The point data loading widget already present on your desktop gives you the choice of loading the point data for all stars or a specific selection of them. Load all data if you have sufficient memory (this may take a few moments). If not all data can be loaded, select a group of stars by dragging the cursor over the star list in the widget. Load the data for the selected stars by clicking the LOAD button.
Use REDUCEPOINTDATAIMAGINGPLOT to plot and edit the point data data in much the same way as you did with the background data. Start by plotting the DelayJitter versus Index, and use the AUTO editor (PL/EAUTO) on it. Do the same for NATJitter (if data present). (The first one has an OutputBeam and Baseline index, the second one an InputBeam index.) The jitter data provide a good first check whether the visibility data is going to be useful or not. Make sure you look at all spectrometers and baselines. If more than one baseline is selected, they are plotted along the X-axis; channels are stacked vertically. Use plots vs channel if you have to flag several entire channels. Use the PLOTAUTO option to automatically edit outliers in PhotonRate and VisSq (select all channels for speed). It is recommended to not flag the data in the laser channels, since VisSq will not be averaged anyway if the background data for these channels has been removed. This prevents the flag table from getting too large. Save the PointFlagTable.
Examine and edit the FDLDelay data on a star-by-star basis, using the Fit utility to detect outliers. You should plot FDLDelay vs Time for this. However, there are usually no outliers if they have been removed from the DelayJitter data, so this step may be skipped and returned to if upon examination of the averaged delays outliers are found.
If reducing 6-way data and photometric and bias corrections are desired, set a default background before averaging (i.e. set all background scans to zero using the defaultbg procedure, or the REDUCEBG DATADEFAULT button). After averaging all data, perform the calibrations by using CALIBRATECHANNELS, and then re-average the data (but do not forget to reload, and expand, the back ground data, and apply any flag tables).