In this chapter I summarise the results in my thesis, and suggest future paths for Lucky Imaging. Future plans for the VLM binary surveys are detailed at the end of chapter 7.
LuckyCam gives diffraction-limited images in the visible on a 2.5m telescope, easily and cheaply. It is clear that the system gives excellent results in a wide variety of conditions – in addition to the wide range of seeings investigated in chapter 4, all but one of the 24 VLM binaries detailed in chapters 6 and 7 were resolved with diffraction-limited resolution.
Lucky Imaging also offers a very large i-band isoplanatic patch, 15-30 arcsec at the NOT. This, along with LuckyCam’s very faint guide star capability, gives excellent sky coverage with natural guide stars. The calibrations, performance measurements and science results presented in this thesis demonstrate Lucky Imaging’s utility for a wide variety of science programmes.
Lucky Imaging degrades gracefully under difficult conditions. Images with diffraction-limited cores were obtained over a very wide variety of conditions, but when it was not possible to obtain them the output image resolution was still very much improved. For example, when operated in B-band , in very poor seeing, or at slow frame rates (chapter 4), LuckyCam still attains a very significant resolution increase even when no diffraction-limited frames at all would be expected.
This ability ensures that LuckyCam-like instruments can be used for a very wide variety of astronomical programmes. Because they offer essentially zero readout noise, and behave otherwise like standard CCD cameras (with the exception of multiplication noise), the frame-selection and recentering methods will always improve the output images compared to a standard system – especially as the image improvement methods can be tailored to both the science being performed and the conditions (chapter 3).
LuckyCam-like instruments are a simple, and thus very cheap method of obtaining diffraction-limited resolutions. A “standard” LuckyCam consists simply of a filter, a reimaging lens, and an L3CCD detector, and yet obtains resolutions in the visible that complicated and expensive adaptive optics systems manage only in the near infra-red on 8-10m class telescopes. Although Lucky Imaging in the form presented here operates most effectively on smaller 2.5-3m telescopes, it increases their capabilities enormously for a very small cost.
It is worth emphasising that it would have been far more complex and time-intensive to perform the VLM binary surveys without LuckyCam. At an average of 8.7 minutes per target (including all overheads), LuckyCam’s zero lock-on time for high-resolution imaging allowed the survey of 110 target stars to be completed in only 16 hours of on-sky time – despite many of the observations being performed though 3 magnitudes of cloud-cover. LuckyCam’s ability to guide on faint stars proved invaluable for the investigation of targets which are not accessible to standard AO systems. It is telling that, in the sample presented in chapter 7, the 14 new close binaries discovered with LuckyCam are all significantly fainter than the previously known systems.
The Cambridge Lucky Imaging group is currently working to further extend the utility of Lucky Imaging, by building a full realtime Lucky Imaging system (rather than the preview system described in chapter 2). This will implement the pipeline described in chapter 3 on DSP (digital signal processor) hardware, and will allow non-specialists to use the system at the telescope without having to post-process hundreds of gigabytes of data.
In the further future, the development of relatively low-noise detectors in the near-infrared will allow Lucky Imaging for diffraction-limited images on large (5-8m) telescopes in the near-infrared. This method would be much cheaper and less complex than an equivalent adaptive optics system, especially as the high sky brightness in the near-infrared necessitates the use of a short-exposure detector for any imaging.
The Lucky Imaging concept, that one can select and only use the very best data obtained by an imaging system, is also applicable to other high-resolution imaging systems. For example, a LuckyCam-like camera placed behind an Adaptive Optics system would allow the selection of those times when the atmospheric turbulence is low, and the AO system is thus producing the very best images. It is possible that such hybrid systems would allow visible-light near-diffraction-limited imaging on 5m class telescopes, and I hope to experiment with such systems in the near future.
We also hope to test a new Lucky Imaging concept in June 2006. Developed by Craig Mackay, CLASI (the Cambridge Lucky Aperture Synthesis Instrument) combines elements of Lucky Imaging and multiple aperture interferometry (Mackay, 2005). The aperture of a large telescope is divided into several sub-apertures. Light from groups of four sub-apertures is combined at a focal plane, to give PSFs crossed with fringes. High-resolution information about the objects in the field can be recovered from the fringes, at scales corresponding to the baselines between the apertures.
Conventional aperture-masking systems require very bright targets, as the detection of fringes usually requires an aperture size of less than r0. However, the use of an L3CCD detector for short exposure imaging allows Lucky Imaging of the PSFs, ensuring that the images are diffraction-limited with large (1m-scale) apertures, and thus observations of faint targets. Multiple detectors observing multiple groups of apertures allow the total aperture to approach that of the fully-filled telescope.
The tests we plan to undertake in June will (hopefully) verify the concept; proposals for the development of the system to give diffraction-limited imaging at 800nm on the 8m Very Large Telescope are currently under review.
The Lucky Imaging concept enables high-angular-resolution imaging using a simple and cheap instrument. With many planned developments of the system, and a bit of Luck, the future of high-resolution imaging in the visible from the ground looks secure.