure 2: Calibration curves for the V ( upper panel) and I ( lower panel) filters for the June $24^{\rm th}$ 1998 run

3.1 Comparison between the two catalogs

Apart from the 13 GGCs presented in the following sections, we re-observed three additional clusters selected from the southern hemisphere catalog presented in Paper I. We covered the same spatial region, in order to test the homogeneity of the two catalogs created using the two different telescopes.

In Fig. 3 the magnitudes of 170 stars (with photometric error smaller than 0.025 mag and V magnitude brighter than 18) in NGC 5897 from the JKT and the ESO/Dutch telescope data sets are compared. The differences in magnitude and color between the two observing runs are: $0.00\pm 0.02$, $-0.01\pm 0.02$ and $0.01\pm 0.02$ in V, I and V-I, respectively. For the other two clusters, we have: $\Delta V^{\rm JKT}_{\rm DUT}=0.02\pm0.02$, $\Delta I^{\rm JKT}_{\rm DUT}=0.01\pm0.02$ and $\Delta (V-I)^{\rm JKT}_{\rm DUT}=0.01\pm0.02$ for 163 stars selected in NGC 6093, and $\Delta V^{\rm JKT}_{\rm DUT}=0.01\pm0.02$, $\Delta I^{\rm JKT}_{\rm DUT}=0.00\pm0.02$ and $\Delta (V-I)^{\rm JKT}_{\rm DUT}=0.01\pm0.02$for 249 stars selected in NGC 6171 (selection criteria as in the case of NGC 5897). Moreover, the slopes of the color differences vs. both the magnitude and the color are always $\leq 0.001 \pm 0.002$. Therefore, the two catalogs must be considered photometrically homogeneous and the measures of the absolute and relative parameters of the CMDs perfectly compatible.

Figure 3: Comparison between the JKT and the ESO/Dutch telescope magnitudes for the 170 stars in NGC 5897 with errors $\leq 0.025$. The differences in magnitude and color between the two runs are: $0.00\pm 0.02$, $-0.01\pm 0.02$ and $0.01\pm 0.02$ in V, I and V-I, respectively. Notice also that the slopes of the differences in color vs. magnitudes or color of one of the runs ( upper panel), are negligible ( $\leq 0.001 \pm 0.002$), making the catalogs photometrically homogeneous