Solving Crime with Only a Hum: The Power of the National Grid

Audio recordings can sometimes be crucial evidence in criminal investigations – footage of a suspected terrorist planning an attack, CCTV footage of a crime, or a secret recording capturing an admission of guilt – but how can the police ensure that recordings are genuine and haven’t been tampered with? Incredibly, forensic scientists have created a technique that can analyse the background ‘hum’ of the national grid to authenticate audio recordings.

The power transmitted throughout the UK is all provided via one source, the National Grid. The frequency of this electricity is typically about 50Hz, and this frequency will be picked up in the background of any audio recordings that are made near a power source, e.g. a pylon, plug socket or light. Due to the unpredictable nature of supply and demand, this frequency varies over time – for example during the ad breaks of popular TV shows, millions of people switch on their kettles and the frequency reduces due to the increased demand.

Crucially, the pattern of these fluctuations is unique over time, so when an audio recording is compared to these historical fluctuations (essentially a digital audio timeline) it can be given an accurate date and time stamp. This is a technique known as Electric Network Frequency (ENF) analysis, and can be used to determine the authenticity of audio recordings – were they recorded when they are claimed, and is the recording continuous or has it been edited?

For the last 10 years, the Metropolitan Police forensic lab has been continuously recording the background sound of the UK’s mains electricity in order to be able to make these comparisons. This technique works particularly well in the UK, because one national grid supplies the entire country.

Dr Alan Cooper, a senior digital forensic practitioner at the Metropolitan Police explains that ENF was recently a critical tool in an important court case. Undercover police had submitted audio recordings of an arms deal as evidence, but the defence had claimed that the police had tampered with the recording. Dr Cooper explains:

“We carried out various forms of analysis, including the mains hum frequency analysis and we found some good quality signals, and that the alleged date and times of the recordings matched with the extracted data from the recordings themselves.” 

This proof that the audio recordings were genuine ultimately resulted in the gang being found guilty and jailed for a total of 33 years for the supply of firearms. Dr Cooper explains that this technique is now starting to be used widely by the force and others across the world, but acknowledges that with the constant flux in technology there will always be challenges ahead.

For more information, the full paper can be found at: The Electric Network Frequency (ENF) as an Aid to Authenticating Forensic Digital Audio Recordings — an Automated Approach