Comparison of flats after a set of changes in the focuser of my imaging scope.

The imaging scope originally had a Bird-Jones corrector at the base of the focuser draw tube. I long-since removed that but the plastic lens holder was left in place. When looking for sources of reflection, I finally got around to a proper analysis of the stops in my scope and discovered a few surprises. One was that the aperture stop of the scope was that little lens-holder.

I finally managed to remove the object after a lot of wrangling and hand-wringing to avoid damaging the draw-tube. It then turned out that there were a lot of reflections from the stock flocking-paint used in the draw-tube. In addition, the tube-wall beneath the now-removed lens holder was bare metal and highly reflective. In order to diagnose the situation and document the remedy, I made these images.

I considered baffling the draw-tube, but it was going to cause even more vignetting since I was already losing light at the edges of the field. Instead, I gave flock-paper a shot. Initially, I just tested with some black construction paper I stole from my kids. This improved the flat so I ran out to the store and got some black velvet and put that in the draw-tube instead. It's even better.

The three images are T-shirt flats taken with identical camera settings and exposures. They are each post-processed to fully stretch the data, so they wind up looking fairly similar. However, I include histograms of the original flats for comparison of absolute numbers (all this is done with IRIS + GNUplot for plotting). As you can see, the overall illumination decreases though I have not made any new stops in the system. That means it was all reflection. In addition, the breadth of the histogram sharpens and gets more peaked with better flocking. I interpret this to mean that the "flat field" is, in fact, getting flatter and there is less spread in the brightness pattern.

There is still bright edges or this dark "donut" shape in the flat. I am going to put velvet on the back end of the tube behind the secondary to try and address that. I'll then compare all this to the theoretical vignetting function I can calculate for the scope.