The stature of a technical journal is determined by the quality of the papers it publishes, which in turn depends upon both the quality of the papers submitted and the quality of the reviews those papers receive. in this editorial I offer a number of suggestions that I hope will assist referees of Optical Engineering papers to provide objective, thorough, and helpful reviews that I can then use in reaching a decision to publish or reject and that authors can use to revise and improve their papers.

Astronomical Optics-Reviewed by Robert R. Shannon

Multiple-aperture optics have certainly come a long way since the Michelson interferometer. Yet Michelson's impressive optical interferometry experiment at Mount Wilson laid a solid foundation for applications to follow. Imaginations of optical scientists were opened to new opportunities presented by multiple-aperture optical configurations. Long baseline and ultralarge effective aperture optical systems were now a possibility for high resolution astronomy and imaging. Dilute apertures in trigued clever thinkers such as Golay. Synthetic apertures at infrared and optical wavelengths became a serious topic of research and development.

The first successful demonstration of actively phase matching, coaligning, and agilely steering a number of mutually coherent optical wavefronts through an array of transmitting telescopes has been performed on the Phasar experiment at the Air Force Weapons Laboratory. In the Phasar experiment, modern wavefront sensing and feedback techniques are used to control wavefronts transmitted through an assembly of three independent optical telescopes. Mutually coherent wavefronts transmitted through the telescope array are controlled to produce an effective single aperture wavefront. The multiple phased wavefronts produce a far-field energy distribution equivalent to the diffraction pattern of a single aperture whose diameter and obscurations match the geometry of the transmitting array. The Phasar experiment demonstrates absolute phasing of a broadband source in the presence of dynamic disturbances. The technologies developed facilitate the construction of large aperture, high quality optical systems by the modular combination of relatively small, inexpensive telescopes.

A description is presented of a system designed to control the path length and tilt components of the optical wavefronts in a multiple-aperture transmitting system. The system has been implemented in the three-telescope Phasar testbed at the Air Force Weapons Laboratory. The system hardware and software are described, and measurements of the system performance are presented. The measurements indicate that under laboratory conditions the current system can maintain an optical path length difference between pairs of telescopes of approximately X/15 and a tilt control error on each telescope of about 40 nrads. The band-widths of the path length control and tilt control subsystems have been measured at 130 Hz and 950 Hz, respectively.

We present a description of a phased array system using the far-field point spread function to estimate the tilt and piston errors. Digital control electronics were utilized to phase one beam to two others that were already in phase to simulate a multiple-aperture telescope system. A control algorithm for estimation of tilt and piston errors similar to the Newton-Raphson algorithm for locating the roots of a function was implemented. A simulation of the system was performed, the system hardware was built, and the system was tested in both an off-line software test and an experimental test-bed. Piston phasing with a capture range of X/4 and closed-loop tilt phasing with a bandwidth of 2 Hz were accomplished.

A multiple telescope array, Phasar, is used to collect light for a phase-conjugate mirror of barium titanate, forming a hybrid receiving/ transmitting system. The observed fidelity of the system output is independent of the type and amount of aberration introduced into the optical path. In all test cases, the measured far-field peak intensity indicates that the fidelity of the return from the three telescopes approaches the fidelity required to yield the ideal in-phase superposition of waves in the far field. The system demonstrates high fidelity phase conjugation for a telescope piston mismatch of up to ±5 Am, for tilt error up to ±700 grad, and for higher order aberrations. A scheme implemented to compensate for depolarizing and polarization-altering effects boosts the amplitude of the point spread function considerably. Near-field diagnostics confirm the fidelity of the phase-conjugate output.

The basic principles for designing and implementing multiple telescope imaging arrays are presented. A generalized treatment is formulated that permits determination of the amount of residual errors that can be tolerated in lateral and longitudinal pupil geometry matching, magnification matching, azimuthal asymmetry, and telescope boresighting. A specific design goal is a modular system of independent telescopes configured to have the equivalent performance of a monolithic aperture of the same diameter as the array. This telescope array, known as the Multipurpose Multiple Telescope Testbed (MMTT), is designed to provide high resolution imagery over a wide field of view. The requirement for wide field of view performance imposes rigid tolerances that must be satisfied at every stage of development. The work described here derives the subsystem tolerances that must be met. Special means by which some of the more rigid requirements can be satisfied are also briefly described.

The design of the Multipurpose Multiple Telescope Testbed (MMTT) is described. The MMTT, whose construction is now nearly complete, is believed to be the first wide field of view phased array imaging telescope. The design philosophy and alignment requirements are discussed and the telescope is described. The array consists of four 20 cm telescopes and operates with visible light. The coherently combined image has a 30 arcmin field of view. A key design goal is the maintenance of lateral pupil geometry to within 1µm of the ideal. Two mirrors behind each afocal telescope combine the four beams into one and are actuated to control both tilt and piston. Together, the two beam combiner mirrors control lateral pupil geometry. A control system senses the lateral pupil geometry error by continuously measuring piston phase changes over the telescope field of view. A discrete time varying Kalman filter then processes these measurements to estimate lateral pupil geometry errors.

This paper presents the issues pertinent to a multivariable control system for a phased telescope array imaging system. A description of the envisioned system, the discrete time state space model of the system, the measurement techniques, estimation/control algorithms, and Monte Carlo simulation results are presented. Initially, an extended Kalman filter is used as the estimator and its simulation performance is compared with that of a linearized Kalman filter with no difference in performance found. Additionally, successful test results from a laboratory two-beam prototype control system testbed are presented.

An array of four telescopes has been fabricated to sufficient precision and stability to serve as a testbed for phased array imaging experiments over a wide field of view. The overall optical and mechanical design philosophy is given as well as the analysis of fabrication and assembly tolerances. Finally, the error budget is developed for the entire system.

We describe the digital optical control system (DOGS), a state-of-the-art controller for electrical feedback in an optical system. The need for a versatile optical controller arose from a number of unique experiments being performed by the Air Force Weapons Laboratory. These experiments use similar detectors and actuator-controlled mirrors, but the control requirements vary greatly. The experiments have in common a requirement for parallel control systems. The DOGS satisfies these needs by allowing several control systems to occupy a single chassis with one master controller. The architecture was designed to allow upward compatibility with future configurations. Combinations of off-the-shelf and custom boards are configured to meet the requirements of each experiment. The configuration described here was used to control piston error to X/80 at a wavelength of 0.51 Am. A peak sample rate of 8 kHz, yielding a closed loop bandwidth of 800 Hz, was achieved.

This paper describes the absolute phasing of two adjacent mirror panels using a 10 W tungsten lamp, a slit, a linear charge-coupled device array, and an oscilloscope. Repeatability of measurement settings equivalent to five one-thousandths of a wavelength was demonstrated using only visual observation of the oscilloscope and a calibrated phase changer. The indicator of phase used in these tests is the intensity distribution across the image, similar in this respect to the Phasar method developed by the Air Force Weapons Laboratory. The phasing of circular apertures as well as small subapertures is shown. A simple near-linear phase changer, to change phase manually by less than one thousandth of a wave, was developed to determine precision and accuracy of the phasing method.

Thinned aperture optical systems (including phased telescope arrays) pose unique problems in specifying or characterizing image quality. Traditional image quality criteria such as "resolution" and encircled energy are woefully inadequate for many thinned aperture applications. Variations in the subaperture geometry that produce only subtle effects upon the core of the point spread function may produce highly undesirable artifacts or spurious images as well as a modulation transfer function (MTF) that exhibits zero (or negligible) values over substantial regions within the cut-off spatial frequency of a filled aperture circumscribing the array. Clearly, some minimum value of the MTF exists below which spatial information cannot be retrieved in the presence of noise. An MTF property called the "practical resolution limit," defined as the reciprocal of the maximum spatial frequency within which no zeros occur in the MTF, thus becomes the image quality criterion of choice for those applications in which fine detail is required from extended objects. This practical resolution limit and its effect upon subaperture configuration will significantly impact the telescope mechanical architecture, stowage and deployment techniques, and perhaps even the booster vehicle selection for future large space telescopes.

McDonald Observatory has joined in the effort to produce a workable design for a spectroscopic survey telescope utilizing a segmented spherical primary mirror, elaborating on ideas initiated by astronomers at the Pennsylvania State University. The spherical primary figure requires that a secondary focus assembly be driven at the tracking rate in an attitude normal to the spherical focal surface while the telescope as a whole, being tilted at a predetermined angular zenith distance, need only be "set" (and clamped) occasionally in azimuth. The spherical primary mirror segments, all figured to an identical radius of curvature, are potentially simpler to manufacture than off-axis aspheric segments. The overall diameter of the primary (that is, the diameter of the smallest circle that circumscribes the entire set of segments) is approximately 10 m, enough to provide a significant aperture increase for current spec-troscopic research. The radius of curvature is 26 m. The glass segments are supported on a fully triangulated space frame of Invar and 304 stainless steels. A structural analysis using standard techniques of finite elements shows that the expected static performance of both the individual segments and the overall space frame present reasonable goals to be achieved by modern engineering design practice. The design has been implemented in a full-scale prototype consisting of a seven-segment central portion of the telescope's primary mirror structure.

Multiaperture vision systems are of interest in the area of robotic vision. A facsimile of the insect eye may be as useful for transmitting information to robotics as conventional cameras, which are analogous to the human eye. A physical model of one multiaperture vision system, the insect apposition ommatidia, has been successfully constructed. To provide an understanding of the multiaperture vision system, the impulse response of a single eyelet, or ommatidium, was developed. Experimental measurements of the impulse response were found to closely match theoretical results.

Imaging correlography is a technique for constructing high resolution images of laser-illuminated objects from measurements of back-scattered (nonimaged) laser speckle intensity patterns. In this paper, we investigate the possibility of implementing an imaging correlography system with sparse arrays of intensity detectors. The theory underlying the image formation process for imaging correlography is reviewed, emphasizing the spatial filtering effects that sparse collecting apertures have on the reconstructed imagery. We then demonstrate image recovery with sparse arrays of intensity detectors through the use of computer experiments in which laser speckle measurements are digitally simulated. It is shown that the quality of imagery reconstructed using this technique is visibly enhanced when appropriate filtering techniques are applied. The signal-to-noise ratio of the process and its dependency on array redundancy and number of speckle pattern measurements is also discussed.

This paper discusses the system design of a wavefront sampling system for a four-telescope optical phased array. This sampling system provides the input to an electromechanical feedback control system that dynamically phases the telescope array. We begin by presenting the basics of the optical phasing problem for a telescope array and then discuss design solutions that provide the necessary wavefront sensing for a feedback control system. The overall system design concentrates on simplicity despite the number of channels, while the optical design embodies a philosophy of invariance to field of view, the use of the first-order optical properties of the entrance/exit pupil relationship, and the exploitation of beam polarization for power conservation and electro-optical chopping.

This paper presents results of a study of a new multiple aperture imaging technique, including a proof-of-concept simulation. The result of the study is a large aperture, imaging laser radar receiver concept that can be used to access multiple targets in a large field of view with no mechanical motion. The receiver diameter is not limited by current optical fabrication techniques. The proof-of-concept simulation incorporates the effects of background radiation and detector shot noise, which set requirements for the receiver subaperture components and the laser transmitter power. The multiple aperture receiver concept involves the area illumination of a target scene with a coherent, multiple pulse laser transmitter. The re-flected radiation from the scene creates a series of speckle patterns at the plane of the multiple aperture receiver, which are sampled by the sub-apertures and then processed to produce an estimate of the image auto-correlation. Background suppression is achieved by a combination of spectral, temporal, and spatial filtering. By Fourier transforming the image autocorrelation, we produce an estimate of the image power spectral density, which is low-pass filtered to minimize noise and then used to reconstruct the image using phase retrieval techniques. Examples of this reconstruction process are presented.

An automatic endoscope, called a rhizoscope, has been developed to study the growth of plant roots in the soil. Unlike conventional methods, this method is nondestructive for cultivated plants. This work has required the development of an original electro-optical sensor and an automated driving system. The rhizoscope is operated by a microcomputer. The root images are stored on floppy disk for off-line processing and analysis. Preliminary test results, obtained by simulation and in a greenhouse, are presented.

A general iterative method of restoring linearly degraded images [R. J. Mammone and R. J. Rothacker, J. Opt. Soc. Am. A4(1), 208-215 (1987)] has been reformulated into a more tractable fixed point iterative procedure. The new formulation is an implementation of the steepest descent algorithm. The slow convergence of the original method is found to be due to its inherent step size. A new method is presented whose increased step size offers accelerated convergence. The realization of the accelerated method is shown to require only a minor modification of the original algorithm. A new stopping criterion is also introduced. Computer simulations demonstrate a significant improvement in the rate of convergence of the new method.

The power penalty estimation for fiber optic communication due to mode partition noise in nearly single mode lasers is discussed. Nearly single mode lasers are modeled here as those with a single dominant mode and n side modes with significant strength. The dropout rate of each side mode obeys exponential distribution. The combined distribution of the dropout rate for n side modes is derived here. This distribution is combined with the statistically independent Gaussian receiver noise in estimating error probability and hence the power penalty. Calculations show that to ensure a 10 -12 error rate, the ratio of the main mode to the side modes must be greater than 57:1 for one side mode and greater than 67:1 for two side modes.

An account is given of a realization of a feedback method to digitize the analog position signal from a moving iron galvanometer. It is employed in a confocal scanning laser microscope for generating digital images. The photometric sampling has to be closely coupled to the position of a mirror that scans a focused laser beam across a microscope specimen. Pictures with low geometric distortion are obtained up to the size 1024 x 1024 pixels.

A laser measurement system for precise and fast positioning of an object has been developed. When the object speed is low, the movement of the object is measured by comparing the phase change of a light beam reflected by the object with the phase modulated by an electro-optic crystal (an active device by which the optical phase can be controlled by the application of a voltage). When the object speed is high, the movement is measured by a fringe counting technique. The system achieves an accuracy of 4 nm and a maximum allowable measurement speed of 1100 mm/s.

Performance data for a cw Nd:BEL laser end pumped by two 500 mW laser diode arrays are presented and compared with the performance data for a Nd:YAG laser in a similar configuration. The phased arrays used as the pump source are cooled by a temperature-controlled heat sink to permit wavelength tunability. Although the absorption band-width for Nd:BEL is substantially broader than for Nd:YAG, the Nd:BEL was found to have a higher threshold for lasing. Both rods gave slope efficiencies of 42%. The dependence of the output power on output mirror reflectivity was measured, with Nd:BEL showing a greater sensitivity to reflectivity than Nd:YAG. The optimum reflectivities were found to be 0.98 for Nd:BEL and 0.97 for Nd:YAG. The maximum TEM00 cw power achieved for each rod at these reflectivities was 250 mW for Nd:BEL and 283 mW for Nd:YAG. The observed electrical-to-optical conversion efficiency was factored into a product of analytic component terms, and excellent agreement was found between observed and calculated efficiencies. It is concluded that under the conditions used in this work, both BEL and YAG hosts perform comparably. The conditions under which BEL might out-perform YAG as a host for diode-pumped Nd are discussed.