The sky was roughly 40% covered with cumulus when the photometry observations was started around 14:30 UT. The first strategy was to select a star and wait for clear sky, but after two interrupted imaging runs (each star was taken in the sequence UBVRI in each frame of a 4x4 mosaic) the strategy was changed to point towards a clear part of the sky not to close to the Sun and look for a nearby bright star (with photometry present in Johnson et al. (1966)). The mosaics were also reduced to 3x3. All filter was exposed in 0.5s except U which had 5s exposures. This was selected so that the background was sufficiently high at the same time as the star should not be saturated independent of altitude and angular distance to the Sun.

A short exposure time could introduce artifacts from the shutter. However, nothing obvious was seen in the images. During the exposure the sky was monitored by eye and through the search telescope. Because of the small aperture which made the f-ratio high (roughly f/112.5), and the short exposure times, any cloud could easily be seen in the images and discarded in the later reduction process. In total 8 stars were observed (see Table 1 in chronological order).

Since only one star at the time was imaged the angle relative to the telescope was kept at all time in order to make the flatfielding easier.

Bias and dark were removed with dark frames with the same exposure times as the science frames. Because of the high f-ratio all dust on filter and CCD window were almost focused. This posed a problem since the filters were changed for each frame in the mosaic and the filter position was not reproduced exactly. However, since the angle relative to the telescope was kept it was possible to use images from all stars in one filter and selecting them into subgroups with close to the same dust position. About 80 images in each filter could then be divided among some 5-8 subgroups. If the subgroup contained 6 or more images they where normalized, and median combined. The resulting flats were used to flat field the images in respective group. The only significantly deviating background found was in the Kochab I-filter. The cause is unknown but a possible explanation is some direct sunlight into the outer tube which has its inside covered with a black cloth which is known to be non-absorbent in I.

After that the images were sorted with respect to filter and mosaic run, aligned to the star and averaged combined. Thus photometry for the stars in Table 1 down to Izar is based on 1-4 combined images, the first Alphecca run has 13-14 images combined, and the rest 7-9 images. The FWHM was around 4-6 arcsec in the combined images. A photometric aperture of 12.1 arcsec was used and the background ring had a thickness of 5.5 arcsec. The gap was in most cases 2.75 arcsec but in a few cases a gap of 8.8 arcsec were used in order to avoid obvious artifacts from the imperfect flat fielding. Along with the ADU counts of the star the background counts. Table 2 shows typical values and the range of values. The lowest values were found at combinations of high altitude, low solar altitude, and great solar distance. The highest values were found at the opposite combinations.

Since the standard photometric values for the stars were given in Johnson R-, and I-filter, those were first converted into Cousins R, and I-filter (Taylor 1966)

(V-R)_C=0.717((V-R)_J-C₂)-0.021 (Eq. 1)

d(V-R)_C=((d(V-R)_J0.717)²+(dC₂0.717)²)^1/2 (Eq. 2)

(V-I)_C=0.902((V-I)_J-(V-R)_J-C₁)-0.087dBJ+0.073+(V-R)_C (Eq. 3)

d(V-I)_C=((d(U-B)_J0.087)²+(d(B-V)_J1.378*0.087)²+(d(V-R)_J0.717)²+(d(R-I)_J0.902)²+(dC₁0.902)²+(dC₂0.717)²)^1/2. (Eq. 4)

In the transformation V_C=V_J is assumed. Furthermore, C₂=-0.018±0.002 if B.C.>4700 otherwise 0, and C₁=-0.009±0.002 if 5400<B.C.<7200 otherwise 0. The Balmer-jump correction is calculated by

dBJ=(U-B)_J-1.378(B-V)_J+0.709. (Eq. 5)

In order to convert the background ADU counts into UBVRI magnitudes two steps were done. First the ADU was converted into electrons and magnitudes by

m_b=-2.5log₁₀(gI_b)+2.5log₁₀(ts²), (Eq. 6)

where m_b is the instrumental magnitude in mag/arcsec, g is the gain in electrons/ADU and I_b is the background counts in ADU/pix, t is the exposure time in seconds, and s is the pixel size in arcsec/pix. The second step is to subtract the zero point Z of the filter. However, in order to do that the stars must be used which means that the atmosphere must be taken into the calculation. The magnitude M of a star above the atmosphere is

M=m_X-k_m'X-k_m''Xc_X-k_mC-Z_m, (Eq. 7)

where m_X=-2.5log₁₀(I_Xg/t) is the telescope magnitude converted from the flux count I_X ADU and normalized to one second, k_m' is the first order extinction coefficient (mag/airmass), X is the airmass, k_m'' is the second order extinction coefficient (1/airmass), c_X is a telescope color (mag) nearby the filter wavelength, k_m is a correction to the standard system, C is a atmosphere corrected color, and Z_m is the zero point. A color can be determined in a similar manner by

C=kc_X-k_c'X-k_c''Xc_X-Z_c. (Eq. 8)

The uncertainties for the background is then

dM_b=((dm_b)²+(dkC)²+(dCk)²+(dZ)²)^1/2 (Eq. 9)

dC_b=((dkc_X)²+(dc_Xk)²+(dZ_c)²)^1/2. (Eq. 10)