The National Physical Laboratory has sponsored fully toleranced optical designs of five standard reference lenses for the assessment of equipment used in measurement of the Optical Transfer Function. This paper will briefly describe those designs, indicating the particular imaging characteristics that each lens simulates and the tolerances associated with each design if the theoretical performance level is to be achieved. The influence of those tolerances on the mechanical design of the lens mounts will be indicated and the difficulties in manufacture arising from them discussed. Examples of three lens designs have been completed to date. The performance of two designs, measured both interferometrically and directly on an OTF measuring instrument, will be compared with the theoretically predicted performance. The performance of the third design, a triplet lens, has only been assessed interfero-metrically. A comparison of these initial measurements with theoretical performance will also be presented. Progress in the construction of the other two lens designs will be reported.
MTF-based simplified test conditions and proposed criteria for the evaluation of the image quality of fixed-focal length lenses and zoom lenses for 35 mm cameras are described. Measurements are made for white light, in the infinite conjugate plane, on-axis and at two 0.7 field (15 mm off-axis) positions, at full aperture and f/8. The focus is set at full aperture for best MTF response at 30 mm-1 on-axis. The lenses are evaluated by examining the MTF responses at 10 mm-1 and 30 mm-1. Previously proposed conditions and criteria for fixed-focal length lenses are also described for reference.
The investigation (reported earlier*) of the reproducibility of monochromatic MTF measurements on a standard wideÃ¢â‚¬â€?angle lens is now being extended to the polychromatic case. This paper reviewed the basic test parameters which must be considered in such a case if accurate results are to be obtained, and discussed the first results of the new intercomparison exercise. *Optica Acta v 22 1975 p 249-264
The design and construction of a standard reference lens for testing MTF equipment operating in the thermal band (2 to 14 μm) is described. The results of some preliminary measurements with the lens are reported and discussed.
The working group of Japan Optical Engineering Research Association (JOERA) was organized in 1974 and made interlaboratory'domparisons of MTF measurements and OTF Standardization. MTF measurements were taken at eight to ten places, using two standard lenses. Standard lens II: Petzval type, f/2, f=50mm, infinite conjugate angular field up to 5°, λ=546nm Standard lens III : Gauss type, f/2.8, f=50mm, infinite conjugate angular field up to ±25°,λ=546nm As a measure for international comparison, SIRA Institute and JOERA exchanged their standard lens and performed measurements. The results of MTF measurements showed good coincide on the axis, but the spread of max. ±0.15 off the axis. We have examined the cause of this error. The measurement using monochromatic light except e-line and white light does not show good coincide. Then this year we intend to investigate this subject in both calculation and measurement.
Some parameters of greater or lesser importance for matching spectral response are discussed for specific lens types. The influence of the deviation of light-source color temperature and deviation of the detector's spectral response are shown for spectral energy distributions frequently used in practice. Although a general answer cannot be given, certain trends and connections can be detected. The effect of different fluorescent phosphors on MTF measurements of stopped-down and high-aperture lenses is shown.
Of the various methods available for evaluation of imaging performance of optical systems those which are most useful for application 'in the field' are those based on measurement of the quality of photographic images of edges. However, with such methods it is usually necessary to estimate the 'best' edge profile in the presence of noise, after which the edge must be differentiated to yield the line spread function and further processed to yield either the point spread function or the MTF. It is thus desirable to have available a reliable curve fitting facility in order to automate the analysis of edge data, or to have a means of rapid assessment manually of the spread functions and MTF associated with a given visually fitted 'best' edge profile. Simple but effective procedures will be discussed which provide widely applicable methods of dealing with edge data both automatically and manually.
Techniques designed for assessing the performance of lenses operating in the 8-13 micron band are described. These include interferometry and pupil scanning, both using monochromatic radiation from a CO2 laser source. In addition a nominally polychromatic line spread function apparatus is shown. Results are presented of MTF using these techniques, along with those from a commercially available polychromatic grating scanning instrument. Measurements have been made on standard audit and operational lenses. Comparisons are made and differences attributable to radiation source emphasised.
In many methods to measure the Modulation Transfer Function( MTF ) of an optical system, a pinhole or a slit is used as a test target. In such methods, the light flux that is diverging from a test target must illuminate the full aperture of the test lens evenly. In practice, however, it is rare to satisfy this condition especially for a measurement of an optical system of high numerical aperture. And the insufficient illumination causes some amount of error on the measured MTF values. This paper describes a method for estimating the error of MTF values that are caused by the difference of illumination condition.
As a realistic visual task subjects had to recognize military vehicles out of six. alternatives. Percentages of correct identifications were determined at three different observation distances in order to interpolate the distance required for 50% correct, which is taken as a measure of performance. The scenery was simulated indoors by slide-projection in favour of better control of conditions and faster procedure. The slides were photographed at the location of an earlier field trial. The performance of a. portable, image, intensifier was measured, in this way as a function of luminance and contrast in object space. It ap-pears that recognition of the present set of objects is visually equivalent to the detec-tion. of a circular disk, and can be predicted from image quality data (displayed S/14, MTF) along that line.
For non-visual optical systems the modern methods of image evaluation (OTF and veiling glare), used in conjunction with an RMS noise statistic, permit the prediction of system performance in terms of stimulus size, stimulus strength and an acceptable signal/noise ratio for a given false alarm rate. However, when it comes to visualisation of other than noise-free images it appears from literature that performance is not uniquely related to these parameters. A plausible and tried modelling of the interactions of noise and quality in visual tasks is discussed in general terms and the implications of the findings on the interplay of noise and quality in various visual tasks are highlighted.
The collimation of binocular telescopes appears to be based on the assumption that the user prefers to accommodate to a stimulus at infinity. Recent work favours the view that for normal observers the preferred accommodative state of the eye is approximately 1 dioptre. Present collimation procedures do not allow the vergence mechanism of the observer's eyes to match their accommodative state. It is shown that the effect of this is to reduce the physiological binocular summation associated with a contrast detection task. A short field trial is reported, where the results obtained in the laboratory were confirmed under realistic conditions. The conclusion is that the present strict tolerances on the collimation of binoculars could be relaxed if the design of binoculars were modified to take into account the accommodative-vergence state of the eye.
The subjectively assessed quality of photographs taken with a 48 in lens has been compared with the MTF of the lens measured in the same image planes. The image quality and associated MTF were varied by the through focus technique, using a range of image planes on both sides of the optimum. The photographs have been examined to ascertain the change in MTF, at the various levels, resulting in a just detectable difference in image quality, and whether a "reduced" MTF could be used which would sensitively indicate the changes. The results will be used to provide MTF production tolerances for lenses used in aerial photography. The subject matter of the photography was vehicles as photographed from a vertical range of 40,000 ft, giving a scale of 1:10,000. Actual photography from an aircraft introduces uncontrollable factors such as vibration and atmospheric turbulence which degrade the image, so the situation was simulated using two stages of ground photography. The first produced a diapositive of vehicle models, using a high quality 35 mm lens and film which caused minimal reduction in the spatial frequency content of the original scene. This diapositive was photographed with the test lens, on to the film most often used in aerial reconnaissance operations, in conjunction with a collimator to provide the infinite conjugate condition and final correct image scale. EROS 200 equipment, employing crossed slits and a rotating grating was used to measure the MTF and the techniques employed to ensure that the image and MTF were recorded in the same plane are described. The spectral bandwidth of the illuminant was restricted to ensure that the same spectral conditions applied during the photography and MTF measurement. This enhanced the performance of the lens and broadened the scope of the experiment. Results indicate that for a combination of this lens with FP3 Aerial Film there are two useful indicators of image quality. (1) The area between the lens MTF (scaled to a maximum input modulation of 0.2) and the film threshold curve. (2) The cut-off spatial frequency. The two indicators are linked geometrically.
In an azimuth scan MTF test the test pattern is rotated so that a continuous scan through all orientations is obtained. Generally the system operates at a single spatial frequency but it need not necessarily be thus limited. Several methods are possible for generating the test target; that found most effective at MVEE is to use a radial grating rotating with a planetary motion in conjunction with a pinhole and detector arrangement as analyser. Initially such a system was used to test sights which had given variations between radial and tangential MTF responses in on-axis tests. For this use the method provided an effective test, and azimuth scan MTF equipment arrangements have subsequently been incorporated in production test equipment. This paper is mainly concerned with additional uses of azimuth scan equipment in more complex evaluation tasks on afocal instruments in which, while measuring the MTF at a single spatial frequency, data are obtained on dioptre setting, curvature of field, astigmatism and graticule focus for axial and extra axial field points. Nominally plane surfaced components can be tested by azimuth scan techniques and the reduced MTF caused by non-flat head mirrors and prisms is illustrated. Methods of presentation of derived data are described and it is shown that some particular defects can be more effectively indicated by the shape of azimuth scan records than by other tests.
The performance of image intensifiers is rated according to their capability of detecting, recognizing, and identifying objects at low light levels. The most important parameters limiting these capabilities are noise and imaging parameters. Noise parameters are normally expressed in signal to noise ratios, noise power spectra and noise statistics, whereas imaging parameters are described by the modulation transfer function and veiling glare. This paper will describe the instrumentation and methods used at the Night Vision Laboratory to determine these parameters. In developing the instrumentation and test methods, special consideration has been given to a clear separation of noise and imaging parameters.
This paper describes the method which is used at RAE Farnborough, to carry out laboratory evaluations of thermal imaging systems. The major performance parameters which are measured are Minimum Resolvable Temperature Difference (MUD), Minimum Detectable Temperature Difference (MDTD), Modulation Transfer Function (MTF) and Signal Transfer Function (SiTF). The reasons for choosing these parameters are discussed, and a description is given of the apparatus and the techniques which are used for the measurements.
Experience with modern television lenses has shown the necessity for quality control. To provide this quality control, test methods have been established to check Geometry, MTF, Colour Performance and Veiling Glare and Flare characteristics. The considerations that resulted in the adoption of these test methods are discussed.
A description is given of the essential points of DIN 58 185 "Optical Transfer Function", in particular of the principles of OTF measurement, the necessary equipment and problems likely to be encountered during measurement. Items of special importance are incoherent illumination, linearity especially of the electronic system and the existence of an isoplanatic region. These have to be checked before measurement. Other points of interest are the lowest spatial frequency for normalization, the selection of azimuth in the test pieces and the effect of auxiliary systems, in so far as these are used. When measuring phase, a distinction must be made between phase and distortion. Calibration should ensure that phase starts with a horizontal tangent. Information is given on the effect of veiling glare.
A draft proposal for a documentary standard on the Optical Transfer Function has been prepared by a task force reporting to American National Standards Committee PH3 (Photographic Apparatus), subcommittee PH3-2 (Lenses, Projection and Printing Equipment). This proposal is currently under review by PH3, and its present form and status are described briefly here. The aims of the proposed document are to provide, for imaging devices in general and particularly for lenses, a consistent set of definitions and nomenclature for the spread, transfer, and pupil functions, the relations among them, and the parameters on which they may depend; to describe basic principles of measurement, using several alternative measuring techniques as detailed examples, and advise on some common sources of error and their minimization: to enumerate conditions that need to be specified in connection with OTF testing; and to provide a basic format and nomenclature for specifying the extent of such tests and for reporting the results. It is expected that subsequent standards, using this document as foundation, can concisely specify particular test limits or measuring methods, as appropriate to particular devices or applications. Some other documentary standards relating to image quality, under current consideration by PH3-2, are also very briefly described.
Activities of the working group on OTF standard in the Japanese Optical Engineering Research Association are introduced. The working group already made three drafts concerning with OTF definition, graphical representations and accuracy of measuring instruments. Now they are planing to make the draft of OTF measuring processing and carring out fundamental researches on white light standard.
In 1946, P.M. DUFFIEUX proposed the application of the transfer function(T.F.)concept to the optics field. A few years later, H.H. HOPKINS, W.H. STEEL, SHADE, LINFOOT and MARECHAL completed the theoretical framework started by their predecessor. This subject has since inspired much work which has been based more on "experimentation and measurement" than on the theoretical aspect. Restricting ourselves in the context of this meeting to the french contribution, we find that during the 1960-1970 period, three laboratories have devoted a considerable part of their research work to this area. In Besancon, Professor VIENOT and his team became interested in this problem very early, developing a method of 910aining the T.F. from the analogue spectral analysis of the point spread response of an optical system . More recently, in Professor FRACON's Paris laboratory, a method for measuq;lg the T.F. of an optical system by holographic recording the wave surface was proposed by Miss MAY. On our part, since 1962, our team at the INSTITUT d'OPTIQUE has been studyg and perfecting a method for measuring the Modulation Transfer Function (M.T.F.) of optical system , which hasiwulted in the development of a measurement instrument, the ACOFAM, manufactured by the MATRA Company.
In 1971, under the auspices of the Dutch Ministry of Defence, a committee started with a standard for the measurement of the MTF of image forming equipment. In a short time a reasonable and workable standard was produced that covered optical and electro-optical systems. This standard was far from perfect but it did contain so many useful starting points for the working out of an improved standard that it was presented to a NATO commission which had the task of preparing an OTF standard for the NATO countries as a whole. The paper deals further with some interesting aspects of the OTF and with problems encountered in OTF standardisation when non perfect imaging devices that have veiling glare, non isoplanatism or a restricted linear range are considered.
Since 1971, five British Standards have been published on various aspects of the assessment of the performance of optical devices. They are: 1 BS 4779: 1971 Recommendations for measurement of the optical transfer function of optical devices. 2 BS 4793: 1972 Recommendations for specifying the optical performance of lenses for TV cameras. 3 BS 4995: 1973 Recommendation for measurement of the veiling glare index of lenses and optical systems. 4 BS 5109: 1974 Glossary of terms used in assessing the performance of optical devices and systems. 5 BS 5275: 1976 Environmental testing of optical devices and systems. The broad principles underlying these standards were discussed, and the content of each briefly outlined.