Feeds

back to article Police use 24/7 power grid recordings to spot doctored audio

Forging audio recordings is a lot harder than it used to be, thanks to a new method of authenticating recordings based on the buzz of the electrical power grid at the time they were recorded. The oscillations of alternating current (AC) produce a distinct frequency – 50Hz in the UK, 60Hz in North America – that varies slightly …

COMMENTS

This topic is closed for new posts.

Page:

Trollface

Re: Who holds the reference data?

Let's hope they archived it and signed it with a fully independent 3rd party otherwise they should sack their IT staff.

2
0
Gold badge

Re: Who holds the reference data?

I give that possibly one, maybe two appearances in court before some enterprising hack creates a website where you can download patterns from all over the world.

Having said that, I lack facts here. Until I see a scientific paper that explains how it works this has about as much value to me as the use of an IP address to prove identity. That only works if you have really expensive lawyers, and even then it's dicey.

0
0
Silver badge

Re: Who holds the reference data?

national grid.

0
0
Bronze badge
Thumb Up

Re: Who holds the reference data?

Best point made so far; the first thing any barrister should do is call into question the police's "proof".

Given criminal cases rest upon "beyond reasonable doubt" it is highly unlikely any case would stand or fall on the issue of a tape recording and some correlation (or not) alone and, if it did, that it would withstand an appeal. It is therefore rather moot in the scheme of thing.

Of course the police are very likely to use claims of being indisputably able to prove their case and get a confession with the promise of going easy if one fesses-up and that's where it has best application. It's often easy to have a suspect to believe a jury won't believe them than get the jury to not believe them. It's simple social engineering.

0
0
Bronze badge
FAIL

I'm all for building enormous Faraday cages around every courtroom in the land to avoid contamination of evidence. .

0
0
Bronze badge

Speed variation, mechanical or digital

Wow and flutter and tape stretch, instability of sample clock frequency - it is trivially easy to introduce (accidentally or deliberately) sufficiently large variations into a recording that it wouldn't match the reference, even before you look at obscuring, removing or faking the tell-tale frequency. It seems to me that an expert witness for the defence can easily discount the validity of the technique.

2
1
Big Brother

It's more useful to think of them using the phase change than the frequency.

The grid tries really hard to keep the frequency at 60.000Hz (or 50Hz, for the countries that are a little slower).

If the load increases, the phase lags. This indicates to the power plant that they need to throw another shovel full of coal onto the fire and let a little more steam into the turbine.

It sounds easy to record that phase difference and match the pattern, right?

Except that there isn't one power plant. The whole point of a grid is there are thousands connected together. So you have thousands of sources trying to push the phase a little faster. Each substation is seeing a different mix of the phase variations. All mixed together, with extra noise added by local loads being switched on and off.

I can see how you might be able to show a proof of concept that invalidates a recording (but not authenticates its veracity) made nearby, within the same substation service area.

The next challenge is the audio recording. Today that's largely digital. Analog audio is hard to do well on digital chips. There is quite a bit of non-linearity. Some of the distortion results in phase changes with sounds level. The audible effect is more pronounced at high frequencies, but it will overwhelm the minuscule phase difference of the power line hum. Even the sample clock of the D/A digital will have enough jitter, correlated with the processor workload and other power draws, to overwhelm any otherwise detectable hum phase pattern.

Perhaps that's why they are publicizing this now. It's a technique that sounded promising, they invested a lot of effort, only to have progress has render it completely useless. They might as well recover whatever value they can by dissuading people from trying to forge recordings.

3
1
Silver badge

you say phase lag, i say frequency change :-)

y' cant have one without the other. in order to go out of phase, something needs to slow down/speed up

all the power stations have to be in phase (theres kit to auto lock them out if they arent - can you imagine how loud the bang would bee if you closed 500MW onto a grid that was 180degrees out!)

i hope your sampler is running a little faster than 60Hz btw!

2
1
142
Facepalm

Good gods, Donald!

How inaccurate do you think modern digital audio gear is??

Mains frequency drifts, due to the phase shifts you're mentioning, between 49.9 and 50.1Hz. That's a piece of piss to record, and has been for probably 40 years! Indeed - you can audible hear the pitch change as you're working on audio with high levels of mains hum recorded over even a 15 minute period of time, if you jump back and forth between different sections. Hell, you can see it visually in the amplitude waveform, without even resorting to frequency analysis!

Don't believe me about the frequency variation? Have a look here - http://www.nationalgrid.com/uk/Electricity/Data/Realtime/Frequency/Freq60.htm

0
0
142
WTF?

Re: Good gods, Donald!

And about phase - If I set a signal generator to give me a 50Hz sawtooth wave, and I record that 50Hz sawtooth wave, I will get a phase-perfect 50Hz facsimile of the sawtooth wave. NOT, as you seem to be implying, a phase shifted mess of a waveform.

0
0
142
Angel

Re: Good gods, Donald!

I may have been a tad harsh last night, lol - I've heard people the same tales about digital audio and phase before - but the reality is that unless you're building a crude ADC yourself, you'll only encounter phase issues right at the nyquist frequency.

0
0

Does it work with tape?

Analogue's not a problem, but mains tape recorders with synchronous capstan motors will track any frequency variations on the supply, so cancelling out the variations in mains hum on the tape. On the other hand it should be a lot easier to detect edits in analogue recordings.

1
0
Silver badge

And all sound recorders are synced to ....

some magical central source ??? I'm guessing that this will make a guaranteed timestamp once in a backhander.

0
0
Bronze badge

Defeatable, if you know about it.

Simply removing the hum isn't going to work - it'll still be evident that the hum was removed deliberatly, which would be very suspicious. No, you just need to make sure that the hum is gone in a way that appears accidential.

So make your recordings on a laptop or battery-operated device, in a room with a continuous background noise (say, a moderatly loud fan) and make sure to record to a 32kbits MP3. Lets see how well they can extract the hum from *that*.

0
0

Re: Defeatable, if you know about it.

Sure, except I expect that in a lot of these cases, the voice on the recording has to be recognizable for the recording to work. ("We have little Johny, listen to him." only flies if you recognize little Johny) And I would be _really_ worried about that fan, unless it is powered by batteries or a guy huffing and puffing on a bicycle.

Anyhow, I'm 100% confident it is possible to defeat this technique under at least some circumstances, just as it is possible to avoid leaving fingerprints at least some of the time. Doesn't mean that it isn't a useful tool to have in the toolbox.

0
0
Anonymous Coward

In Audacity you can easily add a dehum filter

I just found this nifty nyquist plugin within the audacity forum to wipe out 50 or 60 hz frequencies including most of it's harmonics. Just paste the snipplet below into a text editor and save it as dehum.ny and then place it within the nyquist folder within the Audacity installation. You'll find it then within the effects menu of Audacity:

;nyquist plug-in

;version 1

;type process

;name "50/60Hz dehummer 2.0"

;action "Removing Hum harmonics, ( this may take some time ) ..."

;info: Unintentional reverb effect. Gibbs ringing on transients.

;info: if you find a cure for the reverb please email me: uvw111-reverb@yahoo.co.uk

;control choice "Select mains frequency" int "0=50Hz (UK) 1=60Hz (USA) 2=choice" 0 0 2

;control bfreq "Base frequency" real " 10-10000 Hz " 50.0 10.0 10000.0

;control a "Amount of hum removed" real " 1-100 base Q=(100/a)^2" 15 1 100

;control v "Anti-reverberation setting" real "1-100 " 25 1 100

;control frac "Fraction of spectr" real " 0.01 to 1 " 1.0 .01 1.0

(setf freq (cond ((= choice 0) 50.0) ((= choice 1) 60.0) ((= choice 2) bfreq)))

(setq que (/ 10000.0 (* a a)))

(setq anti (/ 10000.0 (* v v)))

(setq mysound s)

(setq r *sound-srate*)

(setq itern (truncate (* (/ (/ r freq) 4) 2)))

(setq d (/ (float itern) anti))

(setq iter (truncate (* (float itern) frac)))

(defun dehum (mysound freq iter)

(dotimes (i iter mysound )

(setf mysound (notch2 mysound (* freq (1+ i)) (* que (1+ (/ i d)))))))

(if (arrayp s)

(vector (dehum (aref s 0) freq iter)

(dehum (aref s 1) freq iter))

(dehum s freq iter))

0
0
142
Holmes

Re: In Audacity you can easily add a dehum filter

Yip. The trouble being though, of course, when you notch out that many harmonically related frequencies from a human voice, makes it sound like they're talking in a 10 foot diameter plastic pipe...

(Sherlock for his pipe fondness, not his sarcasm! ;-) )

0
0

Clearly, there are folks here with no idea about grid frequency maintenance. Power grid frequency wobbles a lot. Long term, it is very stable, as it is kept that way, as already noted, for the benefit of clocks. But short term... Wobbles. And as such even amongst the wow and flutter of a VHS VCR, or the clock inaccuracies of a crystal chip, there is a lot of correlation that can be done...

Have a look at some real time grid frequency data here: http://www.mainsfrequency.com/

1
0
Anonymous Coward

I'm impressed by all the audio lurkers

So...Mr Terrorist just needs to make their recordings using battery powered gear then, right? No hum, no mains, no long stay at Her Majesty's pleasure.

I knew the voice memos thing on my phone was there for a reason.

0
0
Anonymous Coward

laptop in the garden shed for my next blackmail recording then.

0
0

Page:

This topic is closed for new posts.