How to use lucky imaging for deep sky objects (DSO) to obtain high resolution images
Lucky imaging is not magic, but effective. I am using short exposures times, often in the range of 500 ms to 2 seconds per sub.
This has some advantages over more traditional deepsky imaging techniques... with short exposures you do not need an expensive mount,
auto-guiding is also not needed and the frames you take are often sharper than their long exposure counterparts.
With recent camera advancements - with low read noise and high quantum efficiency (QE) - this can work really well
to produce high resolution deepsky images with more details.
lucky imaging best practices - imaging process
1) I always take images with a high gain value (e.g. 120) for 1 second as often as I can.
2) After 5 mins I stop the exposure process and then I save this sequence as one series/sequence, and so on and so on. So I get over the night, for example, 30 or more sequences together.
3) I stack each sequence in Autostakkert to FITS files, the darkframes are already subtracted by Autostakkert.
4) During the image selection in Autostakkert I select the best images of a series, e.g. 50-70% of the single-images (depending on the seeing).
5) I process the FITs files later in pixinsight. In pixinsight there is a function called "SubframeSelector". I use this and throw out the bad ones, the ones with the best FWHM value I keep it.
6) P.S. with the right image processing in PI and especially with "Deconvolution+PSF" function the results are really good and can be compared with results of much bigger telescopes.
lucky imaging best practices - image processing
1) Darks are substracted by Autostakkert, I don't use flats.
2) I use after Autostakkert the classic Pixinsight process, like:
3) StarAlignment, ImageIntegration to get a light FIT
4) DynamicCrop, DynamicBackgroundExtraction
5) ArcsinhStretch to push color, HistogramTransformation for balance
6) PSF to select a suitible star, Deconvolution, trial and error, no fix plan
7) Sometimes I apply it, but it is not always effective: Topaz DeNoise AI
What you can see here and above are effects of seeing. Better seeing, sharper stars and more structures and details in objects, poorer seeing, larger stars and washed out structures.
Particularly, local seeing conditions have a negative effect at long focal lengths. Therefore, the useful selection or a usage rate decides on more or less details in objects.
First, choose the right exposure time... for example, better seeing = short exposure times, bad seeing = longer exposure times. Second, analyze the results e.g. with the
frame selector tool from pixinsight and after it consistently take out images with bad FWHM values. Third, processes (stacking) the images with the best FWHM values.
For short focal lengths, I'm thinking, this procedure is not so important. But if you want to do high resolution astrophotography with long focal lengths, in any case...
Improve resolution of your images
You can further improve resolution of your images by improving your FWHM (Full Width at Half Maximum) - better guiding, making sure you image when seeing is exceptional, using larger aperture, or using some advanced stacking algorithms. As you have noticed, there is deviation between FWHM values in different frames. Use the subs with low(er) FWHM values for further processing. You can give it a try with some smaller pixel camera (like ASI183 - low readout noise, quite high QE). CMOS cameras are also better suited to short exposure subframes, normal CCD Cameras
for long exposures. You can try a drizzle + deconvolution combo with your current setup. I used PixInsight Drizzle Integration and then Deconvolution with PSF model and mask.
Drizzle integration makes sense if you think that image is undersampled.
The most successful type of data collection by the amateur is the transit method. As a planet passes over the portion of the star facing us, the light curve of the star drops for a time. As the planet passes through, the light curve returns to normal.
Taking a series of images of the field surrounding the host star of an exoplanet before, during, and after the predicted times of the exoplanet transit across the face of its host star. The transit method has been very useful for detecting a - hot Jupiter-, namely a large planet whose orbit is close to its host star and where the planet passes directly in front of the star from the perspective of an observer on Earth. Exoplanet transits are typically 2-4 hours long. However, conducting an exoplanet observation also involves beginning the imaging session at least 30 minutes prior to
the predicted beginning of transit and continuing for at least 30 minutes after the expected transit. Thus, it is not unusual for an exoplanet observing session to be 3 - 5 hours in length.
Light curve of Wasp-12, generated with AstroImageJ, an image analysis tool for astronomy. A technique called differential photometry is used to determine the changes in brightness (flux) of the exoplanets host star that might indicate an exoplanet transit. This technique compares the relative difference between the host star and one or more (assumed to be non-variable) comparison or comp stars during the imaging session. Since the difference in brightness of the host star and comp star(s) are equally influenced by common factors such as thin overhead clouds, moon glow, light pollution, etc.,
a change in this difference would be a measure of the effects of the drop in brightness of the host star due to an exoplanet transiting in front of it.