To View all figures/charts/illustrations referenced in this article, click here.  Abstract  Looking for a digital camcorder for use in performing surveillance? Well, technology improvements in CCD and CMOS image sensors as well as the introduction of new high definition display and recording formats have certainly resulted in the introduction of lots of different digital camcorder models. Because of the wide variation in features and price points, it’s more difficult than ever to select the optimal digital camcorder. Certainly, some models will be better than others in consideration of the specific application for which the camcorder is intended and the features that are required. Because there are so many factors involved in making a selection and since new models are frequently introduced, it’s best to provide guidelines useful when evaluating models intended for a specific application. Then, as new models become available, they can be compared to existing models based on the same guidelines. The goal of this paper is to provide guidelines that should be considered when evaluating camcorders for use in surveillance applications. These applications are those that would be used by military, law enforcement and special investigative personnel for gathering evidence or performing electronic news gathering and usually include viewing at a distance. While these are useful guidelines for understanding important camcorder specs, prior to making your own camcorder purchase decision, we suggest that you perform your own research of available models and, based on these guidelines, select those models that are expected to deliver the best performance and perform your own tests under real-life situations. You may learn that other factors may be important to your specific application and that your final decision is a compromise of those other factors. For surveillance, optical image quality is of highest importance. Camcorder resolution and sensitivity are the primary factors to consider, with a variety of related factors such as image and pixel fields-of-view, the pixel size, as well as the frame rate, image stabilization, alternative lens options and lens interchangeability, integration mode, digital zoom and display resolution, user interface, video formats, recording media, manual gain, infrared mode, video compression, external audio inputs, viewfinder, robustness, and video compression.  Camcorder Resolution Video resolution is affected by a variety of different camcorder specifications. It is well known that high definition camcorders deliver higher resolution than standard definition formats. However, it is important to realize that if the camcorder is used to record image sequences that are later to be viewed (rather than using the camcorder just to view the video output in real time), both the optical resolution and the recording format must be considered in determining the ultimate video resolution. Considered separately, the optical resolution depends not only on the detector array size but also the image and pixel field-of-view. HD recording format also affects resolution. Finally, there are other resolution factors that may warrant consideration in order to determine the ultimate camcorder resolution.  Detector Array Size: It seems that every camcorder uses a slightly different imaging detector. While the HD recording format can be 1280×720, 1440×1080 or 1920×1080, the imaging detectors may vary quite a bit. As shown in the table, for a selection of mid-range camcorders, detector array sizes can vary from single chip cameras having arrays as large as 1920×1080 (Canon HF-10 and Sony HDRSR11) to three-chip blocks having detectors as large as 1440×1080. The physical size of the imaging detector also varies from camcorder to camcorder.  Image and Pixel Field-of-View: As shown in the accompanying chart (reference at above link) the combination of the physical size of the camcorder’s imaging detector (for example 1/3”) and the focal length of its objective lens (for example 50mm) result in an image field-of-view (FOV) that is defined as the angle of view from which image information is received. There are separate angles corresponding to the horizontal and vertical directions depending on the aspect ratio of the image. For wide screen video formats, the vertical FOV is 9/16 of the horizontal FOV. All camcorders have zoom lenses where the focal length varies from a wide angle position (with the largest angular FOV) to a telephoto position (with the smallest angular FOV). A smaller field-of-view results in a higher image magnification as would be required to view small objects at a distance. Smaller imaging detectors create narrower angles of view when used on the same focal length lens. For example, a 50 mm lens delivers a horizontal FOV of 5.5° and 3.7° when used with a 1/2” and 1/3” format detector, respectively. Based on the variation of detector sizes and lens focal lengths, the angular field of view for different camcorders varies accordingly. The image FOV can be determined for any camcorder based on a simple formula. The pixel FOV (also known as the instantaneous field of view, or IFOV) is a measure of the resolution of an imaging system. Though related, image FOV and pixel FOV are distinct and a camcorder having the narrow image FOV does not necessarily have a corresponding narrow pixel FOV. As shown in the table (reference at above link), while the Canon camcorders have the smallest image and pixel FOV values, the Sony HDR-SR11 has a small pixel FOV despite an image FOV that is nearly twice as wide as those for Canon. A handy FOV calculator can be used to determine the FOV values for certain candidate camcorders. Pixel FOV is also calculated as are detection, identification and recognition range values for certain specified object sizes at a specified distance. Recording Format: The difference between the different HD recording formats is beyond the scope of this white paper. However, the impact of recording format on resolution is critical to imaging performance for surveillance applications and is described here. As shown in the figure (referenced at the first link in article), standard definition camcorders (with a 4:3 aspect ratio) recorded video with 480 horizontal rows per frame giving a resolution of 640 x 480. (This is for NTSC formats. PAL formats recorded with 576 rows per frame giving a resolution of 640×576). Wide screen versions of these standard definition camcorders with a 16:9 aspect ratio were also available and recorded with slightly larger resolution of 720 x 480 (720×576 for PAL versions). Because of the limitations of the recording format, it would not have been useful to have a high resolution CCD/CMOS detector larger than 720×480 pixels in a standard definition NTSC camcorder except under certain circumstances (like zooming in to view central pixels). However, with the introduction of higher resolution recording formats, higher resolution detectors make sense. Camcorder Sensitivity High definition refers to video having higher resolution than standard definition and most commonly refers to resolutions based on either 720 rows/frame or 1080 rows/frame. Recording formats can either be 1920×1080, 1440×1080 and 1280×720. Consequently, are available with either 720 or 1080 rows per frame recording formats having a resolution of 1280×720 and 1920×1080 respectively. It is important to understand the recording format for any camcorder under consideration. If the full sensor resolution is necessary to perform the surveillance function, then make sure that the camcorder records in sufficient resolution. Had this whitepaper been about digital still cameras, the above discussion of recording format would’ve been sufficient to describe the effects on image quality. However, since video incorporates motion, the fact that some camcorders record “interlaced” video while others record “progressive” video has some relevance, particularly when motion is important to the type of surveillance being performed. If not, then the relevance of interlace vs. progressive scan is less important. Other Resolution Factors: It’s worth mentioning that in those camcorders that utilize three sensors, pixel shift technology has been employed in the horizontal axis to make incremental improvements to the sensor resolution by using interpolation. This complicates the calculation of optical resolution somewhat, but may not be a primary factor. Camcorder low-light sensitivity is usually of primary importance for surveillance applications, particularly when imaging at dusk or dawn, but also when attaching a night vision module that intensifies ambient light. The guidelines for camcorder evaluation are listed below. We recommend a review of the following factors: the Recommended Minimum Illumination Specification, Objective Lens F-number, Detector Pixel Size, Night Vision Module availability. As such, but criteria enabling the comparison of these including: objective lens F-number, pixel size, electronic noise reduction, It is apparent that low-light performance as detailed on the list of the camcorders specifications because different cameras can produce varying results, even though the specifications may be similar. Consequently, once you’ve identified a few different camcorders that have the features of interest and the performance that you require, it is important to test the camcorder’s low-light capabilities. Recommended Minimum Illumination Specification: Quite often, low-light performance specifications are provided on a camcorder’s specification sheet. These are usually described as either “low light sensitivity” or “minimum illumination” and the result is given in lux. The lower the lux rating, the better the camcorder performs in low light. However, because the details describing how these low-light measurements are made is [...]