Thanks Pieter! I read your paper with interest. Dragonfly is a very cool project from an astronomical perspective, but there are also several points that could be of interest for the more general Canon shooter / astrophotographer:
- The optics are found to be essentially diffraction limited, at least <1 degr of the center, which is quite amazing. This means that the image sensor will give increasingly sharp images all the way to a pixel pitch of about 0.63 µm before out resolving the lens. This is 6.8 times smaller than the effective pixel size of a 7D and would imply a 839 MP APS-C or 2.18 GP FF sensor! The lens resolution is likely decreasing from diffraction limited away from the central regions, but this sets the upper usable limit for sensor resolution. In practice, there will be other things limiting the resolution, like turbulence in the atmosphere (for anything shot at a distance).
- Strehl ratios between 0.2-0.8 (where 1.0 is 'perfect') indicates that the lens provides a very high contrast (which we knew). I wonder if the Canon designers have deliberately tried to keep the effective point-spread function constant with wavelength (reducing colour mis-matches in images), deacreasing the Strehl at short wavelengths.
- The precise focus of the lens is strongly temperature dependent: a temperature difference as small as 1C gives a significant shift in focus. This is mostly relevant for exposing an extended period, like in astrophotographical applications.
- The foot of the 400/2.8L IS II provided by Canon shows significant flexure so that care must be taken during tracked exposures.
- For those wanting more images, I found two papers accepted by ApJL providing such for M101:
The project seems to have left the start-up phase only recently (they are apparently still extending the array), so I'm sure we can expect more results soon. Note, however, that they are targeting the faint structures around galaxies, and with only two broad-band filters ('r' and 'g'), meaning the images will probably not be as spectacular aesthetically as narrow-band imaging of more photogenic nebulae. But when they're done with the galaxies perhaps they can put in some nebular filters instead, and give us the most surface-brightness sensitive images of nearby nebulae
Actually, surveying for nearby supernova remnants in H-alpha might be a pretty interesting project scientifically in itself for this Dragonfly.
And then we have the transiting exoplanets, of course, but that has probably been covered pretty well by the already mentioned super-WASP projects and the upcoming NASA TESS mission.