CCTV-101, How Cameras work
Cameras come in many shapes, sizes and specifications. This 101 examines the basics of cameras and features used in 2020.
In this article, we review the basics of cameras, including:
- Core components of cameras
- Analogue vs. IP cameras
- Common components
- Form factors
- Video quality concerns – Day, Night, Sunlight, Shadows
- Bandwidth basics
Core Components of Cameras
Every camera has 3 core components – lens, imager and transmitter.
The lens determines how wide or narrow a view the camera will capture, referred to Angle of View. They are identified by their focal length, which is measured in millimetres (e.g., 2.8mm, 8mm, 12mm, etc.), with smaller focal lengths capturing wider areas and longer focal lengths capturing narrower areas. See our Lens Focal Length Tutorial for more details.
There are two main types of lenses, fixed and varifocal. Fixed lenses are set to one field of view and cannot be changed or fine-tuned to improve the camera’s view. Varifocal lenses allow users to adjust the lens so it may be zoomed in if longer distance is required or zoomed out to capture a wide area.
Varifocal/motorized lenses are much larger than fixed lenses and take up more room, resulting in larger overall camera size, shown here side by side:
Imagers Capture Light
The camera’s imager is the component which captures the light allowed in by the lens, also called an image sensor or simply sensor. The sensor contains a grid of pixels which capture the intensity of light and transmit this information to the camera’s processor.
Imagers are generally referred to in fractions of an inch, measured diagonally from corner to corner, e.g. 1/3″, 1/2″, etc., with most cameras today using imagers between 1/2″ and 1/3″.
Imager Pixel Size Differences
The size of the pixels contained on an imager varies widely, depending on the imager’s size and resolution. All things being equal, the bigger the imager, the bigger the pixels it contains. However, if you add more pixels (e.g., going from 1080p to 5MP) and the image size stays the same, the pixel size decreases.
The transmitter is the portion of the camera which sends video to the recorder or viewer, either converting it to Ethernet in IP cameras or into a high definition analogue format (detailed below). This component could be contained directly on the camera, or attached to the end of a length of cable (called a cable whip).
For example, the two cameras below are very similar in construction, resolution, etc., but one outputs video via Ethernet and the other analogue using a BNC connector.
Analog vs IP Cameras
While analogue and IP cameras have the same core components, the fundamental distinction between the two camera types is encoding – IP cameras have video encoding built-in, analogue cameras do not.
When discussing “analogue” cameras in 2020 and beyond, typically we are referring to HD analogue, which is a generic term for cameras which transmit HD video (720p to 4K currently) over standard coax cables. The main advantages of these models over IP cameras are the ability to use existing coax cabling (making upgrades cheaper) and lower average cost.
However, IP cameras generally offer a wider selection of form factors and features. Audio, analytics, and I/O are rare features in HD analogue models, but common in IP. Additionally, IP cameras may be connected anywhere on the network, instead of requiring a home run to a recorder using coax cable.
For IP cameras, which constitute the overwhelming majority of cameras deployed in 2020, the most fundamental way to do encoding is to use a ‘system on a chip’ typically called an SoC.
An SoC is a single chip (shown below) contained in the camera which essentially functions as its “brain”, controlling image processing, compression and encoding, network access, storage, audio, analytics, and more. All cameras use some sort of SoC, whether IP or analogue.
SoCs are located inside of cameras on one of their mainboards and normally not exposed without doing some disassembly of the camera. Beware, this disassembly generally voids the camera’s warranty and is shown here for demonstration!
That being said, if you need to confirm what chip is inside your camera(s), disassembly may be your only option. We demonstrate this process on multiple cameras in this report: How To See If Your Camera Uses Huawei Hisilicon Chips.
Integrated Infrared (IR)
Many cameras include built-in infrared illuminators which add light to the scene to improve low light images without adding visible light to the area.
The upside is brighter images and better details. The downside is that only black and white images are possible with IR and some illuminators may cause overexposure. We discuss these issues in more detail below.
Many cameras include audio capability, which may be in several different forms, including built-in microphones, input/output jacks, or terminals.
Built-in mics are typically regarded as the simplest method of recording audio as they do not require additional installation of external mics, but they are generally lower performing (noisier and lower sensitivity) than external purpose-built mics.
Using microphone inputs allows users to select a purpose-built mic appropriate to the installation, such as vandal resistant wall mount, recessed ceiling, door phone form factor, etc.
Note that while audio is frequently included in cameras, it is rarely used. In the US and elsewhere, specific requirements for audio recording must be met, and often the party being recorded must actively consent to the recording.
We cover these issues in more detail our Audio Surveillance Guide.
Lots of cameras support onboard storage using SD cards, allowing video to be stored locally on the camera (called Edge Storage).
There are two main ways of using this recording:
- Full time/no VMS/NVR: Edge storage may be used as the only storage, with no central NVR/VMS, sometimes used in small systems.
- Redundant recording: Or it may be used for redundancy, recording to the camera while the network is down and the recorder cannot be reached, then offloaded when network connectivity is restored.
Like audio, while many cameras include SD storage, it is not widely used. SD card capacities have historically allowed for only a few days’ recording, while hard disk drives used in NVRs allow for weeks or months. Additionally, early adopters of edge recording suffered from many reliability issues, which led to a lack of trust in the technology.
However, some manufacturers have begun using edge storage as part of their end-to-end cloud-managed systems, such as Verkada and Rhombus. These manufacturers store video on the camera and allow users to retrieve it via the cloud, instead of uploading all video to be stored in the cloud.
Many SoCs used in modern IP cameras include support for video analytics onboard, but the quality of these analytics has historically been fairly weak, with numerous false alerts on shadows, lights, etc. (see our IP Camera Analytics Shootout).
However, new SoCs with AI engines built-in has begun to change this, with improved performance in harsher conditions. Additionally, more advanced analytics such as facial recognition and vehicle classification are now built into some SoCs. We expect to see this trend continue as more chip options become available and component prices decrease.
See our AI/Smart Camera Tutorial for more details on the basics of analytics.
Cameras are available in several different form factors, with domes, turrets, and bullets the three most common today. Box, cube, covert, and door station cameras are also available, but less common.
Each of these form factors has its pros and cons:
- Domes: Domes’ main advantages are improved aesthetics and better availability of outdoor/vandal options than others, but generally are limited in longer lens options compared to boxes/bullets and must be routinely cleaned to prevent image quality issues.
- Bullets: Bullets are next most common, with the main advantage of longer IR range than domes (generally, not always) and more overt appearance for greater deterrence. However, they are easier to knock out of position and typically not IK10 vandal rated, unlike domes.
- Turrets: Turrets are a relatively new variant of dome which place the imager and IR illuminators into a spherical gimbal. These models typically have the same advantages and disadvantages of domes, but maybe more resistant to IR glare and dirt/scratches than domes due to the smaller surface area.
- Box: Box cameras biggest advantage is flexibility, with the ability to change lenses to cover very wide or very narrow areas, greater than dome or bullet models. However, they are more complex to install, requiring a separate enclosure outside, and have similar disadvantages to bullets in susceptibility to vandalism and aesthetics.
- Cubes: Cubes are essentially mini box cameras without the ability to change lenses, eliminating their main advantage. However, cubes still see some use in lower cost or DIY installations.
- Covert/door station cameras: These form factors are generally used only in the specialized applications they are intended for, such as hiding the sensor in walls/ceilings/doorframes, or placed outside entry points. These segments have grown, but remain niche.
For more details on these form factors and their advantages and disadvantages, see our Camera Form Factor Guide.
One common differentiator amongst cameras is resolution. Camera manufacturers specific resolution based on the physical number of pixels in the imager also called pixel count (not the actual resolving power of the camera). This is most often specified in megapixels, the number of millions of pixels (e.g., 3 megapixel, 5MP, 8MP, etc.), though some use standard HD terminologies, such as 1080p or 4K.
The most common resolution in use today is 1080p (~2.1MP), which, generally speaking, offers the greatest balance of image quality, pixel count, and low light and WDR performance (discussed below). 1080p cameras frequently rank in the top spots of our WDR, low light, and IR shootouts, for example.
Higher resolution cameras are growing, with 3-5MP models jumping significantly in recent years, driven by better chip and sensor availability, as well as lower prices. Additionally, the availability of low-cost 4K and 8MP models in the past 1-2 years has driven some growth in even higher resolutions. However, for the time being, 1080p remains much more common.
Note that VGA is no longer common except in thermal cameras, but is included here for the reference of what ‘standard definition’ refers to.