Users upload a text file of their photometry and the calculator returns a file with corrected photometry. The uploaded file should have the following columns, in order:

• Flux (in instrumental magnitudes or electrons). Flux measurements should be obtained from drz files using a 3-pixel aperture radius.
• Sky background (in electrons). Sky level should be measured locally (e.g around each star, in an annulus 5 pixels wide and r_min=15 pixels from the star center).
• Number of pixel transfers on the parallel register.

The "number of pixel transfers on the parallel register" is simply the distance in pixels from where the star is located to the serial registers of the CCD, along the Y direction. If the Y-coordinate option is selected, the third column should instead show the star's Y coordinate in the FLT frame. In this case, the input file should also include an additional fourth column with the CCD number (1=WFC1, 2=WFC2).

In addition to correcting input photometry, the output file will include a header that explains the meaning of several different flags. Users should carefully check these flags to ensure that photometric data lie within the region where the correction formula is most reliable.

The correction is accurate to better than 4% for stars between 50 e^- and 80,000 e^-. Outside these limits, the error increases sharply. However, for stars brighter than 80,000 e^- the CTE loss is never larger than ~4% (unless the background is lower than ~10e^-). For very bright stars, a larger aperture radius may be used to perform photometry, in order to recover a greater fraction of the lost flux. CTE correction can also be obtained for photometry with slightly different aperture radii, albeit with a loss of accuracy; for larger radii the data will be slightly overcorrected. We hope to expand the region of optimal calibration using data from future cycles.

1. Obtain FLT images using CALACS, and then use Multidrizzle or Astrodrizzle to obtain DRZ data. Alternatively, use the DRZ files from the pipeline as provided by MAST.
2. Multiply the DRZ images by the exposure time. In case multiple exposures are combined into a single drizzled image, the exposure time of the final product should be used. If the exposures were taken with different exposure times, we recommend that photometry be performed on each of the images separately. The results can then be corrected for CTE losses and finally averaged to obtain a more accurate measurement of the stellar flux.
3. Perform aperture photometry with any preferred software (e.g. daophot). Set the aperture radius to r=3 pixels. Measure the background for each star locally (e.g. in an annulus of r_min=13 pixels and r_max=18 pixels, centered on the star).
4. Obtain the date of the observation from the header keyword DATE-OBS.
5. Measure the number of transfers for each star. The number of transfers are best obtained by identifying the position of each star on the FLT image. Use either the task "tran" from stsdas.analysis.dither in PYRAF, or measure the coordinates of the stars directly on the corresponding FLT image. For stars that fall in WFC2 (which is included in extension [1] of the FLT .fits file), the number of transfers Y_tran is simply the y coordinate of the star. For WFC1, which is included in extension [4] of the same fits file, the number of transfers is Y_tran=2049-ystar, where y_star is the y coordinate of the star on that frame. Note that this needs particular care in case the observations are dithered. If the position of the star changes only by a few (< 10) pixels, the correction is still well within the error even for large losses. However, if the dithers correspond to larger shifts, users should carefully check the original position of each star in the FLT files, and then derive the correction using the average value of Y_tran. While doing so, make sure that the parameters used in astrodrizzle (or multidrizzle) do not bias the results when the CR rejection is performed.
6. Load a file with columns for flux, sky, and y-transfer information into this calculator and obtain corrected photometry.
7. Perform the aperture correction.
8. If desired, perform additional corrections (i.e. converting the flux from e^- into e^- s^-1, or applying zeropoints to convert the flux into AB or Vega magnitudes).