The specification of the color of an object is important, sometimes crucially, to its acceptability for an intended application. When quantitative measurements are used to determine the color of an object, an uncertainty is required to satisfy the requirements of traceability. Determining the uncertainty of a color measurement has been a topic of interest for many years, and has received additional impetus with the publication of the ISO Guide to the Expression of Uncertainty in Measurement, which describes methods for estimating uncertainties in a consistent manner. These methods are applied to the case of reflectance colorimetry, where many of the sources of uncertainty are caused by systematic effects and are therefore correlated. Analytical expressions are derived for the combined uncertainties in the quantitative expression of color due to both correlated and uncorrelated effects. The measurement equation is used to derive the uncertainties in reflectance factor from different sources, and to take correlations between signals at the same wavelength into account. The correlations between reflectance factors at different wavelengths are also included in the analysis. These concepts are illustrated by considering uncertainties from common sources for selected colored specimens.

The National Physical Laboratory and the National Research Council of Canada both realize independent scales for regular transmittance in the mid-infrared part of the spectrum. A comparison of these scales has been recently completed. The agreement was excellent, in all cases lying within the quadrature combined uncertainties of the two institutions, and demonstrates the equivalence of the NPL and NRC mid-infrared regular transmittance scales.

In the past decade, scales for infrared spectral regular (specular) transmittance and reflectance have been established at the National Institute of Standards and Technology (NIST). Recently, we have developed an error budget based on direct evaluation of the sources of error in the measurement process. These include the detection system non-linearity, spatial non-uniformity of the detector, misalignment of components, inter-reflection between components, asymmetry of sample and reference measurements, and source polarization effects. Here we describe the evaluation of the linearity of our measurement system that includes a Fourier transform spectrophotometer, integrating sphere and MCT detector. From the linearity results, relative responsivity curves are also established.

CORM Technical Subcommittee OP-1 under Optical Properties of Materials, has undertaken a round-robin inter-comparison of several industrial laboratories. The basis of the inter-comparison is similar to that reported by Verrill and by Rich in that the Ceramic Colour Standards were shipped from one laboratory to the next. There were six participants in this inter-comparison, all possessing at least one instrument capable of reading the spectral diffuse reflectance factor from 360nm to 780nm at 5nm intervals with a 5nm bandpass, as recommended in Publication CIE 15.2. The laboratory temperature was recorded and kept to within 2° C of the nominal target. Some laboratories had second instruments with slightly different realizations of the CIE recommended geometry for diffuse reflectance factor. The results compare the readings from similar instruments (same manufacturer and geometry) and dissimilar instruments (same spectrometer but different geometry or different spectrometer and different geometry). This represents the first time that an inter-comparison of both similar and dissimilar spectrophotometers and reflectance accessories can be reviewed simultaneously.

Primary spectrophotometric Standard Reference Materials (SRMs) from the National Institute of Standards and Technology (NIST) are transfer artifact standards of the highest order in the United States. They are indispensable to spectrophotometer manufacturers for establishing instrument specifications and also to commercial spectrometrology companies for certified value assignment and benchmarking of secondary reference materials. When used judiciously, primary SRMs represent a pivotal link in the measurement traceability chain, and they facilitate quantitative assessment of measurement uncertainty for calibrations of metrological value. This paper provides a concise summary of the contemporary array of primary SRMs for molecular absorption spectrophotometry that can be used to leverage measurement accuracy and traceability to data and results for spectrophotometer operation within the spectral range 235 nm to 805 nm. An introductory benchmark traceability paradigm and a preliminary needs assessment for spectrophotometric SRMs are also presented.

The use of fluorescent-retroreflective materials in transportation safety applications (traffic signing and personal safety garments) has increased significantly over the last decade. The improved photostability of modern fluorescent colorant systems finally allows transportation agencies to make full use of the safety benefits of high visibility fluorescent signing and garments. As the use of fluorescent colored materials has grown, so has the need for accurate colorimetry to reliably and reproducibly describe the properties of these materials. Specifications for materials used in visual signaling are written in terms of absolute chromaticity and luminance factor limits. Therefore, accurate color measurement based on technically sound procedures is necessary to describe these materials. The bispectral or two-monochromator method is the referee procedure for determining the general (illuminant-independent) colorimetric properties of fluorescent materials. Until recently the ability to do bispectral fluorescent colorimetry was limited to a few high level photometric laboratories, almost solely National Standards Laboratories. Commercial bispectral fluorescent colorimeters are now available. The purpose of this paper is to examine the state-of-the-art in the colorimetry of fluorescent-retroreflective high visibility materials. Inter-comparisons were made between bispectral instruments and commercials 1-monochromator instruments. Measurements made on a series of fluorescent-retroreflective and non-retroreflective (diffusely reflecting) fluorescent materials were made on both types of instrument. Ordinary colored retroreflective and diffuse reflective materials were also measured. Representative results of the measurement inter-comparisons are presented and the effects of instrument geometry and illumination are discussed. The results are presented in terms of uncertainty in the determination of luminance factors and chromaticity coordinates. Accuracy is assessed relative to colored fluorescent reference materials calibrated by National Standards Laboratories.

No abstract available.

Fluorescent Colorants are widely used around the world to enhance visibility. The outstanding brightness and cleanliness of the colors lend themselves to applications in safety materials, advertising, toys, magazines, packaging, and other areas. The brightness and cleanliness is a result of the colorants ability to reradiate absorbed energy as visible light, usually shorter more energetic photons as longer less energetic photons. This can give reflectance values of well over 100%, sometimes as high as 300%, in the perceived color. A good working definition of fluorescent color is: A colorant that absorbs light energy and reradiates the energy at visible wavelengths. Light that is not absorbed is reflected, as in conventional color. Emission ceases when the excitation energy is removed.

Variability of spectral radiance factors of a set of fluorescent specimens was measured with eight different spectrophotometers, three designed for the two-monochromator method (2MM) using monochromatic irradiation, and five for polychromatic irradiation (1MM) using a Xenon source with a special filter technique. In four cases a single UV-filter was used that had to be adjusted to simulate the effect of D65-irradiation in the UV-range only. The inter-instrument agreement of 2MM-spectrophotometers compared to a reference instrument was within the range of the repeatability of that instrument (coverage factor k = 2). Differences of spectral radiance factors were larger for reflected radiance than for luminescent radiance, however, for colorimetric differences in most cases the opposite was true with the exception of heavily structured specimens. A certificate of a reference material containing fluorescent whitening agents was used for UV-filter adjustment in 1MM-instruments. For specimens of the same type colorimetric scatter from the reference instrument extended up to 2,4 CIELAB units, whereas in 2MM-instruments it was up to 1,5 CIELABS units only. For chromatic luminescing materials the UV-adjustment was not effective and the colour difference for 1MM readings extended up to 5,4 CIELAB units, whereas for 2MM-instruments it was up to 3,4 CIELAB units.

Lack of photoluminescence is an essential requirement of a reference standard that is used to calibrate color-measuring instruments. This effect is being studied for several white reflecting materials commonly used as diffuse reflectance standards. This paper reports on the findings for two types of Russian opal (MS20 and MS14) and Japanese opal (Everwhite) glasses. Russian opal glass is under consideration by ASTM Committee E12.12 on Appearance of Metallics and Pearlescents as a transfer standard for the calibration of goniospectrocolorimeters and Japanese opal has been used for many years as a transfer standard in paper colorimetry. The accurate characterization of photoluminescence in these opal glasses is important in determining their suitability for these applications. These measurements were performed on the NRC Reference Spectrofluorimeter over the wavelength range 250-800 nm. It was found that the photoluminescent effect was weakest in the Russian opal glass MS14 and strongest in the Japanese opal. Both types of Russian opal glass exhibit a single broad emission band from 320-450 nm (max=370nm) which is excited by UV radiation from <250-320 nm (max=280 nm), excited by UV radiation in the ranges <250-280 nm (max<250 nm) and 250-340 nm (max=310 nm), respectively. The former emission band is responsible for a bright-blue fluorescent color when the sample is irradiated with short-wave UV (254 nm). The impact of this photoluminescent effect was quantified by calculation of the resultant colorimetric errors in the CIE chromaticity coordinates, luminance value, and CIE whiteness for two common illumination conditions: filtered xenon which approximates CIE Illumination D65 and unfiltered xenon which approximates an equi-energy source. For the former case, the colorimetric errors were negligible for both types of Russian opal glass and only significant for the Japanese opal for the whiteness value (0.12% error). For the equi-energy source, the errors in chromaticities ranged from 0.0003 to 0.0007, in luminance value from 0.07 to 0.16, and in CIE Whiteness value from 0.80 to 2.0. Recommendations are given regarding the use of these opal glasses as transfer standards for reflection colorimetry.

To estimate the gonioapparent properties of metallic and pearlescent paint finishes, samples of metal-flake and pearl-mica pigmented paint surfaces were prepared. For these samples, spectral reflectance factors for various geometric conditions were measured by gonio-spectrophotometer. From these spectral reflectance factors, colorimetric values were calculated. Furthermore, spectral reflectance factors were treated by principal component analysis, and flake orientation in the paint layer was supposed.

In the field of spectrophotometry, the value of the quantities depends upon the geometry under which they are measured. Therefore, it is imperative to completely describe the measurement geometry. Many documentary standards specify the geometry for a particular application. However, to accurately specify the geometry, a general, basic understanding of the relevant parameters for describing the geometry is required. A systematic approach for describing the measurement geometry is presented, which will hopefully have a positive impact on documentary standards. The key to describing the geometry is to consider the illuminator and receiver of the instrument as optical systems with pupils and windows. It is these optical systems, together with the reference plane, that determine the sampling aperture of the instrument. The geometry is then completely described by the relations between the sampling aperture and the optical systems of the illuminator and receiver. These concepts are illustrated by considering three configurations of pupils and windows relative to the focal point of an optical system.

The 1986 CIE document 15.2, Colorimetry, was necessarily broad in specifying the use of the gloss trap or specular port in integrating sphere measurements. This has led to a variety of spectrophotometer configurations that adhere to the CIE recommendation. To help uses of these devices determine the performance of their instruments with respect to specular excluded measurements, a procedure has been demonstrated to quantify the effective specular port width of an integrating sphere device. The proposed procedure has been tested on four spectrophotometers, three of which use conventional specular ports of varying sizes. The specular ports of these three devices can be physically measured, however the fourth device uses and alternative method for the specular excluded measurement, and the diameter of its specular port cannot simply be measured. The procedure allows for a relative comparison of conventional devices and those using an alternative method.

The National Physical Laboratory is the national measurement institute of the United Kingdom, and currently disseminates diffuse reflectance scales from 320 nm to 56m1. Visible diffuse reflectance and colour measurement was identified as an area where take up of traceability was variable, with sporadic application of best practice in measurement. A case study was therefore conducted into the feasibility of performing monitoring and calibration of colour-measuring spectrophotometers using the internet to communicate between remote sites to improve take up of traceability and measurement good practice. No data correction functionality exists at present. The final objective will be to launch an internet-based, fully accredited measurement service in the area of diffuse reflectance and colour measurement. This paper describes the rationale and functionality of the system resulting from this case study, and makes recommendations on future developments to enhance this type of quality controlltraceability mechanism.

Harmonic generation of the output of a mode-locked titanium sapphire laser is a highly effective means for producing high power, quasi-cw, wavelength tunable radiation in the 200-500 nm spectral range. The laser is not truly cw, consisting of pulses at approximately 100 MHz. This paper investigates the possibility of stretching these pulses in time using a multimode optical fiber to achieve a truly cw signal: stretching of a 1.5 ps pulse to around 1 ns has been demonstrated. Non-linearity effects in the responsivity of different optical radiation detectors due to the pulsed nature of the radiation have been investigated, and no evidence for non-linearity has been seen for either silicon or GaAsP detectors at wavelengths of 300 or 360 nm within the measurement uncertainties (< 0.15%). Finally, data are presented indicating that the use of mode-locked radiation in transmittance spectrophotometry avoids the problem of interference between front and back surface reflections which has previously hindered high accuracy transmittance measurement using lasers.

Intrinsic standards can become an efficient mechanism for providing traceable reference materials for chemical spectrophotometry. The sealed holmium oxide (Ho₂O₃) in dilute acidic aqueous solution used by the U.S. National Institute of Standards and Technology as Standard Reference Material 2034 is a candidate material for an intrinsic wavelength standard in the ultraviolet and visible spectral regions. Low mass fractions of Ho₂O₃ in glass and as Ho³⁺(aq) solutions have long been favored for use as wavelength calibration materials in molecular absorption spectrophotometry on the basis of their spectral coverage and absorption band shape. Three national metrology institutes of the North American Cooperation in Metrology evaluated the performance of Ho³⁺(aq) certified reference materials under “routine” operating conditions using commercial instrumentation. The resulting data show a substantial level of agreement while also demonstrating that the wavelength comparability of the five instruments used by the participants can actually be improved by calibrating all of the instruments to the consensus band locations of Ho³⁺(aq) solution. A similar approach is being considered to establish intrinsic absorbance standards.

A prototype kit containing artefacts, best practice guide and software for enhancing reflectance measurement data are described. The kit uses measurements made on calibrated artefacts to make other reflectance measurement data more reliable, in terms of zero offset and linearity, while also providing traceability to national reflectance scales. The use of the kit can give agreement between measurements made on the same artefact by different instruments of less than the colour discrimination limit of the human eye of 0.5 ΔE*_ab colour difference units.

The National Physical Laboratory currently disseminates United Kingdom diffuse reflectance scales from 320 nm to 56 μm. The scale for hemispherical reflectance from 360 nm to 830 nm was previously traceable to the National Research Council of Canada and the Physikalisch-Technische Bundesanstalt, and its extension to 56 μm has relied partly on data published by Morren et al. (1972). Measurements of 0°/45° radiance factor from 360 nm to 830 nm were traceable to the Physikalisch-Technische Bundesanstalt. The National Physical Laboratory now realizes all United Kingdom diffuse reflectance scales independently as described here.

A facility has been developed at the National Institute of Standards and Technology (NIST) to provide calibration services for color-measuring instruments to address the need for improving and certifying the measurement uncertainties of this type of instrument. While NIST has active programs in photometry, flat panel display metrology, and color and appearance measurements, these are the first services offered by NIST tailored to color-measuring instruments for displays. An overview of the facility, the calibration approach, and associated uncertainties are presented. Details of a new tunable colorimetric source and the development of new transfer standard instruments are discussed.

In earlier papers, NIST proposed a standard illumination source and optical filter targets with which to assess the state-of-the-art of display measurement. The Display Measurement Assessment Transfer Standard (DMATS) was designed to present the display metrologist with a rectangular array of targets such as color filters, polarizers, and grilles, back-lighted by uniform illumination, to be measured using methods and instruments typically used in display performance measurement. A “round robin” interlaboratory measurement exercise using the “standard” artifact suite would enable a first order assessment of display measurement reproducibility, i.e., measurement variability within the electronic display community. The rectangular array design of the DMATS was anticipated to present stray light and color contamination challenges to facilitate identification of error sources deriving from measurement protocols, laboratory environment, and equipment. However, complications in dealing with heating problems threatened to delay the planned laboratory intercomparison. The Gamut Assessment Standard (GAS) was thus designed as an interim solution to enable the NIST scientists and participating measurement laboratories to begin collecting data. The GAS consists of a 150 mm diameter integrating sphere standard illumination source with a stray light elimination tube (SLET) mounted at the exit port. A dual six-position filter wheel is mounted at the SLET exit port. One wheel holds a series of neutral density filters and a second interchangeable wheel holds various color filters. This paper describes the design and construction of the GAS, its initial performance characterization by NIST, and comparison measurements made at NPL. Possible design changes suggested by the results of the preliminary intercomparison are discussed, as are plans for future interlaboratory comparisons and potential use of the GAS as a transfer standard for laboratory self-certification.

The NIST Flat Panel Display Laboratory (FPDL) is operated through the Display Metrology Project (DMP) of the Electronic Information Technology Group in the Electricity Division of the Electronics and Electrical Engineering Laboratory of NIST. The DMP works to develop and refine measurement procedures in support of ongoing electronic display metrology, and applies the results in the development of national and international standards for flat panel display characterization.

There is a great deal of interest in the sunlight readability of displays. How to quantify this, particularly on a numerical scale rather than pass/fail is an important consideration. Some military standards exist e.g. MIL-L-85762A but are these appropriate to non-military products? We report the use of a method developed by BAE SYSTEMS, the model is based on our ability to see things by virtue of a difference in either luminance or chrominance, or both. The model uses a method called PJND (Perceptible Just Noticeable Difference), based on LJND (Luminance Just Noticeable Difference) and CJND (Chrominance Just Noticeable Difference) thresholds. The basis of this model is a series of acceptance criteria established by subjects in a realistic environment; BAE SYSTEMS have an Ambient Lighting Facility (ALF) which simulates many lighting scenarios that are likely to be encountered in real life. Product mock-ups were taken into this environment and subjects were asked to choose levels of acceptance for differing screen presentations and various lighting geometries. The method results in a single figure of readability, which in effect summarizes the task and environment. This figure can then form the basis of a specification between display supplier and vendor. We plan to develop the model such that it should be possible to predict the best combination of treatments to apply to the display surface to give the optimum and most cost effective sunlight readable display for a given application and product.