TopTechnical DictionarySignal to noise ratio (SNR)

Signal to noise ratio (SNR)

SNR is the power ratio between the signal strength and the noise level in a specific band.

 

SNR is expressed in dB using the following formula:

 

S/N = 10 log ⁡(S/N) dB
 

where:

SNR – signal to noise ratio
S – usable signal strenght
N – noise

 

The higher the SNR value the better. A recommended SNR value for video signals is 48 dB (virtually noise-free). For audio, negligible noise is achieved at approx. 96 dB (CD quality). The main sources of noise in CCTV systems are transmission noise, image sensor noise, and microphone noise as well as noise generated by electronics and quantization. However, those are secondary.

 

Transmission noise

 

The signal strength in a cable decreases with distance. Even if the signal is amplified at the end, the noise generated by the electronics is also amplified, reducing the SNR. The distance, at which the signal is transmitted, the type of the transmission medium used and the frequency of a transmitted signal can affect the SNR. Advanced HD analogue systems require higher band (approx. 50 MHz) and are more susceptible to noise than older analogue TV systems (approx. 6 MHz band).

 

Another important criterion is the cable used. The least efficient and the cheapest is a twisted pair cable with an attenuation of 20 dB over 100 m. It means that the signal strength at 100 m will be 100 times lower. If the signal amplitude at the beginning is 1 V, at the end it will be just 10 mV. A coaxial cable is slightly better with a 6-8 dB attenuation at up to 100 MHz band. An optical fibre with a 0.1 to 0.5 dB attenuation over 1 km and immune to electromagnetic interferences is the best solution available.

 

Noise in the photosensitive elements

 

In this case, the size of the image sensor is crucial - the larger the image sensor, the more light it can take in and the lower the noise. Its design is also important.

 

Two types of image sensors are available: CCD (Charge-Coupled Device) and CMOS (Complimentary Metal-Oxide Semiconductor). Both have their pros and cons. The CCD image sensor is an improvement considering the noise properties. In CMOS image sensors, part of the pixel is taken by the pixel control system which means less light reaching the photosensitive surface and more noise. This problem was eliminated in ERMOR image sensors where the pixel control system is located under the photosensitive element. CCD image sensors have other disadvantages including: relatively low speed, high power input, and high price. Due to their properties, CMOS image sensors are slowly replacing CCD image sensors. Most CCTV cameras use CMOS image sensors.

 

A camera’s noise level also depends on the image sensor quality. The manufacturers use advanced digital video signal processing to reduce the camera’s noise level.

 

Choosing the right camera depends on the amount of light available. The camera’s noise level can be reduced with external lighting that may require additional investment.