3. Data reduction
The data were first binned into 1-ms frames and divided by a deep flatfield image taken during the observation period. A post-exposure shift-and-add sharpening technique was applied to produce an integrated image (Shearer et al. 1996). Because of the large telescope aperture no significant improvement in the image above the seeing limit resulted, but any artefacts due to effects such as telescope wobble were removed. Data from individual nights were summed to produced the integrated V and B images shown in Figs. 1 and 2.
Photometry and astrometry were carried out using the IRAF daophot and GASP packages, respectively. The background level was taken to be the mean of the signal in an annulus of radius and width centred on the object position. Stellar co-ordinates were calculated using the astrometry of Blair & Schild (1985).
Several pulsar candidate positions were chosen as described in Sect. 4. Photon times were extracted from a window with a diameter equal to the seeing width centred on the pulsar candidate. The time series was translated to the solar-system barycentre using the JPL DE200 ephemeris and then folded in phase using the PSR B1951+32 ephemeris in Table 2. The resulting light curves were analysed using the statistic (Buccheri & de Jager 1989). The Fourier power-spectrum of each candidate was also calculated in the vicinity of the first four harmonics.
Table 2. PSR B1951+32 Ephemeris (Princeton Catalog of 706 pulsars)
© European Southern Observatory (ESO) 1998
Online publication: June 26, 1998