Very often, in CCTV, monitors have not been considered seriously. The display quality and monitor size are, sadly, chosen based on their price and physical size to fit a pre-determined area. It is quite clear though, that if the monitors are not up to the same level, or better, quality than the images displayed on them, they will be the weakest link in your CCTV system and basically conceal valuable visual information.

Understanding the visual display quality is even more important these days when we have computer monitors of various kinds entering the CCTV system design, competing with the traditionally analogue, composite (CVBS) monitors.

One very important quality factor is the technology of reproduction, which defines the contrast ratio, dynamic range, gamma, resolution, etc. But another important factor, not understood well enough, is the psychophysiology of human eye and its capability of viewing details.

Seeing details on a computer screen and on a composite monitor is not the same. They “work” with different resolution, different frequency, one is progressive scanning, the other (CVBS) interlaced and they are supposed to be viewed from various distances.

Through experiments and testing it has been found that the most a healthy human eye can resolve is not more than 5 to 6 lp/mm (line pairs per millimetre). This refers to an optimum distance between eye and object of around 0.3 m, as when you are reading a fine text. This equates to a minimum angle of about one-sixtieth of a degree (1/60°). So, 1/60º is considered the limit of angular discrimination for normal human vision. We can use this minimum angular vision for better understanding and optimising the psychophysiology of the viewing.

A known viewing distance parameter in CCTV recommends a viewing distance of around seven times the monitor height. So it is important to understand that viewing distance is a crucial parameter for the psychophysiological experience of seeing details in an image. It is of no use if a viewer gets closer to the monitor, but it is also not going to get any better if he is positioned farther away from the monitor screen.

For the analogue PAL signal, with 576 active lines, the corresponding optimum details viewing distance is obtained from the 5 to 6 lp/mm, projected at the distance where the monitor screen is. So, for example, if we use the 7 × picture height rule, for a 15″ (38 cm) CCTV monitor, whose picture height would be around 23 cm, the recommended viewing distance is around 1.6 m. The best resolving power of the human eye at that distance is simply a factor of 5 from the 0.3m distance when viewing 5~6 lp/mm (because 0.3 m goes around six times in 1.6 m). This equates to around 1 lp/mm (line pairs per millimetre), or 0.5 lines/mm, which is roughly what each of the 576 active TV lines will physically occupy on the screen. This assumes a high-quality, high-resolution monitor of course.

Coming much closer to such a monitor will not show any more details, nor will going further away be of any advantage. Coming closer to a CCTV monitor has the same effect as when a much larger monitor is put in place. If you decide, for example, to put a 21″ colour monitor at the same distance of 1.6 m instead of the 23″ monitor, the picture clarity and video details will appear worse to the viewer. The optimum resolution distance for a 21″ CCTV monitor would be 2.1 m.

Similar logic and calculation will show that viewing a high-resolution computer screen such as a CRT with dot pitch of 0.21 mm is optimum at a distance of around 0.6 m. The majority of LCD monitors do not have such a fine pixel detail, but typically around 0.28 mm, so they actually will look slightly better if viewed from around 1 m distance. A much better looking picture would be shown on a computer screen (viewed from the same distance), where a typical XGA monitor of 1024 × 768 pixels has an actual equivalent resolution of 92 DPI.
To see how this was obtained, divide 1024 pixels with the width of a 14″ LCD notebook screen. Computer screens therefore have a larger area (pixel count) but also a higher frame rate than we use in CCTV. Please note that in order to display such a high-quality image on the screen, the computer has to have a high quality video card with sufficient memory capable of processing that number of pixels (1024 × 768) with the number of colours sufficient to replicate a live scene (24-bit colour, equivalent to 16.7 millions of combinations of the RGB). And another important note: such a display neither complies with PAL nor with NTSC analogue standard, but it is referred to as a computer XGA graphic display.

Printed material has even better resolution per mm than any kind of monitor. This is why you sometimes may see better details when a video image is printed out on a high-quality ink jet photo paper than looking at the same image on a CCTV monitor. We can express monitor resolution in DPI, but it would not make the same sense as in printing. This is primarily because when viewing a monitor screen we (usually) do not stand as close as when reading this book. So, let us assume we do have a very high-resolution CCTV camera image displayed on a very high-quality CCTV monitor, which would be quoted as having 500 TV horizontal lines of resolution. If the monitor has, for example, 38 cm (15″) diagonal screen size, the 500 TVL resolution image means 666 vertical lines across the 30 cm screen width (30 cm = 11.8″). The 666 lines divided by the 11.8″ gives 56 DPI! This resolution is close to the highest we can get when displaying analog video signal, and it is defined by the video standard (PAL/NTSC).

In order to print an ITU-601 compliant TV frame (this basically what we call 4CIF in CCTV) on an ink jet printer, we need to understand this technology too, so that we can optimise the printout quality. It is logical to expect that print size and resolution quality can easily be calculated, since we know, for example, that our ink jet printer has, for example, 1440 DPI. This is not so, however. The dots per inch as described in the specification on your fine ink jet printer (like 720 DPI, or 1440 DPI) refer to the finest dots that can be produced by each of the cyan, magenta, yellow, or black ink nozzles of the ink head. Adding to the confusion is the fact that this is not the same as the DPI described for magazine and book printing. The actual “natural” colours of the ink jet printouts are obtained by a dithering process, which is actually spraying and mixing the ink jet dots with various sizes and combining them to produce a resultant colour. In essence, the ink jet colour printers are binary devices in which the cyan, magenta, yellow, and black dots are either “on” (printed) or “off” (not printed), with no intermediate levels possible. This is conceptually different from the RGB phosphor on a CRT, where each phosphor can have a variety of intensity. A “binary” CMYK printer can only print five “solid” colours (cyan, magenta, yellow, black, and white). White is actually the nonprinted paper background colour, but that is also used. Clearly, this is not a big enough palette to deliver good colour print quality, which is where half toning comes in. This is still the case even with the new photo quality ink jet printers that have additional two colours: light-cyan and light-magenta (for better human flash reproduction). Half toning algorithms divide a printer’s native dot resolution into a grid of halftone cells and then turn on varying numbers of dots within these cells in order to mimic a variable dot size. By carefully combining cells containing different proportions of CMYK dots, a half toning printer can “fool” the human eye into seeing a palette of millions of colours rather than just a few.

A simple rule of thumb by some imaging people, such as Adobe, is to divide the ink jet DPI as specified by the ink jet printer manufacturers with 4 to get the “real” colour dots per inch. Practically, this means that a 720 ink jet DPI printer can reproduce 180 colour dots per inch. To achieve highest quality, it is important to use high-quality printing paper. For best results, inks and photo paper of the corresponding printer manufacturer are recommended.

An important conclusion to make here is that when a digitised and compressed image is exported and given as evidence to a third party (police, for example) it is desirable to have it in the original format, or at least exported to BMP so that no additional compression artefacts are introduced. If you are comparing various image compressions, the most objective way of comparing them is that either when the image is printed out on the same high-quality photo paper, or when they are compared on a computer screen, both should be exported in a non-compressed BMP format.

Vlado Damjanovski, © 2007

• BROWSE / IN TIMELINE
• « Dynamic range of solid state imaging devices
• » Web Site for the Development of Australian CCTV Standards

• BROWSE / IN Viewing quality