3rd Party Test Results for The NAC Memrecam fx K4 Digital High-Speed Camera Systems Color and Monochrome
Executive Summary
Background
In an effort to assist customers and prospective customers to better understand the capabilities of its high- speed camera systems, NAC Image Technology commissioned Electro Optical Industries (EOI) of Santa Barbara, CA to perform an independent evaluation of its Memrecam fx K4 Digital High-Speed Camera Systems, both the color version and the monochrome version. Although similar testing is performed by NAC’s engineering staff during the camera design and production processes, the 3rd party testing was deemed appropriate in order to provide the customer with an objective report on the abilities of the camera systems.
There are a variety of features and functions available with the Memrecam fx K4. However, the basic product of the cameras is high quality images. It is important, therefore, to understand specific information related to the digital images generated by the camera using an easily repeatable testing procedure so that a user or prospective user can determine the level of quality. The “Digital Camera Imaging Evaluation” report published by EOI defines a testing framework and describes features of the digital images based upon that testing framework. The results of the testing are meaningful and can be compared to the results published for other digital high-speed camera systems using the same testing methodology, resulting in a true “apples to apples” comparison.
Image Information
There are a number of important measures of the quality of a digital image. Those key aspects of the digital image described in the EOI evaluation report are: spatial resolution, signal to noise ratio, sensitivity, uniformity, modulation transfer function and intrascene dynamic range. A very brief description of why these items are important is as follows:
- Spatial Resolution – Spatial resolution refers to the camera’s ability to sharply and clearly define the extent or shape of features within an image. It describes how close two features can be within an image and still be resolved as unique. Spatial resolution is stated in cycles/mm or line pairs/mm, two terms that can be used interchangeably.
- Signal to Noise Ratio (SNR) – SNR is an engineering term for the power ratio between a signal (meaningful information) and background noise. The higher the ratio, the better the image.
- Sensitivity – The measure of sensitivity assists users in obtaining proper exposures and in determining the low light capability of electronic cameras. The International Organization for Standardization (ISO) has published a standard for determining sensitivity which assigns electronic cameras with ISO speed ratings, with these ISO speed ratings being intended to harmonize with film ISO speed ratings. Sensitivity is defined by the ISO committee in terms of ISO film speed equivalent at both excellent (SNR=40) and acceptable (SNR=10) image qualities.
- Uniformity – By looking at and comparing the value of pixels in a sample “bright” area of an image, the average value of all the pixels is calculated showing the variance of each pixel from the mean. This gives a measure of the evenness of the pixels and, by extension, the consistency of the image. The smaller the standard deviation from the mean, the better the measure of uniformity.
- Modulation Transfer Function (MTF) – The MTF describes how much a piece of optical equipment blurs the image of an object. The standard test is to see how well alternating white and black stripes show up (that is, their contrast, the difference between the whitest white and the blackest black) in the image, depending on how finely spaced they are. The MTF is a measure of how much bright-to-dark contrast is lost, as a function of the width of the stripes, as the light goes through the system. So it might be 0.95 for broad stripes, which are not significantly blurred, and 0.10 for ones almost too narrow to be viewed by the system. When reviewing the plots of the MTF, it should be noted that the human eye can resolve .05 to .10 (i.e. 5% to 10% Modulation).
- Intrascene Dynamic Range – The analysis of the images were done using a standard 8-bit, *.bmp image. For this type of image, there are 256 gray scale steps (total pixel depth) for a monochrome image and 766 gray scale steps (total pixel depth) for a color image. The measure of intrascene dynamic range refers to the intrascene performance, i.e. the ability to quantitatively detect the very dim and very bright parts of a single image. In other words, the intrascene dynamic range describes the equivalent number of gray scale steps that can be resolved in the image and this result can be compared to the theoretical number of gray scale steps for the monochrome and color images, respectively.
Test Results
Both the Memrecam fx K4 were tested using the EOI 66700 Image Evaluation Sphere. In addition, both cameras were tested with a lens aperture setting of f/5.6, with the exposure set at 1000 μs, and the frame rate set at 1000 fps. The report for the Memrecam fx K4 monochrome camera was published January 16, 2006 and the report for the Memrecam fx K4 color camera was published February 9, 2006.
| Test | Color K4 | Mono K4 | Units |
| Resolution | 16.00 | 16.00 | Spatial Horizontal (cycles/mm or lp/mm) Resolution |
| 16.00 | 16.00 | Spatial Horizontal (cycles/mm or lp/mm) Resolution | |
| Sensitivity | ISO 5,000 | ISO 32,000 | SNR=40 Excellent Image (Noise based ISO rating) |
| ISO 12,500 | ISO 80,000 | SNR=10 Acceptable Image (Noise based ISO rating) | |
| Uniformity | 0.0058 | 0.0109 | Standard Deviation |
| MTF | 10.94 | 11.76 | lp/mm at 50% cutoff frequency |
| 16.00 | 16.25 | lp/mm at 8% cutoff frequency | |
| Dynamic Range | 18.98 | 12.85 | Intrascene Ratio |
| 726 | 236 | Intrascene usable gray scale steps |
Summary
The Memrecam fx K4 camera systems, both monochrome and color, produce demonstrably excellent images, and the test data support this claim. The camera is extremely light sensitive and provides excellent resolution (both in terms of spatial resolution and bit-depth) while still providing “excellent” image quality