Thanks Tim, trying to preserve detail on a >90% moon does take some practice.
I was surprised by the Autostakkert Quality Graph being mostly above 75%. If I look back through my older posts then typically I take the best 25% that are above the 50% quality line. I don't know the reason for this. Capturing in FITS rather than SER is one difference, note that the Max ADU in FITS Liberator is 46,000 which means the 3ms captures did not saturate the sensor. Also there is the camera pixel size. The Atik has 6.45µm pixels. However I have also had good results with a QHY5LII-M (3.75µm), Skyris 618M (5.6µm) and Altair 183M (2.4µm) - for these 3 cameras, drizzle 1.5 in Autostakkert seems to give a better image. On Astrobin, the camera of choice (if you want IOTD or Top Pick) seems to be the ZWO 174MM (5.86µm) coupled with a C14/C11 scope. Some more data capture and investigation required.
Seeing is a lottery. I have captured 20,000 frame SER files to get 10% above the 50% Quality line - that must have been a turbulent atmosphere. Filters do help - Red with lower elevation and the IR685 pass filter for high elevation or when still light. The most stable atmosphere I have found to be in the early morning before sunrise. Some of my best lunar images have been just before twilight with the IR685. Some IR685 evaluation notes from the Baader website:
https://www.baader-planetarium.com/en/d ... -paech.pdf
https://www.baader-planetarium.com/en/d ... -paech.pdf
I have been experimenting with processing the RGB data using only PixInsight. I have managed to produce this:
- Image114d.jpg (958.44 KiB) Viewed 407 times
I am now working on where the colour went. PixInsight processing was as follows:
Stacking - use FFTRegistration to stack the R, G, B data sets. The same file must be used as the reference file (I just used the first red).
- FFTRegistration.JPG (150.46 KiB) Viewed 407 times
Colour combination of the R, G and B stacks using ChannelCombination. The image shows the R, G, B readout values.
- RGB-readout-values.JPG (38.06 KiB) Viewed 407 times
Sharpening using MultiscaleLinearTransform, note how small the values are.
- sharpening.JPG (24.86 KiB) Viewed 407 times
Before carrying out CurvesTransformation for stretching/saturation I applied a mask generated by RangeMask to protect the brighter areas.
- range-mask.JPG (35.37 KiB) Viewed 407 times
IntegerResample and Rotation were used to finish off. This was the final result. I have some work to do here but it is early days yet.
- RGB.JPG (31.55 KiB) Viewed 407 times
I have found a well planned lunar session (using Virtual Moon Atlas) can generate 50Gb of data in a 1 hour session.
Dave
) was just how high quality the average frame was there --averaging above 75% according to Autostakkert. Is this something that you often see or that was peculiar to this occasion? In the few lunar images that I have done the average quality is much lower --a few spikes up above 75% -- and so I limited the selection down to just the top 25% or less.