A novel combination of techniques with a proprietary masked CCD allows economical commercial realization of a fast framing camera which can acquire images at 4,000 frames per second while maintaining the low readout noise of a slow scan scientific grade CCD system. A 16 bit gray scale dynamic range has been achieved in production with this system and even higher dynamic ranges appear possible in the near future with this technique. This imaging system can also be combined with a gated image intensifier to provide shutter times in the nanosecond domain.

Test results obtained from two recently developed multiport Charge-Coupled Devices (CCDs) operated at pixel rates in the 10-to-100 MHz range will be presented. The CCDs were evaluated in Los Alamos National Laboratory's High Speed Solid State Imager Test Station (HSTS) which features PC-based programmable clock waveform generation (Tektronix DAS 9200) and synchronously clocked Digital Sampling Oscilloscopes (LeCroy 9424/9314 series) for CCD pixel data acquisition, analysis, and storage. The HSTS also provided special designed optical pinhole array test patterns in the 5-to-50 micron diameter range for use with Xenon Strobe and pulsed laser light sources to simultaneously provide multiple single-pixel illumination patterns to study CCD point-spread-function (PSF) and pixel smear characteristics. The two CCDs tested, EEV model CCD-13 and EG&G Reticon model HSO512J are both 512 X 512 pixel arrays with eight and sixteen video output ports respectively. Both devices are generically Frame Transfer CCDs designed for parallel bi- directional vertical readout to augment their multiport design for increased pixel rates over common single port serial readout architecture. Although both CCDs were tested similarly, differences in their designs precluded normalization or any direct comparisons of test results. Rate dependent parameters investigated include S/N, PSF, and MTF. The performance observed for the two imagers at various pixel rates from selected typical output ports is discussed.

Range image acquisition can be made in many different ways. In this paper we use an active triangulation method where a sheet-of-light illuminates the scene. The sensor-level signal processing task addressed here is to extract the light impact position in each sensor row. We use a thresholding algorithm implemented on the smart sensor MAPP2200 to extract the range data from the sensor data. The range images can be output with very high frame rates (up to and above video rate). The programmability of the sensor makes it possible to generate range images with arbitrary range resolution or image size. Reducing the image size or resolution gives us higher range image frequency. We also present an error correction scheme not found in any other smart sensor implementation.

Video baseline drift in CCD cameras with variable frame formats is often difficult to correct. Conventional baseline restorers generally do not work well at very high frame rates, since frequently only a short, few nanosecond wide point in each video line is available for reference. A.C. coupled video level drifts as the signal average changes with illumination. Also, low frequency noise (mainly from power supplies) modulates the video signal, especially when high video amplifier gains are required. On the other hand, although D.C. coupled video amplifier output does not drift due to the change in CCD illumination, it does drift as a function of temperature. Also, externally generated noise is directly superimposed on the video signal, reducing the image quality. Either case requires some kind of baseline restoration in order to keep the signal aligned to the desired reference level. This requirement is imperative during the process of testing and evaluating state of the art multiport CCD sensors. We describe a simple method for baseline restoration of A.C. coupled video signals, using a fast voltage comparator and a gated latch. An external strobe pulse must be provided and timed precisely at the point in video line selected as a reference. This reference is compared with a preset external D.C. level. The comparator output would therefore go to its high state when the reference signal is below that level. The comparator output status is latched into a D-type flip-flop by strobe pulses, provided externally, and timed to be coincident with the video line reference points. The inverted output of the latch is used to either charge or discharge the video signal coupling capacitor with a constant current. It tries to shift the video signal reference point, after each strobing pulse, closer to the comparator D.C. bias voltage, in a characteristic zigzag pattern. Best results are achieved when the rate of coupling capacitor voltage change, done by the restorer, matches the baseline drift rate.

A compact and fully integrated, high-frame-rate and high-resolution digital image acquisition and analysis system has been developed. The system integrates high-speed data acquisition, image playback, image processing, and motion analysis features. The imaging device, a multiport, split-frame-transfer, backside-illuminated charge-coupled device (CCD) imager developed by the Army Research Laboratory and the David Sarnoff Research Center, enables system operation at up to 500 frames per second. The system design is described and CCD performance results are given.

The RADAC (Recovery of Attitude and Deformation by Crossed Anamorphoses) system was developed to enable precise measurement of wind tunnel model deformations under aerodynamic loads. The system was developed by the Centre d'Etudes et de Recherches de Toulouse to meet high Reynolds Number testing requirements of the ONERA large wind tunnels. This document presents a summary of the measurement principle. The design and implementation of the system for use in the ONERA S1 wind tunnel is then discussed. Finally, experimental results from development tests and qualification testing in the S1 tunnel are presented.

Los Alamos National Laboratory (LANL) has designed and prototyped high-frame rate intensified/shuttered Charge-Coupled-Device (CCD) cameras capable of operating at Kilohertz frame rates (non-interfaced mode) with optical shutters capable of acquiring nanosecond-to- microsecond exposures each frame. These cameras utilize an Interline Transfer CCD, Loral Fairchild CCD-222 with 244 (vertical) X 380 (horizontal) pixels operated at pixel rates approaching 100 Mhz. Initial prototype designs demonstrated single-port serial readout rates exceeding 2.97 Kilohertz with greater than 5 lp/mm spatial resolution at shutter speeds as short as 5 ns. Readout was achieved by using a truncated format of 128 X 128 pixels by partial masking of the CCD and then subclocking the array at approximately 65 Mhz pixel rate. Shuttering was accomplished with a proximity focused microchannel plate (MCP) image intensifier (MCPII) that incorporated a high strip current MCP (28 uA/sq.cm) and a LANL design modification for high-speed stripline gating geometry to provide both fast shuttering and high repetition rate capabilities. Later camera designs use a close-packed quadrupole head geometry fabricated using an array of four separate CCDs (pseudo 4-port device). This design provides four video outputs with optional parallel or time-phased sequential readout modes. Parallel readout exploits the full potential of both the CCD and MCPII with reduced performance whereas sequential readout permits 4X slower operation with improved performance by multiplexing, but requires individual shuttering of each CCD. The quad head format was designed with flexibility for coupling to various image intensifier configurations, including individual intensifiers for each CCD imager, a single intensifier with fiber optic or lens/prism coupled fanout of the input image to be shared by the four CCD imagers or a large diameter phosphor screen of a gateable framing type intensifier for time sequential relaying of a complete new input image to each CCD imager. Camera designs and their potential use in ongoing military and medical time-resolved imaging applications are discussed.

A high resolution fast readout camera system has been developed around a specialized full frame 512 * 512 pixel format CCD sensor in conjunction with EEV Ltd. An equivalent readout rate of 600 frames per second with a spatial resolution of 24 line pairs per mm at the Nyquist limit is achieved. A binning capability increases the readout rate by up to a factor of four. The system dynamic range is in excess of 1000:1. System features include image analysis for computer with dedicated software, non volatile memory, high resolution and real time displays, instantly reset able, single shot image capture capability and remote head operation.

High speed x-ray imaging can be an important tool for observing internal processes in a wide range of applications. In this paper we describe preliminary implementation of a system having the eventual goal of observing the internal dynamics of bone and joint reactions during loading. Two Los Alamos National Laboratory (LANL) gated and image intensified camera systems were used to record images from an x-ray image convertor tube to demonstrate the potential of high frame-rate digital radiographic videography in the analysis of bone and joint dynamics of the human body. Preliminary experiments were done at LANL to test the systems. Initial high frame-rate imaging (from 500 to 1000 frames/s) of a swinging pendulum mounted to the face of an X-ray image convertor tube demonstrated high contrast response and baseline sensitivity. The systems were then evaluated at the Motion Analysis Laboratory of Henry Ford Health Systems Bone and Joint Center. Imaging of a 9 inch acrylic disk with embedded lead markers rotating at approximately 1000 RPM, demonstrated the system response to a high velocity/high contrast target. By gating the P-20 phosphor image from the X-ray image convertor with a second image intensifier (II) and using a 100 microsecond wide optical gate through the second II, enough prompt light decay from the x-ray image convertor phosphor had taken place to achieve reduction of most of the motion blurring. Measurement of the marker velocity was made by using video frames acquired at 500 frames/s. The data obtained from both experiments successfully demonstrated the feasibility of the technique. Several key areas for improvement are discussed along with salient test results and experiment details.

We have found that a potassium-iodide photocathode of an x-ray streak camera responds to UV light at λ=308 nm. The photocathode surface work function, 6.5 eV, is larger than the 4 eV energy of the UV photon, hence the source of the response is interesting. We will present results on the response of a transmission type potassium-iodide photocathode to the UV light from a λ=308 nm, subpicosecond XeCl laser and from a λ=325 nm HeCd laser. We will test for the nonlinearity of the yield by measuring the number of photons that are needed to be absorbed before a signal is recorded. We will present data on the effect of the UV irradiance on the yield, as well as on the temporal width of the recorded signal. We will give an explanation of the observation and its effect on the dynamic- range response of the streak-camera. We will show that the response is linear with the incident irradiance, up to an incident irradiance of 10⁸ W/cm² and we will explain the observation.

This paper deals with an experimental and numerical approach of laser induced dielectric breakdown in a liquid. The experimental set-up and especially the high speed visualization device involved, are presented. Results on physical characteristics of the phenomena are obtained with the visualization techniques and compared with numerical investigations.

Microchannel plates (MCPs) are incorporated in a wide variety of x-ray imaging and detection devices. Experimental measurements are presented in this paper which are used to determine the MCP and phosphor configuration for optimizing spatial resolution, temporal resolution, and gain in an x-ray framing camera. We investigate with the use of pulsed electron accelerating voltages to attain extraction fields higher than those possible at safe DC levels. In addition, MCPs with exit faces coated with an insulating material, in order to increase the maximum safe DC electron extraction field, were also tested.

This paper focuses on high speed electronic/electro-optic camera development by the Applied Physics Experiments and Imaging Measurements Group (P-15) of Los Alamos National Laboratory's Physics Division over the last two decades. The evolution of TV and image intensifier sensors and fast readout fast shuttered cameras are discussed. Their use in nuclear, military, and medical imaging applications are presented. Several salient characteristics and anomalies associated with single-pulse and high repetition rate performance of the cameras/sensors are included from earlier studies to emphasize their effects on radiometric accuracy of electronic framing cameras. The Group's test and evaluation capabilities for characterization of imaging type electro-optic sensors and sensor components including Focal Plane Arrays, gated Image Intensifiers, microchannel plates, and phosphors are discussed. Two new unique facilities, the High Speed Solid State Imager Test Station and the Electron Gun Vacuum Test Chamber are described. A summary of the Group's current and developmental camera designs and R&D initiatives are included.

We describe a wide band optoelectronic device for short electrical single pulses sampling. The devices are made with microstrip propagation lines on a sapphire substrate and ultra-rapid photoconductive switches are integrated on the same substrate (synchronously to the electrical pulse, a femtosecond laser pulse is used to switch the photoconductive elements). The switches are made of thin semiconductor films deposited by molecular beam epitaxy at low temperature. The propagation on the stripline structure has been simulated and characterized. A 16 channels prototype with 18 ps sampling step has been realized and tested. The analysis of one sub- 50 ps electrical transient has been demonstrated. This type of device should easily allow to reach sampling steps down to 5 ps.

The Department of Transportation, National Highway Traffic Safety Administration (NHTSA) desires to develop qualifications for the film received from independent organizations that perform automobile collision performance tests. All crash tests are recorded on high-speed film. Running 18 cameras at frame rates up to 1000 frames per second is not uncommon. Some of the films acquired during the 150 to 200 ms time-frame of the actual crash are used for computation of both human and vehicle kinematics. Detailed recommendations for performing the tests are outlined by NHTSA in SAE documents. However, no specifications for film quality are defined. Problems arise when unclear and/or incorrectly exposed films result in images unsuitable for analysis. Aspects of the optical data channel that are evaluated in this paper include lighting, lenses, cameras, film, film processing, and timing. Recommendations for reliable data acquisition as well as a set of criteria are also developed.

Miniature laser-launched plates have applications in shock wave physics, studying dynamic properties of materials and can be used to generate experimental data in a manner similar to a laboratory gas gun for 1D impact experiments. Laser-launched plates have the advantage of small size, low kinetic energy, and can be launched with ubiquitous laboratory lasers. Because of the small size and high accelerations (10⁷ - 10¹⁰ g's), improved temporal resolution and optical non-contact methods to collect data are required. Traditional mechanical in-situ gauges would significantly impair the data quality and do not have the required time response.

The ability to image high velocity projectiles requires a precise trigger, short gate periods and known timing data for performance measurement. Data analysis software provides a means for data reduction to useable numerical ballistic parameters. The software relies on knowns of image format, image frame rate, gate period, and timing data for its analytical capability. Prior to applying the software to the data analysis task, the basics of image capture and data annotation must be established to provide a suitable basis for software analysis of ballistic performance. Empirical data can be obtained through the use of simple tools to properly understand the image capture dynamics and establish a precise basis for ballistic analysis.

Over the years `synchro-ballistic' photographic techniques have been used to aid the scientist and engineer in obtaining data of fast moving events. With the availability of `gated video' a new technique, `Ballistic Videography' has emerged. The purpose of this discussion is to review the principles of synchro-ballistic photography and discuss the use of high speed gated video to simplify and increase date yield.

Laser driven plates have been used for several years for high velocity shock wave and impact studies. Recent questions about the integrity and ablation rates of these plates coupled with an improved capability for microscopic stop motion photography led to this study. For these experiments, the plates were aluminum, coated on the ends of optical fibers. A high power laser pulse in the fiber ionizes the aluminum at the fiber/coating interface. The plasma thus created accelerates the remaining aluminum to high velocities, several kilometers per second. We defined `thick' or `thin' coatings as those where a flying plate (flyer) was launched vs. the material being completely ionized. Here we were specifically interested in the thick/thin boundary to develop data for the numerical models attempting to predict flyer behavior.

When created in 210 BC, the 6,000 terra cotta warriors of the Mount Li tomb near the ancient imperial Chinese capital city of Xi'an were emblazoned with dramatic colors. Previously, it had been thought that this polychrome glaze had been entirely consumed in an inferno that swept the tomb shortly after its completion. Careful manual control of laser divestment experiments on the statues during the archaeological excavation of the site has revealed that faint charred vestiges of the original polychromatic glaze still exist. Unfortunately, as a consequence of the long burial, the terra cotta statues with their cooked polychrome glazes are encased in tenacious mineral deposits. Consequently, laser removal of the mineral encrustation requires tedious human implementation in order to avoid damaging the very delicate underlying statue surfaces. A real-time spectral control system for precise automated laser cleaning of archaeological objects is described. It has improved the results, accelerated the process, and relieved the operator tedium in the recovery of this Qin Dynasty polychrome as well as in the cleaning of ancient coins from other archaeological sites.

A proximity focused microchannel plate image intensifier (MCPII) with low inductance planar-lead gating electrodes was successfully operated as an optical shutter with a 220 ps full- width gate image. The MCPII (made by RTC Phillips) was gated on with single-shot, 290 ps FWHM, -500 V pulses from a 50 (Omega) source impedance pulse generator. The limiting resolution of the MCPII's output image in the center of the gate pulse was 8 lp/mm. The characteristic impedance of the 18 mm diameter MCPII was calculated to be 4.5 (Omega) . High-bandwidth microstrip impedance-transforming transmission lines were made out of a substrate with a dielectric constant of 10.2 to interface between the 50 (Omega) impedance of the gate pulser and the 4.5 (Omega) impedance of the MCPII. The shape and width of the microstrip transformer's signal conductor, the lengths of the transformer's sections, and the linearity of the impedance changes were all modeled and two designs were fabricated and tested for comparison along with a shorter, straight-taper design originally used several years ago. The design resulting in the lowest pulse dispersion and the fastest gating speed was one where the impedance changed linearly with distance.

In this paper copper phthalocyanine (CuPc) has been proposed as a potential optical data storage medium at short laser wavelength. Both of the static and dynamic optical recording tests have been performed and the recording mechanism has been analyzed.

We have implemented two different types of 40 ps framing camera configurations on the Nova laser system. We will describe the configurations of both systems as well as discuss the advantages of each.

Large dynamic range (> 1000) measurements of gain profiles for gated microchannel plates are presented. Temporal profiles are reconstructed for any point on the microstrip transmission line from data acquired over many shots with variable delay. No evidence for significant pulse distortion by voltage reflections at the ends of the microstrip is observed. The measured profiles compare well to predictions by a time-dependent discrete dynode model down to the 1% level. The calculations do overestimate the contrast further into the temporal wings. The role of electron transit time dispersion in limiting the minimum achievable gate duration is then investigated by using variable duration flattop gating pulses. A minimum gate duration of 50 ps is achieved with flattop gating, consistent with a fractional transit time spread of approximately equals 15%.

The THX 7887A sensor developed by Thomson-CSF Semiconductors Specifiques is a high resolution frame transfer CCD designed for medical applications (digital X-ray imaging: fluoroscopy and angiography for instance), for industrial (machine vision, pattern recognition or process control, ...) and scientific applications (e.g., optical correlation...). The design of the array is discussed: it consists in 1024 X 1024 photosensitive pixels of 14 micrometers X 14 micrometers . It features on-chip memory and four parallel outputs running at 20 MHz maximum data rate each, thus compatible with frame rates as high as 60 Hz (one frame is made of 1024 X 1024 true pixels without any interlacing). Other specific on-chip capabilities are available: binning 2 X 2 pixels (leading to 512 X 512 or 512 X 1024 format), anti-blooming and exposure time control. Design has been optimized so as to minimize high speed clock number and to facilitate driving of the chip. Main attractive performances measured are low noise and dynamic range higher than 12 bits at 60 frames per second, together with low dark signal, high responsivity and very low optical/electrical crosstalks.

We already presented several times at this conference the status of the ongoing development of Hybrid Photomultiplier Tubes C or Hybrid Photodiodes ). Apart from a quick review of the basic principle, this presentation reports the latest results. Many people and groups have joined the HPD collaboration over the last four years. Presently the majority of the work is done by CERN and INFN (evaluation, application), Canberra (Diode development and manufacturing) and DEP (tube development and manufacturing).

The review of results on magnetic field investigation in high-temperature laser produced and Z-pinch plasmas is presented. These investigations were carried out by means of diagnostics method, based on Faraday rotation of probing beam polarization plane. The specially developed three-channel polarointerferometer in composition with high-speed streak-cameras were used.

Eastman Kodak Company Motion Analysis System Division has investigated many years in developing technology used in our new 24 bit color accurate digital motion analyzer. This paper will describe the method for producing accurate 24 bit color at 1000 (frames per second). It will describe the systems significant imaging performance and new tools for motion analysis.