The narrator discusses the concept of using the mains hum, or the 50 Hz or 60 Hz alternating current (AC) running through power lines, as a form of watermarking. This hum is a result of the imperfections and stray electromagnetic fields in the power transmission, which cause a little bit of power to become vibration, creating a sound at a frequency of 50 Hz or its multiples. This hum can be heard coming from power lines or speakers that aren't quite plugged in properly, and it can be annoying for people recording audio.
The frequency of the mains hum isn't exactly 50 cycles a second, as the grid frequency shifts slightly and constantly over time. This shift is only a tiny fraction, but it's there and the whole grid runs at the same slightly wobbling frequency. The narrator proposes a theoretical method of logging the grid frequency every second of every day, storing it in a database, and comparing it to recordings with mains hum on them. This could potentially create a perfect watermark to prove the time when anything was recorded.
The narrator mentions that the National Grid has recently released seven years of data, which can be used for this purpose. The data would be processed using signal processing to identify a component around 50 Hz or one of the harmonics. However, the narrator advises against presenting oneself as a forensic expert in court, as this type of evidence has been used in court to date.
The narrator then demonstrates the process with some raw footage from friends who make videos. The footage is unprocessed and unfiltered, and the forensic team is asked to timestamp it first. The team uses a signal processing algorithm to extract the peak and constrain it so that it can't suddenly jump up. If a recording was taken on a certain day, then the team only needs to give it a day's worth of national grid data. If the exact time is unknown, the team can process everything, but it will take a bit longer.
The narrator concludes by stating that this forensic analysis isn't a magic tool that always works. Compression can stop it from working, and there might be no background mains hum at all. However, the narrator loves this because it's an accident of technology that's created a tool that would sound like science fiction 20 years ago.
1. The alternating current running through power lines changes direction constantly, completing a full cycle back and forth 50 times a second in Europe, but in some other parts of the world, it can be 60 times a second [Document(page_content="00:00:00.08: the alternating current running through\n00:00:01.68: these power lines changes direction\n00:00:03.68: constantly a full cycle back and forth\n00:00:06.00: 50 times a second 60 in some other parts\n00:00:08.16: of the world but here in europe 50 times\n00:00:09.68: a second but that transmission isn't\n00:00:11.60: perfectly efficient there are\n00:00:12.88: imperfections and stray electromagnetic\n00:00:14.56: fields and a little bit of that power\n00:00:15.92: becomes vibration which makes sound at\n00:00:18.48: that frequency of 50 hertz or 100 or\n00:00:20.80: other multiples of 50. that's the mains\n00:00:23.20: hum that you can sometimes hear coming\n00:00:25.12: from power lines or maybe from speakers\n00:00:26.96: that aren't quite plugged in properly\n00:00:28.96: also that hum isn't exactly 50 cycles a\n00:00:32.08: second the grid frequency shifts\n00:00:33.92: slightly and constantly over time as the\n00:00:35.60: engineers in charge balance supply and\n00:00:37.12: demand it's only a tiny fraction of a\n00:00:39.20: shift but it's there the whole grid runs\n00:00:41.28: at the same slightly wobbling almost 50\n00:00:43.92: hertz no matter where you are\n00:00:46.08: that mains hum can be really annoying\n00:00:48.24: for people recording audio you can hear\n00:00:49.84: it in the background sometimes\n00:00:51.68: but if you were to log what that\n00:00:53.76: wobbling grid frequency was every second\n00:00:55.92: of every day keep all that in the\n00:00:57.52: database and then compare it to\n00:00:59.52: recordings with mainsum on them\n00:01:01.84: then in theory you'd have a perfect\n00:01:03.84: watermark to prove the time when\n00:01:05.68: anything was recorded i talked to one of\n00:01:07.36: the forensic experts who works with it\n00:01:08.96: and also set them a challenge\n00:01:11.76: historically uh what's had to happen is\n00:01:14.08: we've had to record the grid frequency\n00:01:16.16: data ourselves by literally plugging a\n00:01:18.24: recording device into the mains\n00:01:20.16: the national grid have recently released\n00:01:21.76: their data seven years recorded right at\n00:01:24.00: the source it's fairly simple actually\n00:01:26.00: signal