The scientific objectives of multi-object fiber instruments on 8 m class telescopes may significantly depart from those carried out on 4 m class telescopes. Crude anticipation of future programs on these instruments allows to predict that targets will be mainly point sources or of low extension, with magnitudes up to approximately 22. With these characteristics, some observational aspects are investigated: (1) The (blind) positioning error budget constrains the astrometry to within 0.2' rms to keep photon losses at an acceptable level; (2) All-sky atlases from Schmidt plate scans, when available, will provide the relevant information for most programs; (3) Cross- identification or special observing procedures may be required for guide stars if they have a different origin than targets; (4) Atmospheric effects are investigated. Chromatic dispersion is shown impossible to calibrate out. The observation wavelength must be properly related to the acquisition and guiding wavelengths, constraining the use of filters; (5) Other observational constraints like sky subtraction or high spectral stability can lead to specific observational procedures. General observing modes are briefly reviewed. The impacts on the performances, described in terms of `Realizable Multiplex Advantage' (RMA) is described. At best, the RMA can reach one third of the number of fibers.

We describe the conversion of an existing f/8 Cassegrain spectrograph to a floor-mounted spectrograph fed by 94 fibers from the f/5 prime focus of the Shane 3-meter telescope at Lick Observatory. The spectrography forms part of the automated Multi- Object Spectrograph system developed as a collaboration between UCO/Lick Observatory and the Lawrence Livermore National Laboratory. Fibers from a robotic fiber-positioner at prime focus degrade the f/5.5 beam from the telescope (after it has passed through a wide-field prime focus corrector) into roughly a f/4.5 beam. If the 4/8 spectrograph were fed directly with this f/4.5 beam approximately 68% of the light would be lost. A simple optical system has been designed that converts the light from the fibers into the f/ratio expected by the spectrograph. The conversion optics are mounted at the entrance to the spectrograph. We describe focal ratio degradation tests of a variety of optical fibers and the design of the `pseudoslit' which mounts the fibers in a line at the input to the conversion optics.

Fiber-fed spectrographs have now been employed in astronomy for several years. They offer the principle advantages of increasing the number of objects with which simultaneous spectra may be recorded, better stability, an improvement in versatility, and factors associated with the mode-scrambling properties of fibers. However, the nature of these spectrographs pose new questions of certain design parameters. A traditional figure of merit for spectrographs is the throughput-resolution product (A(Omega) R). It is shown that A(Omega) R is no longer a constant as the entrance `slit' is varied. Here, `slit' has a new meaning. In both the first and higher order optics sense, A(Omega) R is investigated and compared with a better figure of merit. This new figure of merit maximizes the potential information content at the spectrograph focal plane. At a minimum, it incorporates the oft neglected instrument response and transmission/efficiency parameters (such as with a grating). Other refinements are made as well. Comparisons may be made both across spectrometer types and as a design guide within a family, where options must be evaluated.

The prime focus of the William Herschel telescope provides a wide field (one degree) for fiber spectroscopy. The production of the fiber feeds, which run from the prime focus to one of the Nasymyth platforms, has required the development of a multiple fiber to fiber connector together with techniques for mounting micro-lenses at the input to the spectrography. The protection of the fibers on the telescope, coupled with careful routing, has led to good focal ratio degradation (FRD) performance. A system for measuring the variation in transmission of the fibers on the telescope has been produced; the technique allows these measurements to be made under normal daytime lighting conditions.

WYFFOS is a spectrograph, currently being assembled, for multi- object spectroscopy using fiber-optic inputs. Initially 126 fibers, later 150 finer fibers (described in a separate paper) will enter it from the prime focus of the 4.2-meter William Herschel Telescope. WYFFOS uses the Baranne white-pupil, Littrow- mount optical system, given a wide 2D filed of view at the grating, high efficiency and a wide range of spectral resolutions. It has a twice-through lenticular collimator and a Schmidt-type cryogenic camera, operating from 350 nm to one micron. We shall install the spectrograph on an optical bench on one of the Nasmyth platforms of the telescope. One advantage of using the optical bench is that the spectrograph's components can be re-arranged, converting the reflection-grating instrument to a transmission-grating instrument. Possible future enhancements include increasing the number of fibers to 300 in the transmission grating mode, and adding a dioptric camera.

The 1.2-m United Kingdom Schmidt Telescope has operated a wide- field spectroscopy service for a number of years. A new common- user system has recently been completed. The instrument, known as FLAIR II, offers many more fiber channels than its predecessors, together with wider spectral coverage and faster turn-round of target fields. This paper describes FLAIR II in its present-day form, highlighting the most innovative aspects of its design, and giving a critical evaluation of its performance. Its impact has been significant, and large data-sets are being accumulated for a variety of object-classes.

The Anglo-Australian Telescope has just received a new prime- focus corrector, designed by Damien Jones (Prime Optics, Qld., Australia) and manufactured by Contraves, USA. The corrector is capable of producing aberration-corrected images over a full 2 degree diameter field of view. It is a 4-element design, the first two elements being rotatable cemented prismatic doublets approaching 1.0 m in diameter, which permit full atmospheric dispersion compensation to high zenith distances. The corrector is at the heart of a new powerful 400-fiber spectroscopic survey facility known as the 2dF. This A$2.3M project represents the largest investment in new technology and instrumentation the AAT has seen in its 20 year lifetime and will provide the telescope with a new and important role for large-scale statistical studies in the latter half of the decade and beyond. This paper will present the performance specifications for the corrector comparing them to the results of the recently complete telescope commissioning test.

In this communication we describe two optical fiber systems contrived to perform simultaneous bidimensional spectroscopy of extended objects: 2D_FIS and HEXAFLEX II. The main component of 2D_FIS is an optical fiber bundle which links the auxiliary Cassegrain focus of the 4.2 m William Herschel Telescope (WHT) with the ISIS spectrograph. HEXAFLEX II has been designed for the Cassegrain focus of the 2.5 m Nordic Optical Telescope (NOT) and it uses the FLEX spectrograph. Both telescopes are at the Observatorio del Roque de los Muchachos, on the island of La Palma. The basic component, the fiber bundle, is similar in both systems although each one is adapted to the particular characteristics of the corresponding spectrograph and telescope. Each one of the bundles is composed of 125 fibers: 95 to observe the object and 30 to obtain the sky background signal. Details of the acquisition procedures are provided. These systems have been tested at the telescope with success. Some examples of results on astronomical objects are shown.

The KPNO fiber-fed, multi-object spectroscopic instrument, Hydra, is being moved from the Mayall to the WIYN telescope. The WIYN telescope has good concentricity between the focal surface radius and the exit pupil. The primary modification to the instrument will allow the fibers to align with the telescope exit pupil while lying along the curved focal surface. Maximum pupil alignment losses will be reduced from 15% at the Mayall to 5% at the WIYN. The new orientation of the instrument requires a different mechanism for accessing the fiber and focal plate for maintenance. We are also upgrading the manner in which the fibers are held in place around the focal plane in order to reduce neighboring fiber interactions beyond the pivot circle. The wavelength calibration assembly is being modified to take advantage of extra room in the instrument. For guiding, instead of using offset guide probes, we are developing an algorithm which will utilize the field orientation probes. The Bench Spectrograph associated with Hydra is also being moved over to the WIYN, however, we do not discuss the spectrograph in this paper.

Sky subtraction with fibres is known to be more difficult than with slits. The scientific objectives relevant to carry out with multifibre instruments, even in the new age of 8-10 m telescopes, mainly deal with objects of magnitude <EQ 22. The requirement in beam switching methods to reach the ultimate performances is controversial. In this paper, we present the results of sky subtraction tests that have been performed on the 3.6 m ESO telescope. The two basic points we wanted to address were the ability of fibres to allow sky subtraction up to the 22nd magnitude, and whether or not beam switching is required. We detail the data reduction steps we have employed. We define all the possible techniques that can be used for sky subtraction and give the corresponding expressions for spectrum extraction. Last, we present the results we have obtained, and conclude that sky subtraction with fibres is quite feasible, even on 4 m telescopes, up to b(subscript $japproximately equals 22 without resorting to beam switching techniques.

A 640 fiber feed for a pair of moderate resolution spectrographs is being constructed for the sloan digital sky survey. The spectrograph slits, optical fibers and plug-plates are mounted together in removable cartridges. The cartridges protect the optical fibers and provide quick-change interfaces to the telescope and the spectrographs. Ten cartridges will be made to allow plugging to occur during the day. The fibers are constructed in harnesses of 20. These are mass-terminated in v- groove blocks at the spectrograph slit. At the input end the fibers are terminated in simple ferrules and manually plugged into aluminum plug-plates. The plug-plates are deformed to match the best focal surface of the telescope, and the holes are drilled to be parallel to the principal ray after bending. We describe a simple technique for drilling accurate holes.

'FUEGOS' (Fibre Unit for European General Optical Spectroscopy) is one of the foreseen focal instruments for the future ESO VLT (very large telescope). The goal is to provide the european community with first class equipment allowing high and low resolution spectroscopy for stellar and extragalactic observations in a large field. This large field of view, 30 arcmin provided by the VLT, is quite unique on 8 m class telescopes. The instrument is made of two parts: a fibre positioner attached to the Nasmyth rotator and a spectrograph sitting on the Nasmyth platform. The two assemblies are linked together with optical fibres. The instrument will include two different modes: the MEDUSA mode for the multi-object spectroscopy and the ARGUS mode for the 2D spectroscopy for extended objects. The multi-object positioner is of the robot type, based on a design using two gripper heads travelling simultaneously on a X-Y carriage. It locks magnetically 80 independent optical fibre bundles at object coordinates on a metallic focal surface. 2D spectroscopy is provided through a fibre anamorphoser in three different fields and spatial samplings. The spectrograph is of the 'white pupil' type, based on a catadioptric design with spherical optics throughout. It gives access to spectral resolutions ranging from 1500 to 30000. The detector is a 2048 x 2048 CCD located inside the camera optics.

We are currently testing an automated fibre positioner for the 4.2 m William Herschel Telescope (WHT). This instrument, known as Autofib-2, operates at the prime focus where it is able to utilize the full 1 degree field provided by the prime focus corrector (PFC). The robotic positioner is able to place 160 optical fibres in the focal plane of the WHT which feeds the light to a dedicated spectrograph (WYFFOS) located on the Nasmyth platform. This paper contains a description of the instrument which highlights the new techniques demanded by the prime focus plate scale and the scale distortions due to the PFC and its atmospheric dispersion compensator. These include robot vision to help achieve the high positioning accuracy and the use of two sky viewing probes to accurately determine the time dependent transformation from celestial coordinates to instrumental Cartesian coordinates. Also presented are the initial technical results on the performance of the instrument and the operational results of particular interest to the astronomical observer.

In this paper we describe the impact that instrument design can have on the overall performance of the Gemini Telescopes. We discuss the interaction between the instruments and some of the other telescope facilities such as A&G and Adaptive Optics. We also briefly describe the instruments that have been proposed for delivery during the construction phase of the project and list some of their features.

The instrumentation plans for the 8 m Japan National Large Telescope (approximately equals SUBARU) are reviewed. The topics described in this article are (1) the interface conditions between the telescope and the observational instruments at each focus, (2) Cassegrain auxiliary optics layout as an example to show the focal area layout, (3) scopes for optical and infrared detectors, (4) the outline of proposals for observational instruments, (5) the domains in the manifold of observational parameters covered by these instruments, and (6) the schedule of construction of instruments.

This article describes the design and construction of a CCD camera for the f/9 Cassegrain focus of the 200-inch (5m) Hale Telescope at Palomar Observatory. This instrument utilizes a set of 18-inch (0.45m) diameter lenses to change the Cassegrain focal ratio from f/16 to f/9, and the optical design prescription of these lenses is reported. For CCDs having 24 micrometers pixels, the f/9 focus gives a reciprocal plate scale of 9 pixels per arcsecond, which is well suited to precision photometry as well as future application to adaptive optics. In addition to the CCD and shutter, the instrument also incorporates an offset guider and an 8-position filter wheel.

The stellar coronograph built by STScl and the OATO operating at the SUSI focus of ESO's New Technology Telescope (NTT), described with particular emphasis on its optical and mechanical characteristics, is one of the most powerful coronographs in operation. Observations with angular resolutions of 0.6 arcsec over field of view of 45 arcsec diameter have been routinely achieved in the observing runs carried out so far. We describe the results from the application of the coronograph to the problem of observing the inner regions in the (beta) Pictoris dust disk.

The Low Resolution Imaging Spectrometer is designed for use at the Cassegrain focus of the Keck 10-m telescope. It provides the capability of acquiring low resolution (R equals 1000 to 5000) digital spectra, as well as 6 X 8 arc-minute moderately high spatial resolution (4.65 pixels/arc-second) direct images. Spectroscopy can be carried out with single slits which are 3 arc-minutes long. In addition punched multi-slits can also be employed which allow for the acquisition of at least forty spectra simultaneously. Since the instrument is designed to be as efficient as possible, it is a double spectrograph, with a dichroic splitting the blue and red light into separate optical paths after the collimator. Only the red side has been constructed thus far. With a 2048 by 2048 thinned Tektronix CCD as the detector the total efficiency of the red side at the peak of the grating blaze is predicted to be nearly 40%. Results of the commissioning observing runs will be described.

An infrared camera has been successfully coupled to the Fourier Transform Spectrometer (FTS) at the Canada-France-Hawaii Telescope (CFHT), making a unique infrared spectrometer. This instrument, the only one of its kind in the world, is capable of simultaneously acquiring both images and spectra of astronomical sources in the 1.0 to 2.5 micrometers wavelength range across its 24' field of view. Successful observations have been made of various targets, including planets and emission nebulae. A description of the instrument, data analysis, and scientific results, is presented below.

A specially designed faint object spectrograph in the near-IR region from 1 to 2 micrometers is proposed for the Japanese National Large Telescope: SUBARU. The proposed instrument called OHS for SUBARU is kind of a pre-optics system capable of eliminating most of intense OH airglow emission lines from the incident beam in the J- and H-passbands. The detectivity for objects in the faintest end is supposedly enhanced with this spectroscopic filter system by removing nearly 95% of the natural sky background: the non-thermal night airglow emission. The sensitivity gain in terms of limiting magnitude in these wavelength bands is expected to be 1 to 1.5 mag, depending on the modes of observations. The expected performance of the prototype OHS when attached to SUBARU will also be presented.

The requirements for medium/low resolution optical spectroscopy with the GEMINI telescopes are discussed. Two different modes have been identified to exploit the exceptional image quality expected from the telescope and its Adaptive Optics (AO) system while also providing a wide range of spectral resolution options over a larger field. Meeting these conflicting requirements within a single instrument poses a major challenge, and it may prove necessary to construct two devices.

The design and the expected performance of the FOcal Reducer/low- dispersion Spectrograph (FORS) for the Very Large Telescope (VLT) of the European Southern Observatory (ESO) are described. FORS is an all dioptric focal reducer designed for direct imaging, low- dispersion multi-object spectroscopy (MOS), imaging polarimetry and spectro-polarimetry of faint objects in the spectral range from 330 nm to 1.1 (mu) $m. The field of view is 6.83 X 6.83 arcmin² with an image scale of 0.2 arcsec/pixel. To take full advantage of excellent seeing, a configuration with an image scale of 0.1 arcsec/pixel can be selected. In the MOS mode, simultaneous spectroscopy of up to 19 objects is possible. Two copies of the instrument (denoted `FORS I' and `FORS II', respectively) are under construction and will be installed at the Cassegrain foci of the VLT unit telescope Nos. 1 and 3.

The MOF is a compact device (15 cm max dimension) displaying a very narrow-band spectral transmission (20 to 70 m$angstrom, single or multiple bands) with intrinsic absolute spectral reference and stability, high peak transmission (close to 50%) and imaging capability. It has been used so far mainly in solar astronomy to detect solar magnetic and velocity fields. Doppler shifts as low as 1 cm/sec equivalent velocity can be measured on the sun (whole disk) using 2 cm aperture telescope and 1 sec exposure time. Its working principle is based on magneto-optical effects on metallic vapors in a magnetic field. An overview of the theoretical model compared with experimental spectroscopic measurements is discussed in the first part. Applications in astronomy, namely helio and astero seismology and magnetographic observations are shown in the second part of this paper.

A multimode spectrograph for tridimensional spectroscopic observations of faint extended objects is under development. The spectrograph has four modes, i.e., the imager mode with narrow (or wide) band filters, the long-slit spectrograph mode for area spectroscopy by means of spacial scans, the Fabry-Perot interferometer imager mode, and the microlens-array (MLA) spectrograph mode. Selection of the mode is made by moving optical parts relevantly in the spectrograph. The principles of the optics of the first three modes are ordinary ones. In the imager mode, a field lens is inserted behind the telescope focus and filters are set between the collimator and the camera lenses, while a Fabry-Perot etalon is inserted instead of the filters in the Fabry-Perot imager mode. In the long-slit mode, a slit is set at the telescope focus, and a grism replaces the filters or the etalon. The principle of the MLA spectrograph mode is essentially same as that of TIGER for the Canada-France-Hawaii telescope. Our spectrograph, however, is equipped with a dual-channel preoptics system for the MLA, one channel acquiring a target object and the other the neibouring sky appreciabily apart from the target object. The image of the sky is projected on the microlenses along a side of the array of the microlenses. The spectra of the sky are used to eliminate of contamination by sky background from spectra of the target object. The design of this spectrograph is primarily optimized for the MLA spectrograph mode at the 1.88 m telescope at Okayama. With this telescope, about eighty spectra of an area about 9# x 15 of an object and fourteen spectra of the nearby sky are obtained with a spacial resolution of 1.3" which matches the average seeing size at the telescope site. Performances of this spectrograph when it is used for other telescopes, including the SUBARU, are also presented.

The VIMS-V is a passive remote sensing instrument under development for the Italian Space Agency to perform high resolution spectral imaging in the optical waveband. Its unique design relies substantially on conventional materials and fabrication techniques to obtain high sensitivity and versatility in a compact, lightweight, and low power instrument suitable for deep space missions. The prototype will fly on the Cassini spacecraft and operate in unison with the Jet Propulsion Laboratory's IR channel to map the surface spatial distribution of mineral and chemical species of Saturn, its rings, and its satellites.

A fundamental problem of adaptive optics systems is the very narrow corrected field-of-view (FOV) which can be obtained due to the fact that turbulence is extended in altitude throughout the atmosphere. The correctable FOV is on the order of 5 to 10 microradians in the visible. Previous concepts to broaden the corrected FOV have been hardware-oriented, requiring multiple wavefront sensors (WFS) and multiple deformable mirrors. In this paper we analyze the average performance of an image measurement and post processing technique which uses simultaneous measurements of a short exposure compensated image measured in an off-axis direction, and an additional WFS measurement taken in the off-axis direction. Results are presented for infinite altitude WFS beacons driving both the WFS for the adaptive optics and the WFS looking in the off-axis direction, a variety of seeing and WFS light level conditions, and off-axis angles from two to six times the isoplanatic angle. We find that this technique provides a significant boost in the average effective transfer function out to at least six times the isoplanatic angle.

The optical design of a general-purpose 1 to 5 micrometers cryogenic IR camera and spectrograph (IRCS) for the 8.2-m Subaru telescope is described. The camera section serves the essential purpose of a slit-viewer in order to permit efficient use of the spectrograph on faint objects. It will also serve as a multipurpose IR camera. The spectrograph section will have a resolving power of (lambda) /(Delta) (lambda) equals 660 to 1600. 1 to 2.5 micrometers or 3 to 5 micrometers will be observed in a single exposure by using gratings and cross-dispersing prism combinations. The slit length will be 3 to 5'. The camera section will have 3 pixel scales (0'.030, 0'.056, and 0'.125) that provide high spatial imaging, 1:1 imaging (high throughput), and `wide-field' (about 2' X 2'). The spectrograph section will have 2 pixel scales: 0'.05/pixel and 0'.125/pixel. The important features of the IRCS are: (1) Two pixel scales are available, one matched to the tip-tilt secondary and the other matched to the adaptive optics system. (2) Switching between imaging and spectroscopic modes is possible. Therefore observational programs can be optimized for the seeing, availability of guide stars, and weather conditions. (3) In some cases deep imaging can be undertaken while long exposures are made in the spectroscopic mode.

Following the availability on the market of IR arrays able to perform ground-based astronomical observations in the atmospheric windows longward of 2.5 micrometers , we started at the Torino Astronomical Observatory a new project aimed at the construction of a thermal IR camera to be installed at the TIRGO telescope. Located in the Swiss alps at 3100 m a.s.l., this Italian facility (1.5 m f/20 IR optimized) provides during the winter months a relatively high number of nights (20%) with first-rate conditions (low temperature and emissivity) for medium-IR observations. In order to fully exploit this potential, we designed an instrument, named Two-Channel Medium IR Camera (TC-MIRC) operating over the entire 1 to 14 micrometers region and optimized for the 2.5 to 14 micrometers thermal bands. TC-MIRC covers such a broad range of wavelengths using two IR array detectors: an InSb device for the 1 to 5 micrometers region and a Si:Ga device for the 8 to 14 micrometers band. The main characteristic of the camera is that both arrays can simultaneously observe the same region. In this way, we can not only approach an almost double observing efficiency (time really spent `on target'), but also use both detectors for correlated observations and testing of unconventional acquisition techniques. It follows that TC-MIRC is a complex instrument presenting several interesting features. Thanks to the presence of two independent optical channels, the user can change the filters, adjust the focus and vary the optical scale on a channel without affecting the acquisition running on the other array. In particular, the possibility of adjusting the optical scale on each detector during the observations allows the user to select the most convenient sampling and field of view on the basis of the actual seeing or diffraction conditions, background level and scientific needs. Moreover, the adopted cryogenic system is entirely based on a mechanical closed-cycle cooler and allows very low-cost operations and easy maintenance on the mountain for long periods of time. In this paper we present the opto/mechanical design and construction, the general structure of the control system and the software architecture. We report the performances reached by the most critical parts of the instrument during the test carried out in the laboratory and at the telescope during the first engineering run in January 1994.

This paper describes the performance of a unique new IR array camera system which provides simultaneous imaging at two wavelengths in the near IR. Two-color imaging is achieved with a dichroic beam splitter which yields two independent beams, one short-wave (SW) from 1 to 2.5 micrometers and one long-wave (LW) from 2 to 5 micrometers . A Rockwell NICMOS 3 256 by 256 HgCdTe array is used in the SW channel and the LW channel has an InSb 256 by 256 array from SBRC. The instrument, which is designed for the University of California's Lick Observatory 3-m telescope and for the f/15 focus of the 10-m W.M. Keck telescope, employs a closed cycle refrigerator and a compact array control/data acquisition system based on transputers with a host 486 PC. On the Lick 3-m telescope the pixel size is 0.7' which gives a field of view of about 180' by 180'. Facilities are also provided for spectroscopy and polarimetry. Recent observational results are reported to illustrate the performance of this system.

We present data on the image quality achieved with the near IR array spectrometer cooled grating spectrometer 4 (CGS4) on the UK IR telescope (UKIRT) on Mauna Kea. A design spot size of 30 micrometers was specified for CGS4, to maintain acceptable image quality with both the 58 by 62 pixel array with which it is currently equipped and the 256 by 256 array which CGS4 was also designed to accommodate. Details are given of the design, construction and alignment method which allow linear tolerances of 50 micrometers and angular tolerances of 25 mrad to be met and maintained at cryogenic temperatures. The instrumental flexure is also discussed. Both laboratory spectra and those taken at the telescope illustrate that design spot sizes of 30 micrometers have been achieved in the near IR. It will be demonstrated that the theoretical resolution of the instrument is attained for resolving powers from approximately 200 to 20,000.

We describe a slitless spectrograph designed for use in the IR region between 2.6 and 5.2 micrometers . The dispersing element is a grism fabricated in silicon by binary-optical techniques. This approach permits the incorporation of aberration correction into the grating element. When combined with simple, all-silicon field and camera optics, the grism forms zero-order images and 3.5 mm- long dispersed spectra on a 256 by 256 element array of 38 micrometers InSb detectors. Interchangeable field lenses provide for 8:1, 6:1, and 4:1 reduction to a final focal ratio of f/2.5.

We present a description of the mid-IR camera system (MIRCAM) currently under development at the University of California, Irvine. MIRCAM is a portable imaging system, specifically designed for diffraction-limited, broadband imaging in the 4.7 micrometers (M band), 10 micrometers (N band), and 20 micrometers (Q band) atmospheric windows. MIRCAM is designed for operation with the new generation of high-background 256 by 256 Si:As IBC arrays currently under development by both Rockwell International and Hughes Aircraft. However, initial operation will proceed with a 128 by 128 Si:As IBC array manufactured by Rockwell International. We have also designed a state-of-the-art control and data-acquisition system that is capable of reading out the 256 by 256 array at a rate of 500 frames per second, or 2000 frames per second for the 128 by 128 array. The fully-reflective optics are designed for Nyquist-sampling at the telescope diffraction limit, and exhibit spot sizes that are considerably smaller than the pixel size over the entire 256 by 256 array. Four interchangeable optical configurations permit use of MIRCAM on the Lick 1-m, IRTF 3-m, UKIRT 4-m, and Keck 10-m telescopes.

The MegaCam is a multi-purpose, wide-field, two-color camera being designed for use at the UH 2.2m telescope. The camera will utilize a Rockwell 1024 x 1024 HgCdTe detector array for 1-2.5 micrometers imaging, and a 2048 x 4096 frame-store CCD (2048 x 2048 active area) for optical imaging. The optics are based on a modified Offner relay design with additional lenses to give a 2:1 magnification in the infrared channel for a field of view of 5'.8 x 5'.8 (0.34 arcsec/pixel) using the f/10 telescope secondary, or a scale of 1'.9 x 1'.9 (0.15 arcsec/pixel) at f/31. This design provides a simple, high-throughput, and compact optical layout. A beamsplitter is placed in front of the IR optics at a low angle of incidence to form the optical image at 1:1 magnification on the CCD, for a field of view of 4'.7 x 4'.7 and 1'.5 x 1'.5 at f/10 and f/31, respectively. The optics and filters are to be housed in a LN₂-cooled dewar. The CCD and IR arrays will be operated with modified SDSU-design controllers. The user interface will have several modes to make simultaneous optical/IR imaging simple to configure and perform at the telescope.

We describe an IR multi-fibre system we are currently building for UKIRT. SMIRFS (Spectroscopic multiple infrared fibre system) will feed the CGS4 spectrograph, providing it with a multi-object capability in the J, H and K bands. For efficient coupling the inputs and outputs of the fibres are microlensed. Because the telescope is normally focused inside CGS4 for conventional spectroscopy the SMIRFS fibre slit is projected in to CGS4 by a spherical mirror. IR optimised Silica fibres are used for the J and H bands and Zirconium Fluoride fibres are used for the K band.

This paper describes the design of an IR cold coronagraph (CoCo) built by SETS Technology, Inc., for use at the NASA 3 m IR Telescope Facility (IRTF) at Mauna Kea Observatory, for the imaging of faint IR sources in proximity to bright sources. The coronagraph is designed to obtain high contrast photometric images by use of an occulting mask and a pupil mask. The coronagraph is to be used in combination with the IRTF NSFCAM, which covers 1-5 micrometers and uses a 256x256 InSb array. The platescale can be varied from 0.06'/pixel to 0.15'/pixel, covering a field of view of 14' and 38', respectively. Selectable apodized and hard occulting masks are mounted on a wheel as the first element in the system to reduce scattered light. Selectable pupil masks are cooled to 77K within the CoCo cryostat. The cryostat consists of a liquid nitrogen can for cooling the optics, masks, and baffles. The CoCo dewar is mounted on a slide in a housing to allow it to move out of the beam path so that the NSFCAM may be used with or without the coronagraph during the same observing period.

The design of a multichannel occultation photometer built under NASA contract to SETS Technology, Inc., for the NASA 3-m IR telescope facility (IRTF) and the JPL Table Mountain telescope is described. This instrument acquires data in four selectable passbands (two 1 to 5 micrometers channels and two 10 to 20 micrometers channels), with very high sensitivity and approximately 100% duty cycle on-source during chopping. The optics are optimized for uniform response across an aperture of up to 20 arcseconds on the IRTF. The cryogenic system is a two-can cryostat with one liquid nitrogen can for cooling the radiation shields, optics, filters, and baffles, and a liquid helium can for cooling the IR detectors. The instrument operates two types of IR detector technologies. The 1 to 5 micrometers detectors are low-capacitance, single-element InSb detectors. The 10 micrometers detectors are blocked impurity band detectors. The instrument also has a 64 by 64 visible CCD array as an additional channel for guiding and visible photometry. A global positioning system unit is incorporated into the system for time and location stamping of occultation events. The instrument design and construction are discussed.

We constructed a near IR camera PICNIC equipped with the 2D array detector NICMOS-3 manufactured by Rockwell. Employing a four aspherical metal mirror system for reimaging optics, PICNIC is an IR imager distinguished with high throughput and small chromatic aberration. In addition to normal imaging mode, the camera has a prism spectroscopy mode and a polarimetry mode. A direct view prism, which can be optionally inserted in front of the camera, make it possible to obtain multiobject low resolution spectroscopy in the wavelength range from 1 to 2.5 micrometers at resolving power of approximately 50. Linear polarization can be measured with combination of the filters and/or the prism. We realized user-friendly operating environment as well as automated observation.

We have just recently commissioned a new 1.0-5.5 micrometers IR array camera for the NASA IR Telescope Facility based upon the Santa Barbara Research Center 256x256 InSb array. The primary features of this new instrument are three user-selected platescales, a variety fo fixed bandpass filters, 1 to 2% spectral resolution circular variable filters, coronagraph masks, polarization imaging capability, an optical guider/imager, and a grism. In this paper we briefly outline the design and performance of the camera system, describe some unique operating modes, and show some recent images.

The present state of the art in scientific InSb focal plane arrays is defined by the Santa Barbara Research Center (SBRC) 256x256 device. In this paper we will discuss the status of the USNO-NOAO 1024x1024 focal plane development effort at SBRC. The project is in its ealy phases, and so this report consists of the device description, design goals, testing plans, and schedule. Drivers for even larger IR arrays include plans for large (8 meter class) telescopes, high resolution spectroscopy, astrometry of large fields, and wide field imaging amongst others. Focal planes of 640x480 have already been demonstrated, and so the step up to 124x1024 seems reasonable at this time.

We report on the development, fabrication and performance of a new class of thermal sensors for far IR and millimeter wave detection. These devices consist of small single crystal samples of ultra-pure, natural or isotopically engineered germanium which have been doped by the neutron transmutation doping (NTD) technique. The concentrations of the acceptor and donor dopants (N_A,N_D) can be accurately controlled with this technique. They depend on the thermal neutron fluence, the neutron absorption cross sections and the atomic fractions of ⁷⁰Ge (for the Ga acceptors) and ⁷⁴Ge (for the As donors), respectively. The values of N_A and N_D and their ratio result in a predictable resistivity of the Ge crystals down to temperatures of a few milliKelvin. The excellent control of the resistivity down to very low temperatrues, together with the development of ohmic contacts working at the lowest temperatures, allows the fabrication of high sensitivity bolometer arrays with over 100 pixels and highly uniform response.

We describe the design and manufacture of SCUBA, which is undergoing laboratory testing prior to commissioning on the James Clerk Maxwell Telescope on Mauna Kea, Hawaii. It contains two arrays, one of 91 pixels optimised for 450 micrometers and the second of 37 pixels optimised for 850 micrometers in close-packed arrays, with each pixel having diffraction-limited angular resolution. Some of the original design features of the instrument are described: the cryogenic system operating at 100mK; the optical layout; bolometer manufacture; and array integration. We illustrate the performance of the instrument with test results obtained during the laboratory commissioning.

Over the past 6 years the instrumentation team at the NASA IR Telescope Facility (IRTF) has designed and built 3 major facility instruments, two cameras and a spectrometer, for use by the IR astronomical community. We will report on many new techniques for cryostat construction that have been developed that deviate from traditional practices. These include aluminum electron beam welded vacuum and cryogenic enclosures, rectangular cryostat formats, use of closed cycle coolers and their performance, use of all aluminum structures, optimization of cryogenic performance without super insulation, activated charcoal getters, and optomechanical mounting. Cryostat performance data and methods for estimating cryostat performance will also be included.

The University of Hawaii and the Rockwell International Science Center are developing a large format SWIR detector array optimized for low background astronomical imaging and spectroscopic observations. This so called HgCdTe astronomical wide area IR imager (HAWAII) device will be based on the technology developed for the NICMOS project, but will incorporate several modifications of this design to improve the performance.

The signal to noise ratio of very faint object spectroscopy is discussed in the context of the gain in limiting magnitude attainable on very large telescopes at low spectral resolution. It is shown that if the multiplicative errors resulting from the sky subtraction are not corrected, deeper spectroscopy will not be obtained by increasing the telescope size. A new CCD observing technique, which we call `Va et Vient' spectroscopy, is described in detail. This is a beam switching method where the elementary exposures are short enough to freeze the sky temporal fluctuations, and where the CCD is read only after a series of a few dozen of exposures. During the integration, the signals from the object field and the reference sky are separated by shifting the charges onto the CCD along the columns. Analytical evaluations are confirmed by laboratory tests and show that this technique is photon noise limited even at high flux and thus can lead to deeper spectroscopy. We give estimates of the results that can be achieved with the new generation of very large telescopes.

The enhancement of the quantum efficiency of charge-coupled devices (CCDs) can be accomplished by several methods. The most dramatic increase comes from thinning the CCD for use in the back illuminated mode. Techniques to improve the QE of back illuminated CCDs include the deposition of backside thin film coatings to reduce reflection losses, surface charging to eliminate the backside potential well, and several aspects of device packaging. We have developed processes in these areas which have led to the post-manufacture optimization of devices for astronomical observations. WE describe our backside oxidation results and the interaction of the oxide with backside charging, the effect of ion absorption on backside charging, the development of one and two-layer antireflection coatings, and a new packaging method to improve near-infrared quantum efficiency.

While large area scientific CCD imagers have high resolution and sensitivity in the visible region, the UV and near IR wavelength regions typically have a decreased response. Therefore, modified designs and advanced techniques are used to broaden the range for high quantum efficiency. For the UV band, deposition of a very thin polysilicon gate allows transmittance of shorter wavelength photons, and frontside coating a completed device with lumogen absorbs lower wavelength photons and re-emits them near the peak sensitivity of the CCD. For the near IR band, high-resisitivity, thick epitaxial material is used for collection of longer wavelength photons. Overall broad spectral band QE is enhanced with backside thinning and antireflective coatings of various dielectric layers. QE has been predicted by calculating the transmission, reflection, and absorption over a cross section of the CCD. Data is presented to confirm the QE increase from 120 to 1000 nm. Design and fabrication are discussed for each improvement technique, as well as trade-offs in cost, device performance, and yield.

Results are presented on the fabrication and characterization of a 4Kx2K three-side buttable CCD produced by Orbit Semiconductor, a silicon foundry in San Jose, California. This first run of wafers was produced to test the ability of Orbit to produce high quality scientific CCDs with the characteristics required for detectors to be used in optical instruments of the Keck Observatory. Also on the wafer are two 2Kx2K devices. Similar devices have been fabricated for us by Loral/Fairchild. Extensive characterization of the Loral devices has taken place over the past few years, so interest is high about the possibility that Orbit might become a second source for similar detectors. This paper presents the first results on the 4Kx2K CCDs, and those preliminary results include measurements of charge transfer efficiency, low-temperature dark current, on-chip amplifier read- out noise, localized charge traps, full well, and responsive quantum efficiency.

We describe the design of an ultra-format, 8192 x 8192 pixel CCD mosaic imager under construction for the Mauna Kea Observatory. The mosaic will be built from a 4 x 2 array of 3-edge-buttable 2048 x 4096 15 micrometers pixel CCDs fabricated by Loral Fairchild. We outline the scientific justification for such a mosaic and the technical details of the 3-edge-buttable 2048 x 4096 CCD design. We also present our strategy for imager packaging and subsequent mosaic construction which will result in a mosaic with individual elements that can easily be installed and removed. This particular CCD mosaic is intended for two existing MKO telescopes: the UH 2.2 m and the CFHT 3.6m. In either configuration, the imager will offer an enormous field of view with excellent spatial sampling.

As astonomical CCD image sensors have incorporated more pixels the total readout time of the arrays has increased to the point where it can significatly reduce overall available observing time. The incorporation of multiple readout circuits on CCDs and multiple readout capability in their associated controllers allow several pixels to be read out simultaneously, reducing the overall readout time of the array. A controller is used to operate a single CCD containing four readout circuits that operate simultaneously, giving a factor of three reduction in readout time, and a small increase in readout noise. Another tactic for reducing readout time is described, that is easily implemented with programmable CCD controllers that reads out only selected regions of interest and discards the uninteresting portions of the image, giving a considerable decrease in readout time and no readout noise penalty.

University College London and The Imperial College of Science, Technology & Medicine are together developing a new large area imaging photon counting system, BIGMIC, for use primarily on very large telescopes. This detector is designed for applications requiring the highest sensitivity and resolution such as, high dispersion Echelle spectroscopy. The system incorporates a specially designed 75mm active diameter image intensifier fibre- optically coupled to a fast scanning 770 x 576 pixel frame transfer CCD. Photon events at the intensifier output screen are centroided to 1/8 of a CCD pixel in both X and Y in order to sample adequately the point spread function of the intensifier, giving a data acquisition format of 6160 x 4608 sub-pixles. The area imaged onto the CCD is 61mm x 46mm with a data acquisition pixel size of 10 micrometers . A hardware windowing facility, built into the detector system, enables the astronomer to select a subset of the imaging area for data acquisition. This permits the user to match the system to specific applications; for example in a number of spectroscopic applications an essentially one dimensional image is required and a detector format of perhaps 6160 x 100 pixels could be utilized.

CCDs are widely used in astronomy for spectroscopy, photometry, astrometry, and many other applications. This array sensor has a periodic structure which defines individual picture elements (pixels). The stability and excellent overall response of this type of detector are well known. Less widely described is the fact that the internal structure of the array gives rise to intrapixel variations in responsivity. These response modulations are particularly relevant when the data is spatially undersampled; a situation that is not entirely uncommon in certain instruments. We have made detailed measurements of the response variations within CCDs, using an experimental arrangement that gave a 2 micron resolution. Optical response has been determined at all sample points within a pixel, at selected wavelengths in the range 400 to 900 nm. Measurements are presented for front-illuminated thick (EEV) devices and backside- illuminated thin (Tek) CCDs. Models of the internal structure have been constructed and used to calculate theoretical response data; these have been compared with the experimental results. An example of an extracted (FOS) spectrum which demonstrates these undersampling effects is discussed.

We describe an image intensifier tube of flexible design, capable of being used for a number of different ground and space-based applications. The tube was originally designed to meet a requirement for an optical photon-counting detector sensitive between 150 and 650 nm with a spatial resolution of less than 20 micrometers . Other performance criteria aimed for were a peak quantum efficiency of greater than 20% and a 200 kHz maximum count rate. The tube has a 25 mm diameter image format which is nominally 2D, though 1D and non-Cartesian imaging formats could be accommodated. Photon detection is via a semitransparent photocathode coated on the detector input window. Liberated photoelectrons are proximity focused on to a microchannel plate (MCP) stack which provides electron gain with a saturated pulse height distribution for photon counting. Imaging is accomplished with conductive charge division readout. The detector geometry is easily modified to accommodate different types of this generic readout type. The imaging electronics consist of several parallel charge measurement channels, the number depending on the readout type. A decoding algorithm provides an x,y photon event coordinate. The detector was manufactured by DEP bv, Holland. The original design utilized an S20 photocathode on a magnesium fluoride or fused silica window but recently detectors have been made with calcium fluoride windows and caesium telluride photocathodes. All detector variants built thus far use a stack of three MCPs in a Z configuration producing an electron gain of between 10⁶ and 2 X 10⁷ depending on MCP conditioning. We present images and performance data from the two types of detector so far built; the original optical intensifier design with a SPAN six electrode image readout, and a variant with a UV photocathode and wedge and strip anode readout. We discuss the problems encountered during the tube development and the solutions arrived at for each type of device.

Astronomers responsible for the planning of astronomical observations are showing an interest in scheduling technology. This interest is related to the increasing cost of acquiring high-quality astronomical data, in terms of the technology of contemporary observing methods, and the human expertise needed to obtain successful results using that technology. Telescope time has always been, and is likely to remain, a scarce resource. Tactical planning to make the best use of the limited time available is now essential, and tools are beginning to appear which help astronomers schedule a set of observations over a number of sequential nights. Such tools are of use to duty astronomers making service observations on behalf of colleagues, as well as helping visiting astronomers make the best use of a short period of stay. The majority of planning programs, such as those inspired by the Hubble Space Telescope, choose to address the longterm scheduling problems rather than offer a shortterm solution that can be modified interactively during the course of observations. The constraints embodied in the development, in conjunction with staff at the Royal Greenwich Observatory, of one such program -- NightWatch are described.

Data management, acquisition, and communication subsystems (DMACS), an assortment of computerized subsystems, implements Local Area Network architecture to simplify hardware and software interfaces between modular subsystems. The subsystems consist of workstations and embedded processors for data processing and control of an airborne observatory. This paper describes the architecture and subsystems of the DMACS for the proposed Stratospheric Observatory for IR Astronomy (SOFIA), a successor to the Kuiper Airborne Observatory developed and operated by NASA at Ames Research Center. In addition, this paper presents a brief overview of the SOFIA objectives.

The Sloan Digital Sky Survey will image (Pi) steradians about the north galactic cap in five filters, and acquire one million spectra using a dedicated 2.5 m telescope at the Apache Point Observatory in New Mexico. We describe the data acquisition system for the survey's three main detectors: an imaging camera, mounting 54 Tektronix charge-coupled devices (CCD); a pair of spectrographs, each mounting a pair of CCDs; and a smaller monitor telescope camera. We describe the system's hardware and software architecture, and relate it to the survey's special requirements for high reliability and need to understand its instrumentation in order to produce a consistent survey over a five year period.

The first part of this paper describes the microcomputer software that controls the individual mechanisms of the Autofib-2 instrument. All previous microprocessor systems built for the William Herschel Telescope used a variant of the Forth language to implement the control system; however, Autofib-2 is the first instrument to be controlled using a UNIX-style real-time operating system. Facets of the implementation will be discussed, including the reengineering of the interfaces in OS/9 C to allow connection to the existing interinstrument communications systems (Utility Network). The algorithms needed to successfully configure 160 fibers taking into account prime focus corrector (PFC) distortions are also discussed. The second part of this paper describes the astronomer-level control system running on the VAX System Computer. Details will be given of the way in which the astronomer's target field is defined and converted into the configuration data to be sent to the microprocessor system, including the mapping of the PFC distortions and the acquisition of new target fields.

To manage the increased size of focal plane arrays (FPA) and the accompanying data volume increase, we have developed a detector control system that allows parallel readout of multiple array amplifiers. This system consists of a data acquisition computer running the Pegasus Software System on a UNIX platform, an interface control computer and a digital signal processor (DSP) to directly control the array. At the heart of the system is a DSP-based controller (Leach SDSU) which serves as a waveform generator and video processor. It is completely software programmable, making it quite adaptable to a host of FPA designs. We have configured our system into a four channel system, reading pixel data out of all four quadrants of an array. Image data is descrambled in the interface computer, producing a sensible image on the image display. We have applied this system to control and readout of large format CCDs and to IR arrays.

An overview of the IRTF's instrumentation software will be presented. We have developed an array data acquisition system based on a PC/AT using 96002 DSPs to clock and readout the array electronics. This PC, commonly known as the Instrument Control (IC) runs LynxOS, a real-time UNIX. Other components of the system include an X User's Interface which provides a friendly GUI interface to the IC, and View FITS (VF) which is the IRTF's quick look data viewer. These components execute remotely on a SUN workstation. The communications between each application is based on ethernet.

The availability of 256 by 256 pixel IR focal plane arrays and the current development of (larger) 1024 by 1024 pixel arrays require the development of control electronics that are flexible and scalable to different format arrays with sufficient speed and precision to perform clocking, readout and real-time processing of array signal outputs. This paper details a design proposal for an array controller system based on multiple digital signal processors. Derivation of performance specifications is presented in addition to the identification of key support functions and special processing modes. Design tradeoffs between various processors, memory, data export, data storage, and display are discussed. The basis for this design is the existing NSFCAM array control electronics which is being incorporated into the CSHELL spectrometer and guider systems currently in use at the NASA IRTF in addition to instruments at other institutions.

We describe the ongoing modernization of the NASA IR Telescope Facility Telescope Control System. A major mandate of this project is to keep the telescope available for observations throughout. Therefore, we have developed an incremental plan that will allow us to replace components of the software and hardware without shutting down the system. The current system, running under FORTH on a DEC LSI 11/23 minicomputer interfaced to a Bus and boards developed in house, will be replaced with a combination of a Sun SPARCstation running SunOS, a MicroSPARC based Single Board Computer running LynxOS, and various intelligent VME based peripheral cards. The software is based on a design philosophy originally developed by Pat Wallace for use on the Anglo Australian Telescope. This philosophy has gained wide acceptance, and is currently used in a number of observatories around the world. A key element of this philosophy is the division of the TCS into `Virtual' and `Real' parts. This will allow us to replace the higher level functions of the TCS with software running on the Sun, while still relying on the LSI 11/23 for performance of the lower level functions. Eventual transfer of lower level functions to the MicroSPARC system will then proceed incrementally through use of a Q-Bus to VME-Bus converter.

The Wildfire Instrument Controller is currently being used in four IR instruments at the Kitt Peak National Observatory. Wildfire is a high performance, versatile transputer based controller which manages the clocking and readout of the 2D arrays along with all other aspects of instrument control. Wildfire utilizes high speed fiber optic links for communication between its two components: the electronics mounted on the instrument and the digital signal processor located in the computer room. This paper is a systems level discussion of the hardware and software that make up the Wildfire Instrument Controller. We present an overview of the system operation along with measured performance data. We also discuss how the Wildfire system can be easily expanded to provide the higher level of performance required for high speed readout of the next generation 1024 by 1024 pixel IR arrays.

The Cassegrain focus of the 2nd unit telescope of ESO's VLT shall be equipped with a 10/20 micrometers camera spectrometer. This instrument shall be built by a consortium of institutes from ESO's member-states responsible for the final design. Still rather detailed internal studies were required to assess the overall scope of work, the repercussions on the telescope design and the required infrastructure at ESO. The internal studies have resulted in a multimode instrument allowing for diffraction limited imaging with variable magnification at 10/20 micrometers , low-, medium- plus potentially high-resolution long slit spectroscopy (approaching or exceeding (lambda) divided by (Delta) (lambda) approximately equals 50000). The optical design allows to take advantage of the rapid evolution in the field of detectors. Novel design features are an exchangeable entrance window, an anamorphic beam expansion and potentially the use of immersion gratings. At the La Silla observatory TIMMI, a small 10 micrometers camera/spectrometer for the 3.6 m telescope (built by the Service d'Astrophysique, CEN- Saclay, France) is in operation to allow both ESO and its community to gain practical experience in astronomy in the 10 micrometers atmospheric window. This instrument is currently equipped with a 64x64 element Ga:Si array. A 128x196 element array is presently in development at Leti/LIR of Grenoble, France, which may be used to upgrade ESO's current observing capabilities. Spectroscopy in TIMMI utilizes Germanium grisms machined via anisotropic etching. This manufacturing technology may also serve to fabricate immersion gratings from high index materials.

Four reflection, grazing incidence X-ray optics offer the possibility of broad band (0.1-10.0 keV) X-ray telescopes that fit within the physical envelope of NASA's Small and Medium Explorer mission. We describe a short focal length (4 m) optical design with arcsec imagery over a 1 degree(s) field of view and good efficiency in the Fe-K band (6.4-7.1 keV).

This paper describes the feasibility design of a high resolution multiple object fiber spectrograph. High resolution is obtained without cross dispersion with large 305 mm ruled length gratings and the use of the fiber link as an image slicer. The grating has a high groove density to provide a free spectral range larger than the CCD detector. The use of microlenses as focal converter at the slit has several implications on the design of the spectrograph. Unlike single object spectrographs, the presence of a significant telecentric slit field, 6 degrees long, makes the standard parabolic collimator inappropriate. A catadioptric collimator is used instead in double pass to image the intermediary spectrum sideways of the slit. The use of the white pupil and the good image quality of the collimator permit various camera options. The one foreseen is a f/1.25 finite conjugate catadioptric camera with a (2048)²24 micrometers pixel detector.

'FUEGOS' (Fibre Unit for European General Optical Spectroscopy) is one of the foreseen focal instrument for the future Europen VLT (Very Large Telescope). It is a multi-object spectrograph widely opened to stellar and extra-galactic programs. In the general concept of this instrument, the fibre link feeding the spectrograph slit from the telescope Nasmyth focus plays a major part. In addition to transporting light with maximum efficiency it is also asked to satisfy a set of very different specifications related to the spectrograph characteristics or to the particular conditions met at the VLT Nasmyth focus. We examine here the engineering solutions that were ultimately adopted and present the first results obtained with a fibre link prototype which was integrated with the help of a dedicated home- made instrumentation. More generally through this presentation, it will be demonstrated that fibre cables coupled with microlenses at both extremities could provide a simple and efficient solution for high resolution fibre fed spectrographs when it is possible to make use of the advantageous plate scale factor of large aperture telescopes.

We define the stability requirements for a high-resolution spectrograph, then show how these can be met at Cassegrain by modern materials, mechanism design, thermal control, and passive and active compensation for structural flexure. We consider the optimization of the information throughput of the spectrograph, in terms of slit-throughput, with the superb imaging performance of modern large telescopes and sites, new developments in image slicers, the prospects for adaptive-optics feeds for spectrographs, and the internal transmission of the optics. We consider in detail the requirements of, and solutions for, high resolution spectrographs for two large telescope projects - the 6.5 m MMT conversion and the two Gemini 8 m telescopes.

Preliminary optical designs have been developed for the blue channel of the Low Resolution Imaging Spectrograph (LRIS) for the Keck Ten Meter Telescope. This paper discusses the configuration- driving factors and performance of the designs, as well as coating and fabrication issues.

We have constructed a computer model for the simulation of spectroscopic astronomical observations of point sources. The aim is to help the observer to plan scientific and calibration exposures, to select the most suitable spectrographic parameters, to predict S/N ratios, and to test data reduction algorithms. The model, which was developed in the IRAF/STSDAS environment in the frame of the Russia / Ukraine / Italy / Germany cooperation for the SPECTRUM UV mission, is rather general and includes the whole chain: Telescope-Spectrograph-Detector. The database required to run the program consists of: (a) monochromatic PSF's of the telescope; (b) monochromatic PSF's of the spectrograph; (c) detector characteristics (pixel size, RON, gain, DC); (d) monochromatic throughputs of each system component. The input of the program is: (a) an observing mode (e.g. slit size, resolution, wavelength range, etc.); (b) the exposure time; (c) the sky background; (d) an input spectrum template. The output of the program is: (a) a 2D image containing the spectrum as it is recorded on the detector; (b) an estimate of the S/N ratio computed by taking into account also the photon noise. Examples of the application of the program to a Rowland mounting spectrograph is presented.

This paper presents the initial on-orbit performance of the corrective optics space telescope axial replacement or COSTAR. At this time the FOC and FOS channels are aligned and the data analyzed. The GHRS is aligned and initial results are presented. The installation and checkout of the COSTAR into the HST was flawless. The optical performance should enable the science lost to spherical aberration of the HST's primary mirror to be fully recovered.

The corrective optics space telescope axial replacement (COSTAR), designed to restore the high image quality that was originally expected from the Hubble Space Telescope, was successfully installed in the observatory during the first HST servicing mission in December 1993. While verification testing in the lab has shown that the imaging performance of COSTAR is nearly ideal, realization of the expected high spatial resolution of the COSTAR-corrected images depends strongly on the ability to align the COSTAR mirrors with respect to the telescope and axial science instruments. We report on the methods used in the weeks following the servicing mission to optimize the HST-COSTAR performance for the Faint Object Spectrograph and Goddard High Resolution Spectrograph channels. The relatively low spatial resolution afforded by the spectrographs as well as defects in their imaging modes presented special challenges to rapid optimization of the COSTAR correction. Nevertheless, the alignment process quickly converged and has yielded images that are superior to the COSTAR specifications.

The alignment of the faint object camera channel of the corrective optics space telescope axial replacement (COSTAR) is described and compared to the alignment plan that had been devised some months before. Images showing the appearance of the point-spread function during the course of the alignment process are presented, and the ultimate imaging performance detailed. A science image of a rich star field shows off the performance improvement over the uncorrected telescope.

An international team led by The University of Texas and Pennsylvania State University is planning a fiber fed high resolution spectrograph for the Spectroscopic Survey Telescope. Design goals are (1) resolution 60,000 with a fiber diameter greater than 1 arc second, (2) spectral coverage per exposure to the limits permitted by the combined bandpass of the SST and optical fibers and the RQE and area of the CCD. The R (DOT) (phi) product will be maximized by use of a high-blaze angle echelle mosaic and image or pupil slicing while the bandwidth will be maximized using cross dispersion and dichroic beam division in a dual spectrograph with large-area CCDs. Critical issues are cross dispersion and order spacing, sky subtraction, echelle and CCD selection, fiber optic feed, and image or pupil slicing. Proposed methods to achieve the goals and meet the challenges posed by the issues will be presented.

The near IR camera and multi-object spectrometer (NICMOS), is a new instrument for the Hubble Space Telescope (HST). NICMOS provides opportunities for near IR astrophysical investigations with HST in the 0.8 - 2.5 micrometers spectral region. NICMOS contains three cameras at different spatial resolutions for imaging, grisms for multi-object spectroscopy, polarizers for polarimetry, and a coronagraphic spot and mask for coronagraphic imaging. A solid nitrogen dewar maintains the three 256 x 256 pixel HgCdTe detectors at their operating temperature of 58K for five years. NICMOS is the next US IR space mission, scheduled on STS Mission 91 in March 1997. It is anticipated that the European ISO mission will be launched and operational by that time and that joint observational programs with the two instruments will be possible.

We describe plans for several science programs in crowded fields using the faint object spectrograph (FOS) which rely critically on the enhanced angular resolution provided by the COSTAR corrective optics. Based on ground-based calibration of the COSTAR and on-orbit performance of the FOS, the anticipated performance of the COSTAR+/FOS should allow many important scientific studies to be completed which have had to be postponed due to the spherical aberration in the HST primary mirror. Particularly impacted by spherical aberration, and thus able to benefit most dramatically from the installation of COSTAR, are spectroscopic observations with FOS in crowded fields. Many of the most important science goals for the FOS instrument involve observations of crowded fields, for which the restored high- angular resolution of HST is essential - not only to isolate features of small angular extent, but also to eliminate optical contamination from the surrounding luminous regions. Spectroscopy of the nuclei of galaxies to obtain rotation curves and velocity dispersions which might reveal the presence of central black holes benefits dramatically from the enhanced angular resolution of HST as restored by COSTAR. We present models based on our current understanding of the dynamics of galaxy nuclei to illustrate the dramatic improvement in sensitivity in searching for black holes made possible by restoring HST's image quality. Two other categories of scientific investigations in crowded fields which will benefit greatly from restoring HST's image quality are spectroscopy of the luminous material (presumably distant galaxies) surrounding quasistellar objects, and spectroscopy of individual stars in globular clusters. The promise of finally being able to carry out these exciting scientific programs with the FOS on the restored HST explains why our team of scientists, along with many other astronomers, look forward to using the restored HST to carry out the scientific investigations for which it was originally intended.

The high level of spatial uniformity in modern CCDs and other array detectors make them excellent devices for astrometric and navigational systems. However, at the level of accuracy envisioned by the more ambitious projects, current technology produces devices with significant pixel registration errors. This paper describes a technique for measuring relative pixel positions to an accuracy approaching 0.001 pixel. The technique has been applied to WF/PC II CCDs which are shown to have 500 nm step-and-repeat errors.

A multi-national consortium of research groups are developing the XMM (x-ray multi-mirror mission) optical monitor to provide a capability for optical identification and photometry of x-ray sources observed by the XMM observatory. This will be the first multi-wavelength facility dedicated to monitoring the variability of diverse sources from the optical through to x-ray wavelengths. Here we describe the system design and discuss progress in the breadboard phase of the development program.

The electro-optical space surveillance (EOSS) technology program at MIT/Lincoln Laboratory develops advanced electro-optical technology for space and missile surveillance. This advanced technology includes visible CCD focal plane arrays and cameras, mid-wave IR (MWIR) Shottky barrier focal planes and cameras, and the detection processing algorithms and processors to optimize the use of these focal plane arrays. Some of these EOSS technologies are highly suited to the detection of faint near earth objects. A near-earth asteroid surveillance system has been designed to utilize recently developed technology and existing telescopes to meet the performance requirements outlined in the NASA SPACEGUARD Study (Morrison, 1992). A detailed model was developed to evaluate the expected detection performance of this system and the proposed Spaceguard system. The components of the designed system have been tested as part of routine device evaluation and field measurements. Results indicate that the proposed system can meet the performance objectives spelled out in the spaceguard study at lower cost and risk than the proposed spaceguard system and with existing technology.

A design for an advanced camera (AC) third-generation Hubble Space Telescope scientific instrument is discussed. The AC is a three-channel spectrophotometric camera with wavelength sensitivity from 115-1000 nm. The AC, if selected, would be launched in 1999 for installation on HST. The axial bay design incorporates optical correction for the aberrated HST primary mirror and evolutionary advances in imaging capability.

We present details of an instrumental configuration dedicated to the detection of large gas-giant planets orbiting main sequence stars via the small drop in stellar brightness that occurs when the planet transits the stellar disk. Our proposed scheme involves the long-term monitoring of the light-curves of tens of thousands of stars using a Schmidt telescope coupled to a CCD via a bundle of 10,000 optical fibres. Using an existing theoretical model of planetary system formation, we calculate a detection rate of 14 transits per year for our proposed system. Some fibre- optic laboratory results and some tests of a 10 fibre multi- object photometer prototype used on the UK Schmidt telescope (UKST) are also presented to support the feasibility of this idea.

In order to explore the nature of the limits on direct extrasolar planet detection we have generated high accuracy broadband background models for several different cases. The simplest assumes an ideal diffraction limited background with shot and read noise errors. More complex models based on phase error maps drawn from real metrology data include the effects of scatter in the optical system. To these backgrounds a planet image can be added at various relative intensity levels. In the simples case, the background dominated by ideal diffraction is so smooth that a median filter is very effective at removing it locally, permitting planet detection at the limits of the flat field error of the detector. The use of more complex filters for the scatter limited case will be discussed.

Fractal compression of deep space and satellite-derived images using iterated function theory makes it possible to compress these complex natural images into a codified template, from which a facsimile of the original image can be regenerated. The coding technique makes use of the fractal properties of mathematical scale invariance and self-similarity to generate a set of transforms which occupy a minimum amount of space. The result is that these images can be compressed to a fraction of their original size, allowing high compression ratios to be obtained. Image compression using fractal techniques is information lossy. This means that the image generated from the template will be an approximation to the original image. For this reason, raw data should not be processed by such techniques. However, for information retrieval systems, where it is necessary to store many hundreds of images, this technique will allow the capacity of such systems to be increased several fold. This paper will give a brief review of fractals and then describe their use in image compression. The mathematics of iterated function schemes is considered. The coding and decoding schemes utilized in the compression and decompression, respectively, are presented. The paper will finish by showing the results of compression and decompression and will quantify differences between the original and facsimile.

The Richardson-Lucy iteration (also known as the EM iteration for image restoration with Poisson statistics) is the most widely used image restoration technique for optical astronomical data. Like all maximum likelihood methods, it suffers from noise amplification in the restored images. Previously suggested methods for dealing with this problem (stopping the iteration early or smoothing the final image) have serious drawbacks for astronomical applications. This paper describes a new image restoration iteration based on the RL method that reduces noise amplification. The method is based on a modified form of the Poisson likelihood function that is flatter in the vicinity of a good fit. The resulting iteration is very similar to the RL iteration, but includes a new spatially adaptive damping factor that prevents noise amplification in regions of the image where a smooth model provides an adequate fit to the data; thus, I call this the `damped Richardson-Lucy iteration.' The damped iteration converges more rapidly than the RL method and can be accelerated using the same techniques as the standard RL iteration. Results are shown for both simulated data and Hubble Space Telescope images.

The Hubble Space Telescope image quality was degraded significantly as a result of the spherical aberration in the primary mirror. Repairs made during the first servicing mission have restored HST's optical performance to the original design specifications. This paper examines not only how effective image restoration has been for the data from the first three and a half years of the mission, but also demonstrates the applicability of the same deconvolution algorithms to postservicing mission images. A variety of well-known image restoration techniques, such as Wiener filters, Richardson-Lucy, Jansson-van Cittert, CLEAN, and maximum entropy, have been applied to HST imaging with reasonable success. However, all techniques have been hampered by incomplete knowledge of the point spread function and the space variance of the point spread function (PSF) in the Wide Field/Planetary Camera, HST's most frequently used imager. After the servicing mission, the COSTAR-corrected Faint Object Camera acquires a space-variant PSF. Restoration is further complicated by time variance of the PSF resulting from both long-term and short-term (orbital period) changes in focus, problems which are not likely to be ameliorated by the servicing mission. Thus, it will be important to continue the efforts in modeling the HST optical system and apply deconvolution methods in order to attain the best scientific results from the telescope.

This paper describes the Bayesian image reconstruction algorithm with entropy prior with space variant hyperparameter. The spatial variation of the hyperparameter allows different degrees of resolution in areas of high and low signal/noise ratio, thus avoiding the large residuals present in algorithms that use a constant balancing parameter. The space variant hyperparameter determines the relative weight between the prior information and the likelihood defining the degree of smoothness of the solution. To compute the variable hyperparameter we used a segmentation technique based on artificial neural networks of the Self- Organizing Map type. Using this technique we segmented the image in 25 regions and computed a different value of the hyperparameter for each one. We applied the method to the Hubble Space Telescope Cameras and to ground based CCD data.

In this paper, we describe a complete pattern recognition procedure for extracting magnetic features in solar images, for tracking each feature from an image to another, thus deriving rotational and meridional motions with unequaled accuracy, and for tracking some of these features concurrently within a multispectral image flow, thus building the corresponding 3D magnetic structure and its evolution over time. The segmentation procedures lean upon the `translation' of the physical knowledge into an algorithmic functional decomposition. This methodology enables a fast and robust design of algorithms and leads to a higher accuracy. A multiresolution elastic matching is then used to track features and extract the surface motion of the sun. This matching implies designing an optimization scheme that fits the anisotropy of the motion as well as a target selection that requires no precomputed threshold. The multispectral images are segmented in a cooperative way, thus giving 3D structures. We introduce a fast algorithm to build the 3D connection between planar sections and show how the volumic information is necessary to tracking, based on the same elastic matching scheme.

Since the digital detectors (like CCDs) presently available for astronomical instruments have two dimensions only, there is an obvious problem for obtaining detailed spectroscopic information on extended astronomical objects (e.g. active galaxies, regions of star formation, globular clusters, etc.); the classical long- slit spectrographic techniques are grossly inadequate. Several so-called integral field spectrographs (IFS) have been used at the Canada-France-Hawaii telescope during the past years to improve this situation. They typically provide hundreds of spectra in a small (approximately 10' by 10') field with a subarcsecond spatial resolution capability and hundreds of spectral elements, both simultaneously accessed. We describe these various instruments and their present performances, review a few illustrative results obtained with CFH telescope and discuss the expected developments: use of IFS in combination with adaptive optics systems for the study of individual objects at spatial resolution near the 0.1' level, as well as their potential capabilities on Very Large Telescopes, with or without adaptive optics.

White and reflective screen materials were examined for their suitability as the flat-fielding screen for the Sloan Digital Sky Survey 2.5-meter Telescope. Candidate materials were found which exhibited high, uniform reflectance from 390 to 910 nm. In addition they are durable and rollable, inexpensive, require a minimum of maintenance and are not a significant source of particulate contamination.

The downhill simplex and simulated annealing algorithms are powerful design tools for optical systems that do not yield to analytic methods of optimization. Use of these numerical techniques have been frustrated by the shortage of computational resources available to investigators. To tap the latent computational resources available in even moderate sized local area networks (LAN's), we have developed software to operate flexibly, a LAN composed of UNIX workstations as a distributed processor for optical design optimization. We have developed a general purpose remote procedure call generator and subroutine library which allow simple specification and use of existing subprograms as remote procedures. This software provides a mechanism for modifying or developing large scientific and engineering applications with inherent parallelism for distributed processing on LAN's. We discuss the performance, bandwidth, and adaptability of this software as applied to the problem of grazing incidence optics design using the simulated annealing algorithm.

The thermal conductivity, specific heat, and elastic moduli of silicon indicate that silicon is an attractive material for the substrates of large cryogenic mirrors. Elemental silicon is available in large single crystals as well as polycrystalline blanks. Silicon is already used extensively in optics for both high-index IR lenses and IR filter substrates. We report on cryogenic tests of silicon mirrors. Our samples show that on cycling from room temperature to 77 K, the dimensional stability is only slightly worse than that of fused silica, which is a known highly stable, cryogenic mirror substrate. The dimensional stability of silicon is much better than that of metal mirrors. The fabrication of silicon mirrors, including a 20 by 40 cm grating blank, for a high-resolution, IR spectrograph now under construction at NOAO is discussed.

This paper describes an image slicer based on a modified version of the Bowen-Walraven slicer. The new design is called Confocal Image Slicer, and delivers a large number of slices, all in focus. This allows the slicing of a wider area of the seeing disk, to the benefit of very high resolution spectrographs like the Ultra-High-Resolution Facility, where the slicer gives access to stars up to 3 magnitudes fainter than the limit imposed by a narrow slit for the same resolution and signal to noise ratio. The basic principle and design philosophy of the confocal image slicer are presented, together with a description of the development of a prototype and a final version now fully commissioned at the Anglo-Australian telescope. Some astrophysical results are also included.

We describe recent orbital test data acquired with the COSTAR/Faint Object Spectrograph (FOS) on the Hubble Space Telescope (HST). These data exhibit possible internal thruput deficit and scattering features in the FOS. We then describe some data from laboratory measurements of COSTAR optical components carried out at Ball Aerospace Corporation and NASA/Goddard Space Flight Center, which we initially thought might bear on some of the orbital test data. We use next, a theoretical model to simulate some of the HST/COSTAR/FOS imaging and spectral performance. We apply the model to the current data to assist in evaluating COSTAR iamging performance, to attempt to isolate the origin of the thruput deficit and scattering features. The results of these studies will permit a better understanding of the limits of HST performance and permit development of optimal strategies for performing spatially resolved spectroscopy in the FOS scientific observing program.

Novel principle of measurement of amplitudes and phases of wavefield in holes of mask inserting in the light ray has been suggested. The method is based on properties of Fresnel images in the Talbot effect.

We discuss the various errors affecting an astronomical spectrograph, whether mounted directly at the telescope, or placed on a table. We then propose a method for measuring these errors, using the Shack-Hartmann principle, which has been successfully used in the field of active and adaptive optics. It is argued that such a measurement can be used to correct the aberrations on-line, leading to considerable improvement in the performance of the spectrographs, as well as leading to much reduced costs.

In an effort to overcome the small size of optical and IR imaging arrays a wide range of techniques for building mosaics of small arrays has been attempted. These methodologies are typically handicapped by one or more shortcomings, including a severe limit on the number of detectors which can be built into the mosaic, physical (non-imaging) gaps in the array, or difficulty of fabrication of imaging optics. We have developed an optical technique which has the capacity to enable extremely large mosaics of 2D arrays, while at the same time exhibiting absolutely no gaps between the elements of the mosaic. Further, through parallel processing of the detectors in the mosaic very high speed readouts of the image data are possible relative to than achievable with comparable large monolithic arrays (if they could even be fabricated). Several laboratory bench prototypes of this imaging system have been built and tested with outstanding results. We discuss the possible optical configurations of the system, techniques for aligning the arrays, and possible configurations for use at the telescope. Examples of test imagery obtained with the prototypes are shown.

Arrays of silicon sensors arranged in a CCD focal plane architecture have become the detector of choice for astronomical imaging in the 300 nm to 1 micron region. However other focal plane architectures offer attractive additional features such as allowing for random access to, and readout of, any subarray on the chip, and the nondestructive interrogation of the signal level of any pixel. A 512 x 512 pixel array utilizing a Charge Injection architecture that offers such advantages has been tested at liquid nitrogen temperature as to its suitability for use in astronomy. Laboratory tests show that using nondestructive readouts results in a (root)N (where N is the number of readouts) improvements in the noise figure (approximately 20 electrons rms noise after 100 reads). Initial images with a CID-38 of a number of astronomical objects are also presented.

The next generation of CCD imagers will undoubtedly be mosaics, in order to overcome (1) the low yield when fabricating large devices, (2) the limited number of readout channels that can be put easily on a single CCD, and (3) the physical limitation of the 4-inch silicon wafer. As a first step towards a 8192 x 8192 CCD mosaic, we have recently fabricated a 2 x 2 array of Loral 2048 x 2048 CCDs. The two-side buttable design of these chips (by John Geary of SAO) allows us to achieve gaps of about 0.6 mm (40 pixels). This prototype mini-mosaic imager, using unthinned, front-illuminated CCDs has been used at the KPNO 0.9 m and 4 m telescopes. As we are constructing a number of scientific-grade mosaics with thinned chips for use at KPNO and CTIO, we are beginning the design and fabrication work for an 8192 x 8192 imager. This will be a 2 x 4 array of Loral 4096 x 2048 CCDs with interchip spacing of less than 0.5 mm. Such a device will have a physical size of approximately 5 inches square and will cover an area of 38.6 (59.1) arcminutes on an edge at the 4 m (0.9 m) telescope with a pixel size of 0.28 (0.43) arcseconds per pixel. This paper discusses results obtained with the 4096 x 4096 minimosaic and design strategies/progress on the larger 8192 x 8192 imager. Specifically, we present designs of the Dewar and mechanical interface for the large mosaic, a physical mounting scheme which will achieve better than 5 micron RMS flatness, and a discussion of the electronics and controller (the CTIO transputer-based ARCON), which will allow us to read out the entire array in less than two minutes. Some strategies for dealing with the large amount of data (128 Megabytes per image) will be presented.

One shortcoming of large area, thinned scientific CCDs has been their lack of planarity. Areas of the silicon membrane that lie outside the depth of focus of the optical system lose resolution and cause a distorted image. Reticon has developed a thinning procedure that eliminates the ripples and bow often seen in thinned CCDs. This paper discusses the thinning procedure used to produce a flat, thinned, 2048 x 2048 pixel scientific CCD. The pixels are 13.5 micrometers square, making the image area of the CCD larger than 1' on a side. The final silicon membrane is between 16 and 18 micrometers thick and is flat to less than 5 micrometers peak-to- peak across the imaging area. Frontside devices have demonstrated a low dark current of 50 pa/cm² at 20 C and full well of 120k electrons when the CCD is run in MPP mode. The charge transfer efficiency exceeds 0.99999 and the readout noise is less than 2.5 electrons when the CCD is operated at 20 kHz.

A new controller concept based on transputer modules and digital signal processors (DSPs) has been developed at ESO. The controller can handle single CCDs or CCD mosaics with 1 to 64 outputs and with up to 256 clocks with software programmable independent high and low voltages. The same basic concept will be used for slow scan scientific and fast auto guiding/wavefront sensing applications. While the transputers are mainly used for data acquisition, communications and supervisory tasks, a DSP with very simple external logic performs the time-critical CCD clock pattern generation. The low noise, high speed clock driver with built-in telemetry is designed for highest operational safety. ACE can easily be interfaced to various types of host computers, PCs, Unix workstations or Local Control Units.

Germanium photoconductors are considered the `state of the art' for detection of the 30 to 200 micrometers band of the electromagnetic spectrum. Ge:Be, Ge:Ga and stressed Ge:Ga detectors have been developed for imaging and spectroscopic astronomy, and are successfully used in facilities such as sounding rockets and the Kuiper Airborne Observatory. Their use is also projected for NASA's Space IR Telescope Facility and ESA's IR Space Observatory. This paper describes a few of the figures of merit for these devices with consideration given to the various ways in which they are employed. Current device performance is reported. We briefly discuss recent advancements in understanding the electrical behavior of these devices as they respond to a step change in photon flux. Also discussed are several 2D photoconductor arrays that are currently in use or under development.

We review recent arguments for using increased spectral bandwidth and exposure times to optimize the signal-to-noise ratio of speckle imaging estimators and discuss the tradeoff between camera exposure time and the number of data frames collected when observing time is fixed. We compare experimental results with a previously-derived expression for optimal exposure time and find reasonable agreement after accounting for frequency-dependent camera noise.

The use of support constraints for noise reduction in images obtained with telescopes that use adaptive optics for atmospheric correction is discussed in paper. The effectiveness of support constraints in achieving noise reduction is discussed in terms of the noise properties and in terms of the type of algorithms used to enforce the support constraints. Both a convex projections and a cost function minimization algorithm are used to enforce the support constraints, and it is shown via computer simulation and field data that the cost function algorithm results in artifacts in the reconstructions, in general. The convex projections algorithms produced mean square error decreases in the image domain of approximately 10% for high light levels, but essentially no error decreases for low light levels.

The Gemini Telescopes are technically advanced 8-m class telescopes which are characterized by their excellent image quality, low emissivity and versatility goals. Gemini has had the foresight to realize the importance of instrumentation support and services to achieve these goals. The instrumentation group of the Gemini Project has developed a preliminary design for an instrument rotator, instrument support structure, and cable/services wrap which satisfies the Gemini science requirements and engineering requirements. This concept will be the subject of a preliminary design review in June, 1994. This paper describes the driving scientific and engineering requirements, the preliminary design, and provides a brief explanation of how the design satisfies the requirements.

The segmented design of the W. M. Keck Telescope primary mirror places several unique demands upon the alignment and adjustment of the telescope optics. These include: (1) careful determination of the optical figures of individual segments (to provide input data for warping harness adjustment), (2) control of the two tilt degrees of freedom for each of the thirty-six primary mirror segments, and (3) phasing or control of the piston degree of freedom for each of these segments. In addition, (4) the proper alignment of the secondary with respect to the primary, although it is a requirement common to monolithic and segmented telescopes alike, is a more subtle and complicated task for the latter because the optic axis of the primary is not readily defined. These four tasks are performed at Keck by the Phasing Camera System.

We analyze the look and feel of each display and discuss its software design. Developed using the Motif Toolkit, each display runs on an X windows work station. In the area of look and feel, we discuss the pros and cons of graphically depicting the instrument light path. In the area of software design, we discuss Motif and Xlib programming techniques.

This paper describes the operation and performance of an IBM- PC/AT based image data acquisition system for large format CCDs. The system currently under development can handle image data from CCDs with pixel organization up to 2048 x 2048. Based on commercially available DT-3852 frame grabber board (M/s Data Translation Inc.) and locally developed I/O interfaces, the system provides an economical and user friendly acquisition, display and quick look analysis facility.

We discuss the present implementation of the data acquisition electronics and the planned upgraded version, together with the related algorithms, of the new Torino IR camera: TC-MIRC. A detailed description of the camera is available in M. Robberto et al., 1994, presented in this SPIE meeting. The electronic structure of TC-MIRC is based on a PC-compatible host computer, using IDL/Windows, and a transputer network for detectors management and data processing; the engineering advantages of an intrinsically distributed environment and the computational scalability are discussed. The data acquisition procedures (standard algorithms for coadding and nondestructive readout, and their use in normal operations) are evaluated, with respect to their present performance and to the upgrade needed to fulfill the design requirements.

We describe a multi-object spectrograph that is currently under design for the f/5 focus of the converted Multiple Mirror Telescope. The f/5 Cassegrain focus will use a three element refractive corrector to produce a 1 degree(s) diameter field of view well matched to multi-object spectroscopy with optical fibers. The optical fibers will be mounted on magnetic buttons and positioned with two high speed robots, using the general techniques pioneered by the AUTOFIB instrument. Our goal is to position the 300 fibers in 300 seconds. The spectrograph optics will be bench mounted on the observing floor, and will be fed by fibers that are 10 meters in length. The spectrograph will have a 250 mm diameter collimated beam to allow a resolution of 4 to 8 $angstrom FWHM over approximately an octave of spectrum between 3400 $angstrom and 1 micrometers . It is our intention to complete the Hectospec before the commissioning of the f/5 focus in late 1996.

The complexity of the structure of many astronomical objects (viz. galaxies, nebulae, comets etc.) is such that details of polarization frequently requires seeing limited resolution. A two channel imaging polarimeter has been described here which can provide seeing limited spatial resolution and highly accurate values of linear polarization by minimizing the harmful effects of atmospheric variations. This can be achieved by using a Wollaston prism analyzer to split a telescope image into two polarized components and simultaneously comparing them to eliminate effect due to atmospheric scintillation. The introduction of a grid at the focal plane, prevents the overlapping of o and e images on the CCD plane, where the image is refocussed. In order to keep the orientation of the two polarized components fixed in the instrument frame, the polarization vector of the image is modulated by mechanically rotating a half wave plate and readings are taken for several positions of the half wave plate to derive the polarization vector. We plan to use the above instrument at the cassegrain plane of 2.3 m V B telescope, India (f/13 beam) to carry out imaging polarimetry of extended astronomical objects. With 1000 secs. exposure time for each frame, on the GEC/EEF P8603A CCD chip, one can measure polarization values with an uncertainty of 0.3%, for the Seyfert N1068 (m_v equals 17.5 per square arc sec) when observed in V band. This uncertainty is mainly dominated by photon count statistics. This instrument is presently being fabricated at Inter University Center for Astronomy and Astrophysics, Pune, India.

Until recently the study of cool clouds of interstellar matter had been limited by the relatively low spectral resolutions provided by existing spectrographs. The Ultra-High-Resolution Facility (UHRF) recently commissioned at the Anglo-Australian Telescope has changed dramatically this panorama by delivering for the first time resolutions approaching one million, near the diffraction limit of the largest echelle gratings available. The instrument shares the east coude room with the University College London Echelle Spectrograph, in what is now one of the most powerful spectrographic installations worldwide. This contribution describes the characteristics of the UHRF, including its design, manufacture, testing, and commissioning. The UHRF incorporates a novel image slicer (described elsewhere in these proceedings), which allows ultra-high-resolution observations on faint objects. Astrophysical results from the first observing runs are presented to demonstrate the UHRF performance in both resolution and throughput.

Instrumentation and observing methods are being developed for a program in optical high-speed astrophysics, an exploratory project entering the domains of milli-, micro-, and nanosecond variability. Current studies include accretion flows around compact objects, stellar scintillation, and astronomical quantum optics. To study such rapid phenomena is not possible everywhere in the spectrum (e.g., X-ray studies are constrained by the photon count rates feasible with current spacecraft). The ground- based optical is a promising region, for which we have constructed a dedicated instrument, QVANTOS ('Quantum-Optical Spectrometer'). It was designed for real-time handling of large amounts of data, for observing also faint sources, and with a time resolution that can be extended to reveal quantum properties of light, such as the bunching of photons in time. Its first version was used to study atmospheric scintillation on timescales between 100 milli- and 100 nsec, utilizing some 25 full nights at a telescope on La Palma (Canary Islands). An understanding of the atmosphere is required to segregate astrophysical variability from terrestial effects, and to find optimal observing strategies. For very high time resolution, light curves are of little use, and statistical functions of variability have to be measured. The noise in such functions decreases dramatically with increased light collecting power, making very large telescopes much more sensitive for the study of rapid variability than ordinary-sized ones.

We discuss a design for an echelle spectrograph based on three prisms as the cross-dispersing optical elements. Since high throughput in our particular system is achieved using an image slicer, obtaining adequate order separation is a challenge. We find that the design described herein yields good efficiencies throughout the visible wavelength range while also providing adequate order separation in the red.

We present the design of the High Resolution Spectrograph (SARG, 25,000 <EQ R <EQ 200,000) for the Italian Galileo National Telescope (TNG). SARG will be a fiber-fed, white-pupil echelle spectrograph, with prismatic cross disperser and two cameras, placed in a temperature controlled room in the telescope pillar. Additionally provision has been made for a cryogenic IR arm extending the spectral coverage up to 1.7 micrometers . Novel features include a new design of the fiber spectrograph interface and an asymmetric white-pupil collimator.

The focal reducer and low dispersion spectrograph FORS which is presently under construction for the ESO Very Large Telescope (VLT) will be an all dioptric focal reducer instrument designed for direct imaging, low-dispersion multi object spectroscopy, imaging and spectropolarimetry of faint objects. A central mechanical component of FORS is the multi object spectroscopy (MOS) unit. The design and the observation capabilities of the MOS are described. Further details on the whole instrument are given in [1]. Two identical copies of the instrument will be built by a consortium of three astronomical institutes (Landessterwarte Heidelberg and the Unviersity Observatories of Gottingen and Munich) under contract and in cooeration with ESO. The instruments will be installed at the Cassegrain foci of two unit telescopes of the VLT.

The concept design of the faint object camera and spectrograph (FOCAS) for the 8.2 m SUBARU telescope (JNLT) is presented. The FOCAS is designed for enabling observations of extremely faint galaxies distributed over a 6 arcmin FOV with spectral resolution of R equals 1000. Fully transmitting optics will produce high quality images better than about 20 micrometers rms diameter over the entire FOV in white light of a wavelength range of 0.365 - 0.9 micrometers . A pair of abuttable CCDs with 4000 x 2000 pixels of 15 micrometers square pixel size are considered to be used. Several observing modes are available with the automated multislit assembly and changable optical elements. Control software coupled with telescope control software enables an automatic observation of several tens of galaxies simultaneously.

We started conceptional design of the multi-object spectrograph for SUBARU telescope. As the limiting magnitudes are deeper, number of objects observable in the focal plane is very large. To minimize loss time in observing night and to realize simultaneous observation of 700 objects, we propose modified plug-plate system by using mosaic fiber plate.

The Florida Institue of Technology and the Phillips Laboratory have developed several advanced visible (0.4-0.8 micrometers ) imaging fourier transform spectrometer (IFTS) brassboards, which simultaneously acquire one spatial and one spectral dimension of the hyperspectral image cube. The initial versions of these instruments may be scanned across a scene or configured with a scan mirror to pick up the second spatial dimension of the image cube. The current visible hyperspectral imagers possess a combination of features, including (1) low to moderate spectral resolution for hundreds/thousands of spectral channels, (2) robust design, with no moving parts, (3) detector limited free spectral range, (4) detector-limited spatial and spectral resolution, and (5) field widened operation. The utility of this type of instrument reaches its logical conclusion however, with an instrument designed to acquire all three dimensions of the hyperspectral image cube (both spatial and one spectral) simultaneously. In this paper we present the (1) detailed optical system designs for the brassboard instruments, (2) the current data acquisition system, (3) data reduction and analysis techniques unique to hyperspectral sensor systems which operate with photometric accuracy, and (4) several methods to modify the basic instrument design to allow simultaneous acquistion of the entire hyperspectral image cube. The hyperspectral sensor systems which are being developed and whose utility is being pioneered by Florida Tech and the Phillips Laboratory are applicable to numerous DoD and civil applications including (1) space object identification, (2) radiometrically correct satellite image and spectral signature database observations, (3) simultaneous spactial/spectral observations of booster plumes for strategic and surrogate tactical missile signature identification, and (4) spatial/spectral visible and IR astronomical observations with photometric accuracy.

The characterization and performance of a scientific-grade SBRC 256 x 256 InSb FPA with the IRCAM3/ALICE system is described, together with examples of anticipated astronomical use.

The characterization and performance of a science-grade SBRC 256 x 256 InSb FPA with the IRCAM3/ALICE system is described, together with anticipated examples of astronomical usage.

The Cryogenic Optical Bench (COB) is a 1-4 (mu) IR array camera with multiple cold spectral and spatial filtering capabilities which can be combined in a variety of configurations. The array is driven by a transputer based high speed data system using fiber optic links and dedicated processors. We describe the instrument functions and the mechanical, optical, electronic, and cryogenic implementation. COB is a facility instrument at Kitt Peak National Observatory, available for use by scientists worldwide on the basis of scientific merit.

The design of a polarizing Fourier transform spectrometer, developed for submillimeter and mid-infrared astronomical spectroscopy, is presentd. Results from recent observing runs are used to illustrate its performance.

Michelle is a facility-class long-slit spectrometer for the mid- IR being built at the Royal Observatory Edinburgh for the UK Infrared Telescope (UKIRT). With diamond turned, all-reflective optics to achieve high throughput and a cryogenic mechanism to select one from five diffraction gratings during operation, it will be capable of taking spectra at resolving powers ranging from 100 to 30,000 over wavelengths from 8 to 25 micrometers . A separate optical path will provide the ability to switch in well under a minute from spectroscopy to taking fully sampled diffraction limited images of the same field. Two mechanical coolers will maintain its optics at below 25K, while a Joule- Thomson (JT) stage on one of the coolers will keep the Si:As hybrid detector array at less than 10K. We present the predicted performance of the instrument, along with its opto-mechanical layout and the means by which it can be easily converted for use on the proposed Gemini telescope. The design of some unusual mechanisms is discussed.

Conventional search techniques for extrasolar planets have not yielded indisputable evidence of such planets. The small sample sizes of previously employed searches ensure that a null result provides little or no information on the numbers or distributions of such planets. The photometric technique, wherein monitoring is performed to detect light curves distinctively characteristic of a planetary transit, has been a tantalizing prospect, but has always been rejected because of its observational intensity. Moderate aperture (0.5-1.0 m) automated telescopes and multi- color CCD imaging arrays with image recognition software form the instrumental foundation for a global network to conduct such a search. Catalogs of cool dwarf stars provide a program star list with the advantages of orbital geometries and photometric properties enabling detection of planetary transits with a network of modes proportions. We reaffirm the viability of the photometric approach and outline the methodology of the search. Computer modelling of various strategies of conducting such a planetary search has been undertaken and we outline the resulting network configurations and observing strategies suggested by this effort.

Two methods to obtain the autocorrelation central value from clipped photon-counting data are presented. They are based on the statistical analysis of the total number of counts in one frame. One of the methods requires a smaller number of frames to provide a good statistical accuracy while the other has a wider intensity range of applicability. These are the first techniques that overcome the problem of the autocorrelation central hole due to the clipping in photon-counting detection.

Current astronomical CCDs (Charge-coupled devices) are limited in size. However, there are increasing demands for larger devices for spectroscopic and direct imaging use. A comparison of the areas of a direct Schmidt-survey plate, and the largest CCD likely to be available in the foreseeable future, serves to illustrate limitations of current electronic technology. Similarly, modern spectrographs can serve either high spectral resolution or multi-object capability, but not both. One could conceive of an updated analogue of a classical coude spectrograph, which would provide both simultaneously, were the enormous detector areas required available. It is with these two long-term applications in view that we consider the prospects for large-scale CCD mosaics. Spectroscopy in particular would benefit from mixed devices in the same mosaic e.g. blue and red- optimised, IR devices etc. We address how optimally to provide the drive waveforms with (1) minimum interconnections and complexity, (2) the ability to mix CCDs, (3) the ability to tune the individual CCD waveforms for optimum low-light level performance, and (4) to conceive an architecture which is indefinitely expansible. Our solution is to multiplex the drive waveforms to the CCDs as well as the signals from them. Minimum chip-count in the sequencing electronics is ensured by using high-density PLD technology (Programmable Logic Device), which enables a module of sixteen or more CCDs to be sequenced from one PLD plus memory. We describe a laboratory prototype, and describe how this could be developed into a system with all the drive electronics for large numbers of CCDs immediately behind the focal plane array. We also summarize our software system for efficiently generating and editing the bitmaps which define the CCD waveforms.

The Cryogenic Spectrometer (CRSP) is a longslit astronomical spectrograph which has been in service at Kitt Peak National Observatory since 1988, utilizing a 58 x 62 Santa Barbara Research Corporation InSb array in the dispersive focal plane. We have recently completed an extensive upgrade to the instrument which includes: installation of a SBRC 256 x 256 InSb array in the focal plane. CRSP is thus the first astronomical IR spectrograph to utilize the new 256 x 256 InSb focal plane. By comparison to the 62 x 58 focal plane, the 256 x 256 array has significantly less dark current (<1 e/s vs 50 e/s) and lower read noise (30 vs 350 electrons for a single read), resulting in improved performance for low background observations. In addition, the smaller pixels yield plate scales which are well- suited to sampling the typical seeing at Kitt Peak. This yields significant gains in the reduction of systematic errors associated with the extraction of point-source spectra against the challenging background of the IR night sky, which is dominated by emission lines of OH and by thermal emission from telluric absorption lines at wavelengths > 2.3 micrometers .

We have built an imaging polarimeter for use at mid-IR wavelengths (i.e. N band or 8 - 13 micrometers ). The detecting element is a 128 x 128 element Si:Ga Focal Plane Array, supplied by Amber Engineering, USA. The polarimeter itself provides diffraction limited images on a 4 m class telescope and has a field of view of about 32 arcseconds of sky with 1/4 arcsecond pixels. The two orthogonal planes of polarization can be imaged simultaneously so that each integration provides an independent measure of polarization. This instrument had a successful commissioning run in July 1993, and the polarization beam splitting assembly will be commissioned in May of this year.

New two-mirror fast coma-free systems with zero spherical aberration are proposed. Two large classes of these systems are investigated, 'front focal' systems and 'behind focal' systems. Well-known classical two-mirror systems are special cases of these classes. Some new aplanatic two-mirror systems cannot be found and investigated by means using third-order aberration theory, Gaussian optics and well-known optical computer programs (BIEM-3 for example). These systems can be used in telescopes, radiotelescopes, systems for cosmic rays explorations, and light collectors, (e.g., sun energy, etc.). These new systems can be used for investigations in a large spectral region, from far UV to far IR, or in radio wave observations, for astrophysics and space science. It is necessary to have special mathematical theory and software for designing of these new systems.

We describe the high resolution echelle spectrometer (HIRES) now in operation on the Keck Telescope. HIRES, which is permanently located at a Nasmyth focus, is a standard in-plane echelle spectrometer with grating post dispersion. The collimated beam diameter is 12', and the echelle is a 1 x 3 mosaic, 12' by 48' in total size, of 52.6 gr mm^MIN1, R-2.8 echelles. The cross disperser is a 2 x 1 mosaic, 24' by 16 ' in size. The camera is of a unique new design: a large (30' aperture) f/1.0, all spherical, all fused silica, catadioptric system with superachromatic performance. It spans the entire chromatic range from 0.3 (mu) to beyond 1.1 (mu) , delivering 12.6-micron (rms) images, averaged over all colors and field angles, without refocus. The detector is a thinned, backside-illuminated, Tektronix 2048 x 2048 CCD with 24-micron pixels, which spans the spectral region from 0.3 (mu) to 1.1 (mu) with very high overall quantum efficiency. The limiting spectral resolution of HIRES is 67,000 with the present CCD pixel size. The overall 'throughput' (resolution x slit width) product achieved by HIRES is 39,000 arcseconds. Peak overall efficiency for the spectrograph (not including telescope and slit losses) is 13% at 6000 angstrom. Some first-light science activities, including quasar absorption line spectra, beryllium abundances in metal-poor stars, lithium abundances in brown-dwarf candidates, and asteroseismology are discussed.

In this paper we present (1) the optical system design and operational overview, (2) laboratory evaluation spectra, and (3) a sample of the first observational data taken with HYSAT. The hyperspectral sensor systems which are being developed and whose utility is being pioneered by the Phillips Laboratory are applicable to several important SOI (space object identification), military, and civil applications including (1) spectral signature simulations, satellite model validation, and satellite database observations and (3) simultaneous spatial/spectral observations of booster plumes for strategic and surrogate tactical missile signature identification. The sensor system is also applicable to a wide range of other applications, including astronomy, camouflage discrimination, smoke chemical analysis, environmental/agricultural resource sensing, terrain analysis, and ground surveillance. Only SOI applications will be discussed here.

Lick Observatory has recently developed a near IR camera for astronomical imaging. The new camera has been built around the Rockwell International Inc. NICMOSIII 256 x 256 HgCdTe focal plane array. The dewar and optics were manufactured by IR Laboratories in Tuscon AZ, while the electronics, telescope interface, and data system were built at Lick Observatory. The Lick IR camera was designed primarily as a 1 to 2.5 micron Cassegrain imager, however, the dewar can also be operated as a high resolution spectrographic camera for the Lick Coude' Spectrograph in the 1 micron spectral region. At the Cassegrain focus the instrument is used to image faint galaxies, diffuse IR sources in star formation regions, and other astronomical objects. The camera operates at f/17 with focal plane scale reductions of 1: 1.25, 1: 2.5, and 1: 3.75. Thus the optics provide 0.2, 0.4, and 0.6 arcseconds per pixel when used in the Cassegrain mode at the 3M telescope, and 0.6, 1.2, or 1.8 arcseconds per pixel when the camera when used on the 1M telescope. The 1: 1.25 scale of the camera gives a resolving power greater than 60,000 when used with the Coude spectrograph at 1.5 microns. Standard H, J, K as well as narrow band special purpose filters, for spectroscopy, are contained in two independent motorized filter wheels.

We are developing a submillimeter continuum camera for the Caltech Submillimeter Observatory (CSO) located on Mauna Kea. The camera will employ a monolithic Si bolometer array which was developed by Mosley et al. at the NASA Goddard Space Flight Center (GSFC). The camera will be cooled to a temperature of about 300 mK in a ³He cryostat, and will operate primarily at wavelengths of 350 and 450 micrometers . We plan to use a bolometer array with 1 x 24 directly illuminated pixels, each pixel of dimension 1 x 2 mm², which is about half of the F/4 beam size at these wavelengths. Each pixel is 10 - 12 micrometers thick and is supported only by four thin Si legs formed by wet chemical etch. The pixels are doped n-type by phosphorus implantation, compensated by boron implantation. Signals from the bolometer pixels are first amplified by cryogenically cooled FETs. The signals are further amplified by room-temperature amplifiers and then separately digitized by 16 bit A/D converters with differential inputs. The outputs of the A/D converters are fed into a digital signal processing board via fiber-optic cables. The electronics and data acquisition system were designed by the Goddard group. We will report the status of this effort.

Very deep images in the near IR could be recorded from ground based telescopes, if the strong night sky emission lines of OH were blocked and the continuum transmitted. For example, in the H band, of width 330 nm, 87 line groups contribute nearly all the emission. If 1 to 3 nm bandwidth sections of spectrum were removed at each group, the total background--lines plus continuum--would be reduced by 95 to 99.8%, depending on the continuum strength, while the light from the object under observation would be reduced by only approximately 40%. This would greatly enhance sensitivity to faint diffuse objects, such as distant redshifted galaxies. Our camera passes the 2D dimensional image through a large Littrow spectrograph. A long spectrum is formed, such that the full image width projects to a spectral resolution (lambda) /(Delta) (lambda) equals 1500. After reflection at a mirrored `staircase' along the focused spectrum, the light retraces its path through the spectrograph to form a high resolution white image at the original entrance. By placing a comb-like mask along the dispersed spectrum, the OH lines are removed from the rebuilt image, while an optical arrangement involving a polarized beam-splitter efficiently separates the original and rebuilt images. A small scale demonstration in the laboratory with a 10 cm grating proves the method. The source was a resolution chart 6 x 8 mm at f/15. Working in the optical with 10% bandwidth, the spectral resolution was up to 50,000 for individual image resolution elements, 370 for the full rebuilt image. A CCD image of the result image shows 100 x 130 spatial resolution elements. A design optimized for use with a 2 m telescope has a grating size 50 cm x 66 cm and 20% bandwidth. It gives a nominal field of 3 by 5 arcmin and is limited by seeing rather than aberrations. The 2.5 m paraboloidal reflector needed for such an instrument could be built inexpensively as a spinning liquid mirror.

The Optical Aperture Synthesis Technologies project is aimed at the development of technologies for space interferometry. One of the key technologies is the path length stabilization (or cophasing) of light beams from a guide star. A setup has been developed for testing a system for controlling the cophasing (a delay line) in combination with a system for measuring the cophasing (a cophasing interferometer). The light from a simulated guide star passes a delay line system and it is imaged by the cophasing interferometer on a detector. From the captured (white light) fringe patterns the optical path difference (OPD) is determined in real time. The OPD data are used to drive the delay lines in a control loop system for fast stabilization of the optical paths to a sub-wavelength accuracy. The major design drivers for the testbed development were the optimal mechanical and thermal stability (nanometer level), the control loop bandwidth, the OPD stabilization accuracy and the wavefront quality. The instrumentation (optomechanical breadboard and control system) that has been developed for this project is described, and a first set of test results.

The influence of the atmosphere reduces the angular resolution of long exposures taken with large ground-based telescopes. On the other hand, short exposures contain the complete high-resolution information about the observed object in hidden form. Such short exposures are called speckle interferograms because of their appearance. The hidden object information can be deconvolved from a set of such speckle interferograms by various speckle methods. The speckle masking method can even yield images of the observed object with diffraction-limited resolution. The investigation of the required resolution in dynamic range for the digital recording of the speckle images leads to the simple conclusion: one bit is enough to feasibly detect speckle images for the high- resolution methods. In this article, the applicability of the one-bit detector is investigated in detail. This is done by means of an extensive computer simulation, in which first only a simple object structure - a binary - is used. With the dependence of the image quality obtained by bispectral analysis the useful range of the detection threshold is searched for. In these experiments the following observational parameters are varied: the seeing conditions, the separation and the magnitude difference of the binary. Next, the influence of detector noise and of photon noise is considered. Finally, the simulation of the observation of a NGC 3603-like star cluster is performed. This experiment, in which all disturbing effects are included simultaneously, shows clearly the applicability of the one-bit detector to speckle imaging of more complex objects. The advantages for building and using such a detector motivate the transfer of the concept of one-bit correlation - the correlation of interferometric data detected with a dynamic range of only one bit - from astronomical radio interferometry to speckle observations.

Delta-doped CCDs, developed at JPL's Microdevices Laboratory, have achieved stable 100% internal quantum efficiency in the visible and near UV regions of the spectrum. In this approach, an epitaxial silicon layer is grown on a fully-processed commercial CCD using molecular beam epitaxy. During the silicon growth on the CCD, 30% of a monolayer of boron atoms are deposited on the surface, followed by a 15 $angstrom silicon layer for surface passivation. The boron is nominally incorporated within a single atomic layer at the back surface of the device, resulting in the effective elimination of the backside potential well. The measured quantum efficiency is in good agreement with the theoretical limit imposed by reflection from the Si surface. Enhancement of the total quantum efficiency in the blue visible and near UV has been demonstrated by depositing antireflection coatings on the delta-doped CCD. Recent results on antireflection coatings and quantum efficiency measurements are discussed.

Apart from the limitation set by the brightness of the object to be occulted, there are two reasons for the reduced sensitivity of the method of the lunar occultations (LO) in its current implementation. First, the integration of the diffraction pattern over the aperture of the telescope suppresses high spatial frequencies. Secondly, the application of broad-band filters has the same influence due to the wavelength integration. We illustrate a method which is suited to overcome both problems simultaneously. This method (1) increases the sensitivity of LO in terms of both limiting magnitude and angular resolution, (2) is especially suited for very large telescopes in combination with modern 2D detectors, (3) enables the estimation of the brightness profiles as a function of wavelength, and (4) leads to a better resolution of close binaries of different spectral type. A comparison between the application of this novel method and the use of a single element detector is made by numerical simulation.

There is no doubt that astronomy with the `new, improved' Hubble Space Telescope will significantly advance our knowledge and understanding of the universe for years to come. The Corrective Optics Space Telescope Axial Replacement (COSTAR) was designed to restore the image quality to nearly diffraction limited performance for three of the first generation instruments; the faint object camera, the faint object spectrograph, and the Goddard high resolution spectrograph. Spectacular images have been obtained from the faint object camera after the installation of the corrective optics during the first servicing mission in December of 1993. About 85% of the light in the central core of the corrected image is contained within a circle with a diameter of 0.2 arcsec. This is a vast improvement over the previous 15 to 17% encircled energies obtained before COSTAR. Clearly COSTAR is a success. One reason for the overwhelming success of COSTAR was the ambitious and comprehensive test program conducted by various groups throughout the program. For optical testing of COSTAR on the ground, engineers at Ball Aerospace designed and built the refractive Hubble simulator to produce known amounts of spherical aberration and astigmatism at specific points in the field of view. The design goal for this refractive aberrated simulator (RAS) was to match the aberrations of the Hubble Space Telescope to within (lambda) /20 rms over the field at a wavelength of 632.8 nm. When the COSTAR optics were combined with the RAS optics, the corrected COSTAR output images were produced. These COSTAR images were recorded with a high resolution 1024 by 1024 array CCD camera, the Ball image analyzer (BIA). The image quality criteria used for assessment of COSTAR performance was encircled energy in the COSTAR focal plane. This test with the BIA was very important because it was a direct measurement of the point spread function. But it was difficult with this test to say anything quantitative about the aberration content of the corrected images. Also, from only this test it was difficult to measure important pupil parameters, such as pupil intensity profiles and pupil sizes and location. To measure the COSTAR wavefront accurately and to determine pupil parameters, another very important test was performed on the COSTAR optics. A Hartmann test of the optical system consisting of the RAS and COSTAR was conducted by the Goddard Independent Verification Team (IVT). In this paper, we first describe the unique Hartmann sensor that was developed by the IVT. Then we briefly describe the RAS and COSTAR optical systems and the test setup. Finally, we present the results of the test and compare our results with results obtained from optical analysis and from image tests with the BIA.

A high detectivity CCD camera has successfully been used to detect the faint light of a neutron scintillator. We present our first prototype of such a neutron detector with an active area of up to 28*28 cm² and we show several high-resolution radiographs taken with it. After some further improvements, this detector will be sensitive enough to detect a few single neutrons on an area as large as 0.5*0.5 m², thus making it feasible for small-angle scattering with sufficient size and very high resolution.

A new non-destructive readout scheme for IR array detectors, which allows image sharpening by on chip tracking has been tested with the ESO IR array camera IRAC2. This camera is equipped with a large format NICMOS3 256*256 MCT array detector. The effect of the readout algorithm is equivalent to a first order wavefront correction of images degraded by atmospheric seeing. Correction is performed in real time for long on chip detector integrations. Results obtained at a 2.2 m telescope on the pre main sequence binary star S CrA, the Circinus galaxy and the cluster NGC3603 will be presented. Possible applications of this method at large 8 m telescopes will be discussed.

To detect faint signals in the presence of a high background, differential techniques are often used. In IR astronomy this has led to the so-called chopping and nodding techniques. The introduction of array detectors especially for imaging in the thermal IR region of the astronomical spectrum requires an adaptations of these techniques which also takes into account the pixel nonuniformity of the array detector. I describe one possible imaging algorithm and the associated requirements for the chopping mechanisms of an 8-m telescope.