CCTV camera technologies have advanced rapidly alongside consumer electronics like camcorders and digital cameras. It can be unnerving to find ourselves gripped by a tide that sweeps us out of the Analogue Sea into the enormous Digital Ocean. However, with the navigation knowledge offered below you should remain afloat.
CCD or CMOS?
The ‘imager’ or ‘sensor’ just inside the front of the camera is commonly either a CCD or CMOS chip. For those who are desperate to know… ‘Charge Coupled Device’ and ‘Complementary Metal Oxide Semiconductor’. Which is better? In truth, each has its pros and cons.
For instance, CCDs using the ‘interline transfer’ method are cheaper but can exhibit vertical streaks in images around bright points of light. This isn’t true of their ‘frame transfer’ siblings but these are more expensive. It’s not all plain sailing for CMOS chips which can show ‘rolling shutter’ problems where fast moving objects have warped shapes. Nonetheless, in the real world, be aware of these but usually either is fine because light sensitivity, dynamic range and signal-to-noise ratio and many others factors in a camera’s design far outweigh any agonising you may do over CCD vs CMOS.
More video surveillance resources
Historically, industry standard sensor sizes are listed such as: 1/4-inch, 1/3-inch and 1/2-inch for ‘standard definition’ video where the picture width:height (known as ‘aspect ratio’) is 4:3 for non-widescreen video. Analogue cameras comply with the CCIR and PAL standards resulting in images made up of 720 pixels horizontally and 576 pixels vertically. Known as ‘D1’ pixel sampling, this is now referred to as SD (Standard Definition) so as not to confuse it with HD (High Definition) which uses higher numbers of pixels, described later.
Alternative video standards such as NTSC in the US created their own ‘SD’ size owing to their longstanding 525-line television systems. Smaller than D1, their images are digitised at only 640×480 pixels. You probably know this as VGA too when used to describe computer display sizes. Be aware when implementing your Operational Requirement that many IP video systems employ cameras that create these lower resolution images. However, they might employ higher resolution cameras such as SVGA (800×600). The list goes on:
XGA (1024×768) 0.79MP
SXGA (1280×1024) 1.3MP
SXGA+ (1400×1050) 1.5MP
UXGA (1600×1200) 1.9MP, etc
Listed alongside here are the numbers of megapixels (MP) for each, from simple multiplication. You can see how quickly we move into megapixel territory. Compare these with D1 (0.41MP) and you could conclude that one UXGA can replace between four and five D1 cameras. Stop and think some more. If your Operational Requirement from our previous article reports that you need, for instance, a person to be 100% of picture height for ‘Identification’ with an SD PAL camera, we can say the person is simply 576 pixels tall. Now with that yardstick in hand, how does our 1.9MP UXGA camera compare? Giving us 1200 pixels top to bottom, the 576 pixels needed for Identification are 48% of that 1200. Call it 50%. A half, meaning that our shiny new UXGA camera might replace only two D1 cameras. Clearly not the four or five that we might have been led to believe by manufacturers’ megapixel marketing blurb. Important to realise.
Digital cameras have given us ever larger imaging chips, now with video capabilities. Megapixel sizes are increasing and these CCTV systems are getting more affordable as pixel-crunching computer power advances. 3, 5, 11, 16 or 29MP: how many would you like in one camera? At this level, high quality photographic lenses should be fitted for good resolution and efficient light gathering. On the latter point it is clear that if more pixels are fabricated onto a CCD or CMOS chip of a given size, then each pixel must become smaller. This leads directly to less light-gathering capacity for each pixel and a camera offering diminished low-light sensitivity. Good lighting will be vital.
Image refresh rates
Analogue SD cameras have always produced 50 images per second. These 50 ‘fields’ each comprise 312.5 television lines that are then ‘interlaced’ to create a 625 line ‘frame’ 25 times a second. For years this has been taken as read. Now, does a 29MP camera give you 25 ips (images per second)? No. Currently, 2 ips, limited by the sheer amount of data that needs to be processed. Bear this in mind when designing a MP system. Check with manufacturers because this might be too slow to meet your Operational Requirement if you need to capture fast-moving details.
CCTV’s use of digital video means it is no longer constrained by the PAL limitations. Consumer level cameras can routinely offer as high as 60ips in HD recording, and some 120ips at the lower VGA resolution, which we can expect to increase. HD is now commonplace so let’s look at it in more detail.
HD offers widescreen images with a 16:9 aspect ratio: a shape that is almost universal in domestic TV. There are two sizes: HD720 and HD1080 (known as ‘full HD’) with different image sizes: 1280×720 pixels and 1920×1080 pixels giving us 0.9MP and 2.1MP, respectively. Both broadcast standards are, of course, capable of 25 ips, but choose carefully with your ‘CCTV standard’ suppliers! The designation HD1080i refers to interlaced picture scanning (see above) while HD1080p denotes ‘progressive’ scanning whereby television lines are treated sequentially as opposed to interlaced odd-even-odd-even, etc. Where the progressive scanning takes ‘instantaneous snapshots’ it can avoid the comb-edged tearing that is characteristic of moving objects viewed by interlaced scanning.
As promised, a word about the options for signal formats that emerge from these cameras. The many transmission methods were outlined in our previous article. Analogue signals are CCIR/PAL standard as composite video with amplitude 1 volt peak-to-peak. This cannot cater for the newer HD signals. One format for HD is the Serial Digital Interface (HD-SDI) standardised by SMPTE (Society of Motion Picture and Television Engineers). Interestingly, existing co-axial cables might be used to carry this over limited distances. Practically all of the massive megapixel camera signals use IP video over fast data networks. Such IP video networks are now commonplace with HD and digitised CCIR/PAL video too.
For CCTV users who are willing to plot a course into new waters, the advances in camera technology make for an exciting voyage of discovery but, as ever, keep a well-trained eye peeled for sharks.
IFSEC International 2018
Join IFSEC Global live at Europe's only dedicated integrated security event. Register for free.
Meet over 600 exhibitors, test more than 10,000 of the latest security products, and discover best practice and future trends in an unrivaled seminar programme.
- Frank Gardner to chair the Keynote Arena
- Former US Secretary of Homeland Security to take Keynote stage
- Live attack testing in the LPCB/BRE Global Attack Zone
- Your chance to get hands on with the latest security innovations thanks to the brand new Show Me How feature