NIST cooperates with the three services (Air Force, Navy, and Army) to realize a uniform night vision goggle (NVG) calibration system with low measurement uncertainty. These efforts include transferring the NIST detector-based radiance and luminance responsivity scales to the primary standards laboratories, developing radiometric models to estimate uncertainty components in NVG calibrations and field measurements, and standardizing the spectral distribution of the test set of the light sources used in calibrating NVG.
NIST cooperates with the three services (Air Force, Navy, and Army) to realize a uniform night vision goggle (NVG) calibration system with low measurement uncertainty. These efforts include transferring the NIST detector-based radiance and luminance responsivity scales to the primary standards laboratories, developing radiometric models to estimate uncertainty components in NVG calibrations and field measurements, and standardizing the spectral distribution of the test set of the light sources used in calibrating NVG.
A gain measurement technique for the calibration of night vision goggles (NVG) is proposed and evaluated. This technique is based on the radiance measurements at the input and output of the NVG. In contrast to the old definition which uses a non-International System of Units (SI) traceable luminance, the “equivalent luminance unit”, the suggested technique utilizes the radiance quantities that are traceable to the SI units through NIST standards. Due to the implementation of the scaling coefficients originating from the NVG spectral responsivities, the same NVG gain is expected within both techniques. The suggested method was evaluated at the NIST night vision calibration facility and the experimental data were compared to the results obtained with a commercial NVG test-set. The comparison of the radiometric quantities obtained using the two different methods indicated differences up to 15 % due to different calibration conditions. However, at proper calibration, equal NVG gains within both the suggested and old gain definitions were measured for the goggles equipped with a filmless image tube. The NVG gain uncertainty analysis including the effect of the no-moon night sky radiation was performed for goggle types A, B, and C.
The luminance gain of an image intensifier-based device may be considered as a ratio of the output luminance to the input luminance. This simplified approach, however, cannot serve as the universal NVG gain definition due to significant variations in the input and output photometric quantities. When NVGs become more sensitive to infrared (IR) radiation than to visible, the use of photopic units may lead to NVG testing artifact because most of the night sky radiation detected by NVG is in the near-IR range above 800 nm. But for the luminance defined by the International Commission on Illumination, the V(λ) photopic function characterizes only the visible range while it is indefinite in the IR interval.
The NVG manufacturing and testing procedures, however, need certain criteria to characterize the NVG gain. In the common procedure used, the fixed-level IR input radiance of 2.48x10-6 W sr-1 m-2 produced by an LED peaking at ≈ 820 nm is treated as “equivalent luminance”. Accordingly, the NVG gain definition (old) which is still in use for NVG calibrations is:
Gain = output luminance / input luminance,
(1)
where the input IR-LED radiance of 2.48x10-6 W sr-1 m-2 is assumed to be equal to the “equivalent luminance” unit. By characterization of the NVG output and input with a mix of different photometric and radiometric units, the definition (Eq. 1) is the currently established value for NVG gain. The paradox of this approach is that a photometric unit cannot be used in the IR range where the V(λ) function is not defined.
The suggested new NVG gain definition is based on the input and output radiance quantities:
Gain (new) = K1 . K2 . R2 / R1,
(2)
where R1 is an input radiance, R2 is an output radiance. Coefficients K1 and K2 characterize the radiance spectral features at the NVG input and output, respectively. Both coefficients are constants for the same type of NVG with standardized spectral characteristics. New gain definition (Eq. 2) satisfies the following conditions:
Several types of NVGs with different spectral characteristics at the present are standardized: Type-A, Type-B and Type-C. The spectral responses for these NVGs are presented in Fig. 1 with blue diamonds, red squares, pink diamonds and green diamonds for NVG Type-A, -B, -C** and -C*, respectively. The difference in spectral response between the Class C* and Class C** curves is mostly due to a response flatness. Due to the same type of photocathode used, all NVGs have similar spectral characteristics in the upper wavelength range.
The data presented in Table provides the comparison of the NVG gain measured using the old luminance ratio definition and the output/input radiance ratio required for the new gain definition.
Source | Peak WL, nm | Out/In luminance ratio (for gain_old) |
Out/In radiance ratio (for gain_new) |
Out/In data ratio |
---|---|---|---|---|
1 | 462 | 2177 | 430 | 0.2 |
2 | 527 | 2384 | 471 | 0.2 |
3 | 565 | 2430 | 480 | 0.2 |
4 | 626 | 2063 | 408 | 0.2 |
At this point, the coefficient K1 has not been applied yet. As seen, the output/input radiance ratio used for gain measurements in the suggested new definition is about five times lower than the gain obtained from the output/input luminance ratio using the old definition. The reason of the discrepancy is due to artificially defined “equivalent luminance” unit used for the NVG input characterization. To create the reference to the old NVG gain definition and to match with the data of previously calibrated NVGs, it is reasonable to use the ratio of 4.95 as the coefficient K1 in the new NVG gain definition. Taking in account that K2 coefficient is unity for the photopic output measurements, Eq. 2 (new NVG gain definition) becomes
Gain = K . output radiance / input radiance,
(3)
where the coefficient K = K1 equalizes the numerical gain values in the new and old NVG gain definitions. For the NVG of arbitrary type, the coefficient K must be adjusted according to the change in the NVG spectral response. Within the new NVG gain definition, we suggest the coefficient K equal to 5.42 and 4.71 for NVGs type A and C, respectively. The suggested values for coefficient K are based on the known spectral responses of the type A and C NVGs, prorated to that of type B. The Ki coefficients are obtained from the integrated NVG input quantities as follows
Ki = KB ∫λSi(λ)N(λ)dλ/∫λSB(λ)N(λ)dλ,
(4)
where Si(λ) is a normalized (standard) NVG responsivity, index i denotes the i-type NVG, SB(λ) is the normalized (standard) responsivity of NVG type B, and N(λ) is the spectral radiance of the 2856 K blackbody. The new definition comprises that the source radiation (820-nm LED) is located within the spectral responsivity of the NVG input as in the case of the NVG Types A, B and C. If NVG is more sensitive to visible light, the LED should be selected properly, and the coefficient K should be re-defined. The NVG-type scaling coefficient implemented in the new definition makes the gain values in both the new and old definitions the same.