mirror mirror


ahead of the forthcoming dark frame / deep field exhibition at BREESE LITTLE, we're previewing a number of artists featured in the show. Sophy Rickett's Objects in the Field (2013) and the Observation series (1991/2013) are the result of a collaboration with Dr Roderick Willstrop from the Institute of Astronomy at the University of Cambridge. Rickett's works (which employ unseen test shots from when everything was still done on film) juxtapose deep time with history on a human scale by resurrecting these astronomical photographs. although they are only a few decades old, they are already technologically obsolete, making Rickett's work analogous to the archaeology of astronomy itself

Rickett's work with Dr Willstrop has opened up a fascinating chapter in the history of astronomy. to find out more about the science behind the art, we asked the retired astronomer/physicist more about the unique innovation that forms the basis for these images: his development of a working three mirror telescope, which allowed for a wider field-of-view than had been previously possible


super/collider: what is it like seeing these images in a gallery setting so long after they were taken?
Dr Roderick Willstrop: very gratifying. modern methods of making enlargements give much better results than were possible even with the best lenses of the 1990s and an enlarger which was state-of-the-art in 1956 – when it cost half as much as the annual salary of a Junior Assistant Observer! at the time the photographs were taken I was able to make enlargements up to 20 times, but only of a small part of the negatives. the enlargements Sophy has created are much more impressive because they cover the whole of each negative what is your own personal favourite image in the exhibition? I think the image of the Andromeda Nebula, Messier 31, is the one that shows the smallest, sharpest, point-like images, and it is therefore my favourite

what is the best thing you've ever seen through a telescope?
either Jupiter with its Great Red Spot, seen in the 3.9 metre Anglo-Australian telescope (during testing, not wasting valuable observing time!) or, for obvious reasons, the Sun

how were the images you originally created different from what was previously possible?
the Three Mirror Telescope was designed to be an alternative to the Schmidt camera, especially for apertures greater than about 1.5 metres (60 inches). there is a practical limit to the size of Schmidt cameras. these instruments use a large spherical mirror to focus the light, and a thin lens placed at the centre of curvature of the mirror to sharpen up the images. the lens needs to be convergent near its centre, and divergent near the edge, and is therefore more difficult to make than a spherical surface. the separation of the lens and mirror is twice the effective focal length, resulting in a long tube, which needs a large, expensive, dome


very few Schmidt cameras have been built larger than the Palomar Schmidt and the U.K. Schmidt Telescope, each with an aperture of 1.2 to 1.25 metres (48 to 50 inches) and focal length 3.0 metres, giving a focal ratio (focal length/aperture) of f/2.5

if the corrector lens were removed from these telescopes, the mirror alone would produce images 100 arc seconds in diameter, worse than the resolution of good human eyesight, and more than twice the apparent diameter of Jupiter. the lens can correct this defect perfectly at only one wavelength (colour) which is usually chosen to be in the blue or green part of the spectrum. In red light the lens is too weak, and in the violet and ultra-violet it is too strong, producing images two or three arc seconds in diameter. this was not very important when the most sensitive photographic emulsions had coarse grain, but from the 1970s onward with the introduction of emulsions like IIIa-J with finer grain it became a more serious matter

three large Schmidts, the two noted above and one at the European Southern Observatory, have therefore been fitted with achromatic corrector plates, having two components, or crown and flint glass, on the same principle as the achromatic lenses of all refracting telescopes. the Palomar (renamed the Oschin) Schmidt and the U.K.S.T. should be able to produce images 0.25 arc seconds in diameter, or smaller, at all wavelengths from 400 to 800 nm (violet to near infrared)


the Tautenberg Schmidt is slightly larger in aperture than the Oschin and U.K.S.T., and was given a focal length four times its aperture, alleviating the problem to some degree, and a larger Schmidt in China avoids the colouring effects of lenses by using an inclined reflecting corrector mirror. the inclination of this mirror must be great enough so that the large spherical mirror does not cause any obstruction to the incoming light. ideally, the corrector mirror should not be rotationally symmetrical, and the field of view with good definition is more restricted than in the refracting Schmidt

the Three Mirror Telescope is based on a design by the French optician Maurice Paul, published in 1935. Paul's design used a concave paraboloidal primary mirror, a convex spherical secondary mirror placed two-thirds of the way from the primary mirror to its focus, making the light nearly parallel again, and a third mirror, a concave sphere to focus the light mid-way between the second and third mirrors. Paul commented that this arrangement is free of the optical defects of coma and astigmatism, and provided that the focal length is at least four times the diameter of the primary mirror it will also be free of spherical aberration for all practical purposes. sadly, Paul dismissed the design as impractical because the obstruction caused by the second and third mirrors is always at least one-third of the diameter of the primary

James Baker, an American astronomer, independently discovered the design and published a brief description in 1945. More than 20 years later, he published designs of systems of two mirrors that could give the Palomar 200-inch (5.08 metres) telescope a larger field of good definition than could be done with auxiliary lenses, and found that at the focal ratio of f/3.3 of this telescope the secondary mirror should be slightly non-spherical to maintain good image quality. the largest field of view in Baker's designs was 1.1 degrees in diameter, requiring a secondary mirror about 60 inches, and a third mirror 80 inches in diameter, the latter just above the 200-inch mirror, probably supported on a pedestal rising through the 40-inch hole in the primary. none of Baker's designs was built


in 1983 Donald Lynden-Bell urged me to search for an optical system that could give a field of view and optical definition comparable to the Schmidt camera, but with a much shorter tube length. After investigating the two-mirror arrangement proposed by Andre Couder in 1926, and finding that to avoid coma the two mirrors must be separated by twice the equivalent focal length, as in the Schmidt, I was alerted to the three mirror arrangement, by then known as the Paul-Baker system, by a colleague, Ed Kibblewhite

some ray-tracing work soon showed how the field of view could be enlarged beyond Baker's 1.1 degrees. the primary mirror was given a large perforation, and the third mirror placed behind it. The secondary mirror was brought closer to the primary. Soon it appeared that the secondary mirror should be only half way from the primary mirror to its focus, and the third mirror the same distance below it. to obtain a field of view 5 degrees in diameter the diameters of the mirrors were increased (for a constant focal length) until the primary mirror had an aperture of f/1.6, a larger relative aperture than the 8-metre Gemini telescopes then proposed (and later erected at Hawaii and Chile)

at this aperture I found that the primary mirror needed a small departure from a paraboloid, and both second and third mirrors should be non-spherical to obtain the best images over the whole field. the shapes of the mirrors were defined by power series with terms in R^2, R^4, R^6 and R^8; the R^2 terms fix the axial radii of curvature and these were kept constant while the other 9 (three for each mirror) were adjusted. after several days at the computer I found that the ray-theoretical image spread could be made less than 0.1 second of arc over the central degree, and less than one-third of a second of arc over the full five degrees, when allowance was made for some loss of light in the outer parts of the field. the illumination was uniform over a field two degrees in diameter, and vignetting amounted to 40 per cent at the edge of the field

can you explain the mechanism briefly?
well, not briefly, but here goes: if the first two mirrors were both paraboloids the twice-reflected light could be made exactly parallel (an arrangement first described by Marin Mersenne in 1636), the third mirror would then also need to be a paraboloid to focus the light, and the result would be no better than that provided by a single paraboloidal mirror, which gives images suffering (mainly) from the triangular blur of coma. making the second mirror spherical, on the other hand, provides just the same correction as the lens of a Schmidt camera, so that the third mirror can bring the light to a sharply focused image if it, too, is spherical


and you managed to make it work?
in 1985 I made the mirrors for a 'working model' of the design as it was at that time. this had an aperture of 10 cm, a focal length of 16 cm, and a field of view of 4 degrees. the best photograph taken with this showed spiral structure in the Whirlpool nebula, Messier 51, the first spiral galaxy discovered by Lord Rosse in 1845. this photo helped to persuade the Science and Engineering Research Council to make a grant to finance the construction of the prototype of 50 cm aperture in 1986-88

as this was an experimental type of telescope it was important to site it within easy reach of workshops (for alterations and adjustments) and the observer (myself). an existing wooden hut with roll-off roof in the grounds of the Institute of Astronomy at Cambridge provided a low-cost 'dome'. unfortunately IoA Cambridge is situated on Madingley Road, a major traffic route with closely spaced sodium street lights. a red filter (3 mm Schott RG630) has always been used to keep the sodium light out, but it cannot absorb the far red light from domestic lighting, security lights, etc. under these circumstances, and given the small size of the prototype, it is not surprising that it has not made any significant discoveries. the important function of this telescope was to show that the design led to a reliable working telescope, and to encourage funding of telescopes of much larger aperture and the same, or similar, design. indeed, the 8.4 metre Large Synoptic Survey Telescope now under construction in the USA has adopted a design which is a blend of Maurice Paul's original concept and my prototype, modified to use an array of CCDs. these detectors must be kept cold (-30 C) and dry if they are to work well at the low light levels found in astronomy. they are therefore used in a vacuum; the front window of the CCD chamber is comparable in size to the largest Schmidt camera corrector lenses, but to withstand atmospheric pressure it is given the form of a deeply curved meniscus lens


selected works from Sophy Rickett's Objects in the Field (2013) and the Observation series (1991/2013) will feature in dark frame / deep field exhibition at BREESE LITTLE, for which super/collider is media partner

private view: Wednesday 3 June, 6–9 pm curator-led tour: Thursday 18 June, 7pm super/collider event: Wednesday 15 July, 7pm summer party: Thursday 23 July, 6–9 pm