Search

The calibration of densitometers and microdensitometers is discussed in terms of the physical standards necessary to accomplish it. The need for calibration of these instruments is briefly considered, particularly for those forensic investigators looking at evidential photographic materials with a microdensitometer. The calibration process is reviewed for both instruments to prepare for a new procedure to follow. Since a great deal of microdensitometer work is carried out on materials about whose characteristics nothing is known, the problem of obtaining data in terms of diffuse (visual) density is almost impossible. However, a bootstrap calibration, based on the scan of a standard step tablet and scans of selected portions of the specimen is shown to be feasible. It requires a calibrated densitometer to quantify the results and has reasonable accuracy within the necessary assumptions. An example is given.

The day-to-day evaluation of a scanner system’s performance, a problem in quality assurance, is addressed by deriving a technique for establishing it’s system cutoff spatial frequency through the analysis of the edge slope. It is shown that the slope value at the midpoint of the system’s imaged edge (edge response) is related (and, for some conditions, equal to) its cutoff frequency. From the analysis, a parameter emerges that describes the sampling aperture as ‘‘hard’’ or ‘‘soft,’’ signifying whether its shape is best described by geometrical or physical optics, respectively. The physical implications of this description are discussed and results are substantiated experimentally using a microdensitometer.

This paper attempts an analytical treatment of flare light in microdensitometers by developing a model based on scattering from the air-glass interface. Verification of the model is conducted via a breadboard mockup on an optical table, and on an actual microdensitometer. It is, seen that the simple model predicts values for flare irradiance well within an order of magnitude. It is further shown that a sample trace is not only a convolution of the target with the projected sensor aperture but, to some extent, a convolution of the target with the irradiated area in the sample plane as well. Based on this analysis, it is shown that flare can be corrected for provided its magnitude is held constant. This requires that, during a full scan, the target always remains effectively within the irradiated area.

The Modulation Transfer Function (MTF) of an x-ray focal spot has been obtained without the aid of a microdensitometer. The modified method described in this study utilizes radiographic magnification techniques to eliminate the need for a microdensitometer. In this study a radiographic magnification factor of 25X or higher has been employed to obtain the slit-camera images of x-ray focal spots. The Line Spread Function (LSF) is obtained by scanning the focal spot slit-camera image acquired with the modified method on an ordinary densitometer equipped with a 0.2 mm aperture. The MTF is subsequently calculated from the LSF and compared with the MTF that was obtained by means of the traditional method. The results show good agreement between the MTFs obtained by these two methods.

A survey is given on the principle, the theory and the practical performance of an instrument that measures transmittance modulation over an extremely wide range of spatial frequencies. In a conventional microdensitometer, a microscope lens magnifies a small section of the sample onto a narrow single slit in front of the light detector. The Multiple-Sine-Slit Microdensitorrteter MSSM uses laser two-beam interference to project a sinusoidal intensity distribution of a great number of periods directly onto the sample. The total light flux transmitted while interference fringes scan across the transmittance pattern is collected by an integrating sphere in contact with the sample. The absence of any imaging elements gives the MSSM a constant instrument MTF over its entire working range, which at present spans from 4 to 1500 cycles/mm. By the large scanning area and the great number of periods in the scanning function a narrowband characteristic is achieved that efficiently suppresses grain noise and improves signal-to-noise ratio. Due to these features, MTF's could be determined up to 300 cycles/mm for coarse-grained emulsions and up to 1500 cycles/mm for high resolution emulsions as they are used for holography, optical data storage and integrated circuit manufacture.

In the development of modern linear microdensitometry, the trend of optical system design has been towards the condition of underfilling the efflux optics, with total collection of light after it passes through the sample. The system transfer function is therefore attributable to the influx optics, and the sampling aperture is the light distribution impinging on the sample, reduced from an illuminated slit or circular aperture through the influx optical system. The maximum frequency response of the system is obtained when the sample is illuminated with the impulse response of the influx optics. However, the theoretical impulse response can only be realized by imaging a delta-function and this is photometrically impossible. Similarly, because the system images an illuminated aperture onto the sample, scanning with a pure, geometrically-characterized slit or spot is not possible due to lens response and diffraction. These two problems are investigated, for both coherent and incoherent illumination. For both impulse response and slit image, the MTF is investigated, and its deviation from the ideal is calculated. The results are characterized in terms of RMS-MTF differences over the spectrum for 10, 5, 2 and 1% levels. The controlling parameter is the ratio N/R, where R is the reduction factor employed for the influx optics, and N is the number of resolution elements contained within the nominal object slit width. The study shows that there are significant differences in these values for the same RMS difference level, with coherent and incoherent illumination, and that there are compromises to be made with both kinds of illumination. The results of this study facilitate calculation of system response for any configuration of object slit and influx optics (within the linear microdensitometer system), and defines limits on slit sizes for operation with impulse response and geometrically characterized slit images for the sampling aperture. The effects expected with the linear polarization associated with laser illuminat

Linear systems methodology is proposed which quantifies flare as part of the spread function or Modulation Transfer Function (MTF) of a microdensitometer with general application to any optical detection system. Because the irradiance system, the sample's scattering properties and the optical components alter the flare portion of these functions, an empirical method, such as edge-gradient analysis, should be used to determine their profiles. Applying straightforward linear systems theory, one has a method of correcting for flare which is usually considered to be a "nonlinear and unpredictable" factor. More importantly, this method is applicable for any object distribution (for a given optical system and sample scattering profile). Partial spread functions and partial MTFs are introduced which permit quantifying the degradation produced by flare, this quantity being denoted as the partial flare factor. The partial flare factor is compared to a more restrictive black-overlay flare factor. Examples are discussed for optical detection systems with and without flare.

The capability of high-resolution microdensitometers is cur-rently limited by the inability to focus accurately and to maintain focus under dynamic conditions. This paper describes a device for establishing and maintaining submicrometer focus tolerances under dynamic operating conditions that would normally result in gross defocus. The device utilizes pneumatic operating principles in a closed-loop feedback system. The precision and repeatability of establishing the focal position is such as to virtually eliminate the need for visually obtaining best focus. The manual through-focus positioning traditionally done by the operator is eliminated. The focus position is established through a gauge indicator needle that is positioned in submicrometer increments from slightly above the target surface to 20 micrometers below. The system is capable of dynamically maintaining focus over a range of more than 150 gym, and it can accommodate object thick-nesses of up to 1/4 inch. Since the pneumatic servo system employs an air nozzle, it not only provides automatic focusing, but also a film hold-down method that was determined to be the best of a number of possible candidates.

The scalar theory of microdensitometer performance developed in previous papers is revised and expanded to include coherent illumination. The central ideas of scalar microdensitometry are combined from several basic sources and summarized ; this paper serves as a convenient single source of microdensitometer theory. Eight distinct variations of instrument configuration and operation are identified, and the image characteristics and conditions for linear microdensitometry are developed for each. The concept of effective incoherence is summarized and discussed. Further consideration of the problems associated with aperture sizes, determination of the sampling aperture size and image vs sample scanning are presented. An experimental test program, carried out with the Mann-Data Microanalyzer, by experienced operators, under closely controlled conditions to test the theory, is reported. Among the conclusions is that the instrument manifests its best performance with overfilled optics and sample scanning, and that an increased reduction factor on the influx side would be useful with image scanning. Several fundamental problems surface during the course of the investigation. These are discussed and the need for further study in certain areas is emphasized.

Problems in designing advanced microdensitometer systems are discussed. Requirements for systems capable of generating data suitable for direct use as computer input and some equipment design considerations are reviewed.

The Isodensitracer is a two-dimensional, scanning microdensitometer which rapidly produces density "contours" by means of a cyclic dropped line code. For more than three years this instrument has proven its utility for making rapid photometric and geometric analyses of photographic imagery for such diverse applications as moon mapping, spectrum analysis, and medical X-ray interpretation.

The evaluation of image quality parameters directly from the image is often required when the imaging system point spread is not known. The derivation of the parameter values from microdensitometer scans across the images of inferred edges in the scene format is a popular technique. This paper presents the results of a study to determine the errors incurred and the simplification that results from making certain approximations in deriving the values of two particular parameters. The measures of image quality sought are "resolution", which is defined as the area under the transfer function, and "passband", which is the area under the square of the transfer function. An expression is derived for low contrast edges which provides a simple relation between the edge density trace and the line spread function which is independent of film gamma, and the error in resolution and passband resulting from its use as an approximation for edges of higher contrast is given. The second approximation considered is due to the practical requirement that the density trace of the edge is truncated. Closed-form edge-density functions derived from transfer functions consisting of weighted superpositions of exponential and gaussian terms were terminated by "visual" inspection and used to evaluate the errors introduced in the computations of the resolution and passband.

The increased resolution capability of modern photographic emulsions and optics has created a need for increasingly precise and automated microdensitometry. This article describes a laser interferometer controlled microdensitometer with greatly improved positional accuracy of 0.5 micrometer.

A review of the current developments in microdensitometry is made, with emphasis on the investigations leading to the current level of understanding of optical performance. The classical microdensitometer is then analyzed according to the principles of the theory of partial coherence. Conditions for the insuring of linear operation are derived, and the idea of effective incoherence at the source aperture is presented with a discussion of the implications. The various microdensitometer configurations are subjected to analysis, and the four possible variations (viz., overfilling, underfilling, with two possible locations for the sampling aperture) are thoroughly evaluated. The new concept of linear microdensitometry is discussed and summarized briefly. The current concerns of microdensitometry are then presented and considered. The restrictions on maximum sample frequency as a result of the partial coherence in the illumination is a major concern of this paper, and tables are presented for typical microdensitometer configurations that delineate the kinds of limitations that can be expected.

Tradeoff studies and tests on the various subsystems of a microdensitometer have resulted in a recommended design configuration for a Linear Microdensitometer. The photometric subsystem includes laser sources suitable for color operation, an illumination system capable of producing a 1 um spot on the film, a wide angle collection system (NA = 1.0), a photodiode detector, and digital density conversion. Performance of the photometric system was verified by breadboard tests. The results include an MTF (square wave response) of 99 percent at 200 Qp/mm and 90 percent at 500 kp/mm, and linearity (object independence) to greater than 400 Qp/mm. The photometric precision is ±0.01 D from 0 D to 2 D and ±0.02 D from 2 D to 4 D. These specifications are attained at scanning speeds up to 8000 samples per second. A two-beam optical system and automatic focus control ensure photometric repeatability. The two-axis stage is driven by phase-locked loop servos incorporating interferometric position sensors. The servos permit scanning at skew angles, obviating stage rotation. The stage sub-system is precise to within ±0.5 pm over 5 inches of travel. Stage speed is variable from 0.4 mm/sec to 40 mm/sec. Human factors features include rapid sample acquisition by virtue of a viewing system with magnification variable from 16X to 200X and simplified operator control through a CRT graphic terminal.