You will first be prompted for the pixel coordinates corresponding to the center location, via the standard LEXTRCT method to enter a pixel position. It is possible (and generally correct) to specify the location to an accuracy of 0.1 pixel.
Next you will be asked for the width of rings in detector units. Lextrct bins the radial profile in to radial bins of this width. The initial default is to use a width of one map pixel per bin. If the bin width is smaller than one map pixel, then it is possible for some of the rings to contain no data, i.e., gaps will appear.
You will now be prompted for the maximum radius in detector units. You should chose a radius that is large enough to include the PSF and some background, however, not so large as to include any background variations.
Next, you will be prompted for a range of azimuth to include. Use 0,360 to use all the data. Angles are always measured from [0,360) which means an angle can be 0.0 exactly, but not exactly 360.0. If you enter a range of 0,90 this in interpreted as [0,90), likewise an entered range of 90,180 will cover the range [90,180). With this convention, you can easily enter two adjacent ranges, such as 0,90 and 90,180 and be fairly confident that no data (pixels) will be included in both profiles.
For the question, 'Should errors be based on scatter (N)?', if you answer Y then the errors will be based on the standard deviation of all the data included in a bin. Note the innermost bin often contains only one point, in which case the scatter is set equal to the mean. If you answer N then the program assumes photon counting statistics.
Finally you will asked for the background level as the number of counts per detector pixel. The program attempts to estimate this from the outer few rings of the radial profile. Since the outer rings typically contain a large number of pixels, the total power is sensitive to the exact level of the background.
If you have not read in a background file, then Azim uses the mean in the outer three rings to estimate the background. This estimated background is subtracted from the encircled energy curve but not from the radial profile function.
If you read in a background file and subtract the background from the data map, then no background is estimated (or subtracted). For this case errors are incorrectly calculated. (If you do this you should fix the code to work correctly for this case).
For the encircled energy profiles, the error is based on the square root of the number of enclosed counts. Since each radial bin includes the data from the previous bin, plus the current ring, the data is highly correlated. Thus the error bars are much bigger than the observed scatter. Thus, these errors can be used to calculated the error in a single point on the curve (such as the half power radius) but should not be used to fit the total encircled energy profile.