I have created a Master Dark by median combining the Havas Sept. 21, 2002 120 second darks
I subtracted this dark from all 20 of the Sept. 21, 2002 R images. I then aligned all the images to remove the telescope tracking error and median combined the images. The result was converted to jpeg (all processing done with MaxIm_DL v3.08):
None of the objects found by Charles (marked by a triangle), Jan (the square), or Steve (the three lines) show up. I have circled two sets of stars that are quite faint and are visible on the NEAT and DSS images to illustrate that my processing has preserved faint details. The spots being pointed to by Charles, Steve and Jan appear to be incompletely removed noise pixels.
To show that there really is not anything at the spots pointed out by Steve, Jan, and Charles, I had MaxIm_DL zoom in on the FITs image to show clearly the individual pixels. The result is seen below:
Steve's "star" is entirely due to a few pixels (not enough for a star image) on shavas-15R.fts. This is illustrated below:
The processed FITS image can be downloaded for your examination by shift-Left-clicking on this link, saving the file, and unzipping it to fetch the .fit file.