Annual Meeting of the Deep Sky Section, 2011 March 12

Using the f/2 HyperStar System for Deep Sky Imaging

Dr Arditti explained that the HyperStar system was a means of reconfiguring widely available Schmidt-Cassegrain telescopes (SCTs), which typically had focal ratios of around f/10, into Schmidt camera configurations which had much faster focal ratios of around f/2. This change enlarged the telescope's field of view by a factor of around five, allowing wide-field images to be taken with short exposures. Dr Arditti explained that he had been using such a system for around 2½ years, but that, with the exception of the manufacturer's specifications and some enthusiastic reviews by Greg Parker⁸, he had been surprised by how little he had seen written about people's experiences of using it. In this talk, he would describe his own experiences of using it from Edgware, London.

Dr Arditti explained that the HyperStar system had evolved out of an earlier product, Fastar, which had been developed by Celestron in the 1990s but which they had discontinued in 2005. Though many Celestron SCTs were still marketed as being Fastar compatible, now the only such kits available were manufactured by an independent Arizona-based company, Starizona, under the new brand name. In practice, conversion kits were now available from Starizona for many models of SCT which weren't marketed as being Fastar compatible, including a few Meade telescopes such as the 14'' LX200. As well as offering support for a wider range of telescopes, Starizona also claimed to have reduced the spherical and chromatic aberration of the earlier system.

The purpose of converting an SCT into a Schmidt camera was explained in terms of the length of exposure needed by astrophotographers to record satisfactory images of various objects. For unresolved stars, the amount of light collected from the star scaled simply in proportion to the area of the aperture of the telescope. In other words, the required length of exposure was inversely proportional to the square of the diameter of its aperture. But for extended regions of resolved nebulosity, what mattered was the focal ratio – i.e. the focal length divided by the aperture diameter – of the telescope: the required length of exposure was inversely proportional to the square of this ratio. The reason for the difference was that longer focal lengths equated to higher magnifications, and higher magnifications meant that the light was more spread out across the sensor, with each pixel receiving less of it.

For effective deep sky imaging, telescopes needed both large apertures and also fast – i.e. small – focal ratios. In other words, it was important for the telescope to have a short focal length. Historically, one way of achieving this had been the Schmidt camera design of telescope, invented by Bernard Schmidt in 1930 and notably adopted by the 48-inch Samuel Oschin Schmidt Telescope used for the Palomar Observatory Sky Survey in 1949-58. In common with SCTs, Schmidt cameras brought light to a focus using a catadioptric combination of a refracting correlator plate followed by a spherical short-focal-length primary mirror. But from there onwards, the light path differed. In an SCT, the light went on to a secondary mirror, usually attached to the centre of the corrector plate, which bounced it back down the length of the telescope tube to emerge to the eyepiece through a hole drilled in the centre of the primary mirror. In total, the light traversed the length of the tube of the telescope twice between its reflection from the primary mirror and its reaching a focus, making the Schmidt-Cassegrain a relatively long focal length configuration, with a slow – i.e. large – focal ratio of typically around f/10.

Schmidt cameras achieved much faster focal ratios by removing the secondary mirror and bringing the light to a focus directly from the primary mirror, usually midway along the tube of the telescope. The light only traversed around half the length of the tube before reaching a focus from the primary mirror, equating to a much shorter focal length, and in turn a much faster focal ratio and larger field of view. Historically, a significant disadvantage of Schmidt cameras had been that their focal planes were heavily curved, requiring the use of non-flat photographic plates. Also, the placement of the focal plane in the middle of the telescope's optic tube, where it was impossible to mount an eyepiece, made them useless for anything other than photography. The first problem, at least, could now be alleviated by the use of a field-flattening lens immediately in front of the camera, at the slight expense of some chromatic aberration.

The process of converting an amateur f/10 SCT into an f/2 Schmidt camera required the secondary mirror and its mounting to be completely removed from the centre of the corrector plate, and replaced with a new assembly which could be clamped in its place. This new assembly comprised of a lens – to correct for the spherical aberration normally corrected by the secondary mirror of an SCT, and also to flattened the image plane – and a mounting for a camera – which could be either a dedicated astronomical CCD or a generic SLR camera. The size of the image on the sensor was well suited to the 27mm-diagonal chips used in most DSLRs.

However, as the camera was placed directly in front of the telescope's aperture, there was a strong incentive for it to be as compact as possible. In addition to the camera itself, cables had to be strung across the aperture to get to the camera; at the very least, a power cable and a USB cable would be needed. Hence, whilst DSLRs were supported, a more compact CCD might be preferred. The speaker used a QHY8 CCD camera that he had bought with his HyperStar system, and had found that the obstruction posed by the cables running to it didn't seem to cause serious image degradation on his 11'' Celestron, except for some diffraction spikes around stars. He supposed that they would probably be barely noticeable on a 14''-aperture telescope, but would pose a very serious problem on telescopes smaller than his own. He added that he had taken especial care to clamp the cables firmly to the telescope's tube at the edge of the aperture before running them on to his computer. If this was not done, any snag in the cables would put force directly onto the camera, and thence onto the thin corrector plate onto which it was clamped. Another consideration was that placing any cooled camera so close to the corrector plate could cause it to dew up very rapidly, and the speaker added that he always made sure to use a heated dew shield.

The speaker added that each model of camera needed its own unique adaptor to interface it to HyperStar in order to achieve a sharp focus. The focussing mechanisms of SCTs could typically only move the primary mirror through a total distance of around 20mm, which was plentiful in the SCT configuration as the secondary mirror acted as a diverging lens, extending the focal length of the telescope by a factor of around five so that the eventual image plane could be shifted back and forth by as much as 100mm. In the Schmidt camera configuration, however, there was no such extension, and the sensor within any given camera needed to be finely positioned within the narrow focal range of the instrument by means of its own specific adaptor.

Turning to describe the process of converting an SCT for use with HyperStar, Dr Arditti explained that whilst a few Celestron SCTs – those marketed as Fastar compatible – needed no modification to receive it, most needed to have a conversion kit applied first. These kits were sold by Starizona, and were available for any Celestron SCT of aperture between 6'' and 14'', as well as the Meade 14'' LX200, and comprised a new baffle tube, a new secondary mirror holder, and a counterweight. Describing the process of applying such a kit as 'fraught', the speaker explained that the telescope needed to be almost completely dismantled. The corrector plate needed to be removed from the front of the tube, and the secondary mirror and its holder unscrewed from the plate's centre. Dr Arditti remarked that his corrector plate had turned out to be glued in place, and that significant prising had been needed to free it from the baffle. Furthermore, he had not realised until too late that the corrector plate should not be rotated relative to its original orientation – one of its functions was to correct for the astigmatism of the primary mirror – and hence it was worth making a Tipp-Ex mark on its outer edge and on the baffle to show how it should be aligned.

Once the corrector plate was isolated, a new assembly could be clamped in place of the old secondary mirror holder; this new assembly could receive either a HyperStar unit, or a new replacement for the old secondary mirror holder. In addition, a counterweight was added to the telescope's eyepiece holder when in the HyperStar configuration, to balance the added weight of the camera at the front. Once a telescope had been converted for use with HyperStar, it was relatively straightforward to switch it between the SCT and Schmidt camera configurations; the process took perhaps a few minutes in a well-lit room. In the dark it was more fiddly; the speaker had dropped his secondary mirror on a recent attempt, though luckily it had come to no harm.

Turning to describe his experiences of using HyperStar, Dr Arditti identified a few potential limitations of the system. Firstly, it seemed difficult to use it with many autoguider systems, since there was nowhere where the light path could be split. Even SBIG cameras, which combined an autoguider and imaging CCD in the same unit, seemed ill-suited since their large packaging would obstruct too much of the telescope's aperture. In practice, he explained that he had resorted to using a separate guide telescope on the same mount when taking long exposures. Second, use of colour filters needed careful planning. Filters could be inserted into the HyperStar assembly, but the camera had to be removed to do so. This made it impossible to take images in multiple colours whilst preserving the alignment of the camera, and so colour imaging was difficult if not using a colour camera.

Perhaps the most serious ongoing nuisance that Dr Arditti had experienced was that very fine collimation seemed to be required. Having a fast focal ratio meant that the Schmidt camera configuration had a very small depth of focus – i.e. the image moved very quickly out of focus if the sensor was moved back or forth. Any slight misalignment of the sensor from the optic axis of the telescope made it impossible to get the whole field into focus at the same time. The speaker had found that his system generally needed re-collimation after every re-pointing, which he supposed to be caused by the weight of the camera warping the alignment of the assembly that it was attached to. This re-collimation process was fiddly as the adjustment screws were by now concealed inside the dew shield and behind the camera. A long screwdriver was needed, and when pointing close to the zenith, also a step-ladder. Long test exposures were often needed to test the collimation, since clear detections of many stars across the field were needed to ensure that they were all cleanly focussed, and the process often took over an hour.

After finding that he often couldn't achieve collimation at all, the speaker had investigated further. Dismantling his QHY8 camera, he had concluded that the sensor itself had been poorly collimated, by mounting it on the chuck of a lathe and turning it slowly by hand whilst reflecting a collimation laser beam off the surface of the CCD. He had, however, been able to improve matters by adjusting the screws used to mount the sensor until the direction of reflection was constant as the camera was rotated.

To close, the speaker showed a few of the fruits of his labours, though he added that he was generally more interested in the technical challenge of acquiring images than in processing them, which often got put off for 2–3 years. First, he showed an image of M31 stacked from 31 ten-minute exposures taken in 2009 September. He remarked that he had limited himself to exposures of a maximum of ten minutes as he lived below a flight path into Heathrow, and aircraft often passed through the widened field of his HyperStar setup. Other easy targets had included the close neighbours M42 and NGC 1977 in Orion, galaxies such as M109, and globular clusters such as M13. Some of these appeared surprisingly small in his large field of view.

Lastly, he showed the deepest image that he had ever taken with the system: a 12-hour exposure of a field in Cassiopeia, stacked from 84 ten-minute exposures, encompassing M52, NGC 7635, NGC 7538, and a vast number of field stars. The image had been taken over the course of successive nights between 2009 July 25 and 2009 September 12, but the speaker added, to laughter, that he had only got around to processing the data the previous night, for inclusion in his talk.

Following the applause, Nick James remarked that he had always been a little wary of supporting the whole weight of a camera on the corrector plate of an SCT. The speaker agreed, but said that in practice the corrector plate didn't appear to warp under its weight. An audience member asked whether the speaker felt the effort had been worthwhile, and how his results compared with what could be done with small-aperture refractors. The speaker thought in retrospect that his results were broadly comparable to what could be done with a refractor, but that he had nonetheless enjoyed the challenge of installing HyperStar.

The Director then invited Owen Brazell, usually best known as an observer of planetary nebulae, to talk on this occasion about galaxy clusters.