Processing Lunar AVI's Using RegiStax 5
(by Bob Pilz)Sharpening
After
stacking, we have
an image with greatly reduced noise, but the stacking process also
introduces a
fair amount of softness into the image. A proper sharpening workflow is
necessary to pull out all the detail inherent in the stacked image. The
Wavelets
tab
contains a powerful
sharpening function called Wavelets. I use it as the first step in my
overall
sharpening workflow.
Sharpening workflow is a huge subject unto itself, and even controversial, with everyone seeming to have their own favorite method. I am no exception. Rather than duplicate what I’ve written before, I’ll refer you to an article I wrote for Selenology Today Journal #9. It can be found at: http://digilander.libero.it/glrgroup/journal.htm
Example
The figure below shows the results of several of the above stages of processing. Due to space limits, this shows just a small area North of Hainzel and has been enlarged 1.5x from the native recorded size.
The first section shows the best single frame from a recording of 15,000 frames. There is some detail, but the noise is very evident and prevents any additional sharpening.
The second section shows the results of creating an initial reference frame as described earlier for use during the Align stage. The visible noise is much reduced as one would expect from stacking frames (in this case the 10 best frames at each of 9 align points spread evenly around the frame), but the detail is starting to smooth out. Although the noise is less, it still is not low enough to allow any significant degree of sharpening.
The third section shows the results after completing optimize and stack stages and shows a stack of 400 frames (the best 400 at each of 112 align points). A stack of 400 reduces the noise by a factor 1/20 ( i.e. 1 / SQRT(#frames)). The noise is very low, but the detail has been even more smoothed out by the stacking. However, the detail is there and is just waiting for sharpening to bring it out. We must have faith J.
The last section shows the results after sharpening and contrast/brightening adjustments. All that processing did its job and greatly improved on the overall best single frame. The entire final can be found here: http://www.pbase.com/bob_p/image/109445362/original .
Figure 10.