The calibration of the five different CCDs used in the imaging survey allows us
to derive a homogeneous set of standard magnitudes for all detected objects.
Instrumental magnitudes are related to standard magnitudes by the
transformation Eq. obtained from Eq. (1):
where 
We assume that no correction for
galactic absorption is necessary because our fields lie at high galactic latitude
and the widest extent of the full survey is only 1.53
deg.
Because we have a priori no knowledge of the colours of the detected objects in
the Johnson-Cousins system, we invert the set of Eq. (2) so that the colour
correction appears in terms of observed colour. We give the equations
for the V band where two colour coefficients are measured:
Most objects detected in one filter are identified in the other 2 filters (see
next subsection).
For the galaxies detected in only one filter, we derive the
magnitudes in the other
filters using the standard colours B-V = 1.0 and V-R = 0.5.
These mean
values are derived from the mean galaxy colours obtained for the survey
and reported in Sect. 9.2.2.
To check the reliability of our colour coefficients listed in Table 3
and the resulting standard magnitudes, we estimate V in two possible ways using
Eqs. (3) and (4). The mean
differences in the 2 corresponding magnitudes for 7000 objects observed at the NTT
and 5000 objects at the 3.60 m are 
mag.
This test demonstrates the reliability of our measured colour coefficients.
Because the R and V frames cover almost the same area of the sky,
the final V magnitudes are calculated in terms of V-R colour (Eq. 4).
Although R and V frames cover almost the same area of the sky, the B frames
cover only about 80% of area observed in R and V,
partly because of the
smaller CCD size in the EMMI blue channel.
Within the common area of the B, V and R bands, we define a colour completeness
rate for objects selected in the B band
as the fraction of B detected objects with associated V or R or both detections
(in bins of 1 mag). We calculate the analogous curves for the V and R
bands
inside the same area. These various curves are shown in Fig. 7 (click here).
Typically, the completeness drops below 90% 1 mag brighter than the limit of
the catalogue in each band. At the limit of the spectroscopic catalogue
, the completeness in B and V is greater than 95%.
Figure 7: Colour completeness of B, V and R detected objects
in B-V and V-R colours. The completeness is estimated in bin sizes of
1 mag. The coding of the curves is specified inside the graph