Today, for the first time ever, we can reveal some of the innermost workings of TV Licensing's sinister and secretive television detector vans.
Capita Business Service's Chris Christophorou, acting on behalf of the BBC as Licensing Authority, applied to Canterbury Magistrates for a search warrant under Section 366(1) of the Communications Act 2003 on 28th June 2011.
In order to obtain that warrant Christophorou had to provide evidence to a Magistrate, on oath, that an offence was suspected at their target address. In doing so he revealed that TV Licensing's detection capabilities rely, at least in part, on light emitted from television receiving equipment.
We consider the science behind the BBC's TV detection equipment to be very dubious indeed - a stance further reinforced by Christophorou's admissions below. We are exposing it for public scrutiny, because of the distinct possibility it has been given undue evidential weight when TV Licensing apply for search warrants. We feel even more of a duty to share this information in the week the transparency, accountability and lawfulness of the entire UK media has been brought into focus.
Below is the complete text of the sworn statement Christophorou gave Magistrates on 28th June 2011:
______________________________________________________________
THE INFORMATION of Chris Christophorou, this being his deposition in support thereof, is taken on oath this 28th June 2011 before me the undersigned, a Justice of the Peace for Canterbury Magistrates Court, on an application for a search warrant authorising any employee of Capita Business Service Ltd, authorised on this behalf by the British Broadcasting Corporation, with or without constables, to enter the premises at [REDACTED](herein referred to as “the Premises”) at any time within one month of the date of the warrant and to examine and test any television receiving equipment found there.
AND the said Chris Christophorou upon his oath says:
1. I am employed by Capita Business Services Ltd and apply for a Warrant under Section 366(1) of the Communications Act 2003 to enter and search the Premises and to examine and test any television receiving equipment found therein and I produce a copy of the authority of the British Broadcasting Corporation.
2. In a letter from “The Legal Occupier” dated 10th August 2010 the correspondent stated that they did not have a television set at his Premises and that as a result of a number of TV Licensing enquiries he was withdrawing the common law right for TV Licensing officers to visit the Premises: “This letter denotes prior written warning that such a visit will constitute trespass and harassment, and that I am withdrawing the implied right of access to my property…”. The correspondent added further that they were under no legal obligation to add a name and would not be doing so.
3. A response was sent from the TV Licensing (Customer Relations) Department on the 19th August 2010 acknowledging the request but pointing out that TV Licensing reserve the right to use other methods of enquiry to ascertain the licensing situation.
4. It was decided to seek permission to use television detector equipment to see if it would provide an indication that a television receiver was being used at the Premises. An authority under the Regulation of Investigatory Powers Act 2000 and the Regulation of Investigatory Powers (British Broadcasting Corporation) Order 2000 (sic) was sought and granted for the use of a detector van at the Premises.
5. A television display generates light at specific frequencies. Some of that light escapes through windows usually after being reflected from one or more walls in the room in which the television is situated. The optical detector in the detector van uses a large lens to collect that light and focus it on to an especially sensitive device, which converts fluctuating light signals into electrical signals, which can be electronically analysed. If a receiver is being used to watch broadcast programmes then a positive reading is returned. The device gives a confidence factor in percentage terms, which is determined by the strength of the signal received by the detection equipment and confirms whether or not the source of the signal is a “possible broadcast”.
6. On the 31st March 2011 at 18.36 hours the detector van was positioned near the Premises. When the detector camera was pointed at the window of the Premises a positive signal was received indicating a TV receiver was in use receiving a possible broadcast with a confidence factor of 97%. A television aerial and satellite dish was also seen affixed to the Premises.
7. A letter was subsequently sent to “The Legal Occupier” at the said Premises on 10th April 2011 advising that TV Licensing had recently obtained information providing us with reasonable grounds to believe that TV receiving equipment was being used on the Premises and advising that if a licence was not obtained then a search warrant will be applied for.
8. At no time during this period was there a record of a television licence in respect of the Premises. In the circumstances it is believed that television-receiving equipment is in use at the Premises. As stated the occupier has been sent a letter urging them to buy a licence and warning them that a search warrant would be applied for if they failed to do so. Television Licensing records reveal no trace of a current television licence held by any person authorising the use of television receiving equipment at the Premises.
9. From the information I have been provided with, there are reasonable grounds for believing that an offence under section 363(1) and (2) of the Communications Act 2003, is being or has been committed at the Premises, that evidence of the commission of the offence is to be found at the Premises and that it is necessary to apply to the Court for a Warrant under section 366(1) of the Communications Act 2003 to search the Premises and test any television receiving equipment found there.
10. The request for a search warrant is a last resort. All other reasonable methods of obtaining evidence have been exhausted.
11. The issue of a warrant in these circumstances is a necessary and proportionate measure. Entry will not be granted unless a warrant is produced.
TAKEN and SWORN this 28th June 2011
Before me the undersigned
A Justice of the Peace for Canterbury Magistrates Court.
______________________________________________________________
From the BBC's previous response to our FOIA requests we know all too well that TV Licensing employees are not adverse to criminal activities. Assuming Christophorou's evidence, particularly point 5 above, is an honest representation of how detector vans work then it's no wonder the BBC wants to keep their technology under wraps. In a nutshell they're employing bad science, possibly to baffle non-scientifically minded Magistrates into granting search warrants.
Are we meant to believe that their detection equipment can filter light from a TV set from all the other light sources in a property, despite it being reflected from several walls and refracted through several different media? It can do this when a lens is focused on a window, perhaps at an angle, perhaps with curtains and blinds fitted, perhaps with double glazing?
Handheld TV detection equipment?
Even if the technology does work, highly implausible as that seems, it can only detect a "possible broadcast" is being received. My understanding of English is such that even with a confidence factor of 100% that a "possible broadcast" is being received, that's no proof that an "actual broadcast" is.
I detect big time bullshit with 100% confidence factor.
Please forward our article to as many engineers and physicists as you know, because I'm sure they'll say the same thing.
Edit (10/9/12): The occupier of the property searched by TV Licensing, a gentleman by the name of Steve Heather, has had his day in court. He was found not guilty of TV licence evasion, which casts into even sharper focus the evidence presented to obtain this search warrant.
Edit (9/6/13): We have added some further information to the foot of Steve Heather's victory post. BBC documents have been released that confirm the authenticity of the deposition statement published above. The documents also confirm that TV Licensing (or rather Capita) refused to allow Chris Christophorou, the goon who laid this "detection evidence" on oath, to appear for cross-examination at Steve's trial. Now why might that be?!
Edit (10/9/12): The occupier of the property searched by TV Licensing, a gentleman by the name of Steve Heather, has had his day in court. He was found not guilty of TV licence evasion, which casts into even sharper focus the evidence presented to obtain this search warrant.
Edit (9/6/13): We have added some further information to the foot of Steve Heather's victory post. BBC documents have been released that confirm the authenticity of the deposition statement published above. The documents also confirm that TV Licensing (or rather Capita) refused to allow Chris Christophorou, the goon who laid this "detection evidence" on oath, to appear for cross-examination at Steve's trial. Now why might that be?!
They certainly seem to be baffling the magistrates with bullshit hahaha.
ReplyDeleteThere's a lot of it about.
My wife is always urging me to watch some program or other on iPlayer, but I cannot stand the drivel and propaganda the BBC produce, nor any other mainstream media source.
To me, it's all an arm of government. Their system doesn't seem to make room for the likes of me though.
;-)))
There is a way in which it might work.
ReplyDeleteAs it says in point 5, they are looking at the fluctuating light signal. So your detector filters the received light with narrowband notch filters centered on the red, green and blue frequencies used in color TVs. You compare the fluctuating pattern on each of the RGB components with the pattern actually being currently transmitted on each broadcast TV channel. By running a simple correlation function between the detected patterns and all the broadcast channels, you can determine which station is being watched. This would indeed provide a confidence figure, as stated in point 5.
The longer you leave the detector running, the more data you gather, and the more confidence you get that the detected channel is correct.
It's important to remember that TVs only emit red, green and blue. They cannot make any other colors (despite what our eyes and brains think!), so by notch-filtering on those specific colors they can filter out a lot of the background light that is on other frequencies. The problem then isn't quite as hard as it first appears.
However, I have no idea whether this system really works reliably in practice. It's one of those things that you have to try in order to get good data. It sounds very difficult to me, but it does not sound totally implausible.
Pat, what you speculate is indeed interesting. Whilst it would be possible to attune detection devices to the RGB wavelengths, I think it would be darn near impossible to have 'comparison software' running. Any possible comparison software would probably need to be pointed directly at the television, or certainly without any obstacles, to get a reliable comparison to 'a broadcast signal'. Given also the number of channels these days, the software would have to compare it to ALL currently broadcasting channels.
ReplyDeleteNormally their vans and equipment would be at an angle to the house, likely through curtains, possible deflections off walls. Unless they are packing a Sun supercomputer in the back of the van, that is a heck of a lot of 'correction' to be added into the assessment.
It is likely that they have equipment to dectect specifically the RGB frequencies - but that will NOT determine whether it is a broadcast signal or a DVD being watched. ie. They can be 97% or better that you are watching television-like equipment, but not necessarily a broadcast signal to that equipment.
I guess they could possibly marry the RGB detection with some sort of tuner detection, however that too would have its 'grey' areas, given that all TVs have built-in tuners (whether or not de-tuned) that would register as a positive tuner signal. ie. That positive would register even if watching a DVD on the equipment.
In theory, if they are relying solely or in greater part on detecting RGB light fluctuations, then the solution would be have the TV in the back room, should people actually want to watch the rubbish on broadcast TV. My advice would be to instead expand the DVD collection, much more enjoyable. Except for the regular threatograms from TV Licencing.
Well done on getting hold of the search warrant application!
I did say I didn't think it'd be easy. ;-)
ReplyDeleteLOL, Pat!
ReplyDeleteI found this rather intersting thread on a forum, with posts written by an apparently ex-EO.
On page 2, he talks about the detection equipment:
#6 Finally, I notice you never mentioned TV detection devices claimed to be used by TVL.... I know a lot of people believe these don't exist and are a myth created to scare people. Can you categorically state whether they actually do or don't exist?
ANSWER:
They do. Categorically. At least the technology does. In four years I never saw it used. And I was told that a handheld device was to be introduced (in 2007). I was once an electronics technician. In every analogue TV and radio there is a device called a local oscillator. This is used to reduce the broadcast carrier frequency to a more manageable level. Every old TV & radio is also a transmitter!!!. This was the signal used to detect if a TV was in use. With the advent of digital TV they now use a light spectrum analyser. Every TV broadcasts (shines) a very specific range of colours (frequencies) from the screen. The analyser can determine that a broadcast programme is being watched. Even from behind closed curtains.
http://www.tvlicenceresistance.info/forum/blogs/i-was-an-enforcement-officer/15/?PHPSESSID=84499f3f4b94ae73d0f8caa5a7e5cdbe
I know other forums and sites have discussed the local oscillator business.
(btw, I changed my screen name from "A Bandit" to "BringItOn". I felt like a bandit, even though I was not breaking the law!)
Very interesting!
ReplyDeleteOK, I've had a few more thoughts about this. Now I'm not suggesting this is actually how they're doing it; I'm just thinking through it as an interesting engineering problem.
If you're just looking at the overall R, G and B signatures over time i.e. the total R, G and B content of the picture, as opposed to the image content, then you have quite a low bandwidth signal.
Say we only take the frames (rather than the fields). We have 25 fps, so that's 25 samples for each colour. Let's say we quantise to 8 bits linear per sample (i.e. no compression, for simplicity).
Then per colour we have 25x8 bits per second. So our overall bandwidth for all three colours is about 600 bps.
That can easily be transmitted over even a 2G cellular link. So you grab the light signature from a house for (say) 5 minutes, and upload it to HQ along with a time stamp stating the exact time that the recording started.
At HQ they are storing a simple file which consists of the R, G and B signatures of all broadcast channels at all times.
When HQ receives the recording from out in the field, it runs a mathematical correlation against all the stored signatures in its database for that timestamp. Thus it calculates a confidence value for each channel. Given that the sample was obtained over several minutes, even a slightly elevated confidence on one channel is a good indicator that a TV show really is being watched.
This system would eliminate a lot of the complexity in the field. Their job (and tech) is very simple. The hard work is done at HQ, and even then it isn't that hard.
The big question here, as we've both been saying I think, is whether the real world is just too darn noisy for this to work at all in practice. I would love to see some experiments! Ah, if I were 30 years younger, I'd have the time to go and do the experiments myself. :-)
Unfortunately, it is indeed possible to reconstruct the image on screen - it's part of the TEMPEST family of attacks in information security terms. I've seen the original radio variant demonstrated live at Cambridge University, monitoring the display on a laptop using something resembling a TV aerial and some clever electronics; the same technique works with light as well with a high-speed photodiode. The few metres delay variation from multipathing aren't significant; 3m equating to just 10 nanoseconds of difference.
ReplyDeleteI'm no fan of the TV tax - I'd love to see the BBC switched to a subscription basis, and feel the government screwed up by not mandating conditional access during the digital switchover - but from a scientific perspective, I'm afraid the technique they describe is indeed perfectly viable. Fancy database about light correlation? They can look at the picture and go "oh, that's Coronation Street" then check Radio Times to confirm it's the current broadcast!
Thank you for your comment Jas. I must say I remain unconvinced. As it goes this case is heading to trial next week. If TV Licensing are confident of their evidence the trial will go ahead. If they're not so confident I would expect them to withdraw the case at the last moment. We'll see what happens.
ReplyDeleteConsidering the political minefield actually using that technology might entail, they might be better withdrawing rather than admit they use fairly recently declassified electronic espionage technology to see through the curtains and read your TV screen, even if they do have it!
ReplyDeleteThere's a thought: if they use this to process the TV signal ... that's the broadcast TV signal, for which you need a license. If they are using it, do they have a license for each and every "detector van"?
I would truly love the irony of seeing the telly-tax goons prosecuted for non-payment of their own protection racket, or privacy laws (is there a loophole big enough to permit snooping through the curtains at someone's TV/computer screen to view the contents without the owner's knowledge or consent, even if it is on the off-chance they might be committing a crime?)
Between the legal and political ramifications if they are snooping like this, withdrawing might well be their best bet. Having said that, since this is a prosecution, the defendant should have a right to see all the evidence beforehand, shouldn't he? Could it be too late for them to withdraw?!
The method claimed is known as Van Eck phreaking and is more effective with cathode-ray displays. Because of persistence of the vision the eye sees a full image which in actuality is being drawn by line by line - the raster. Pointing a camera in which the film is replaced by a photodiode, at an image on the target's screen, will recover much of the video signal. Recovering the sync information can be regenerated to yield the original video information.
ReplyDeleteWill it work in one of Capita's detection vans? How about the detector's equipment being swamped by the 100Hz flicker of sodium lights? How about curtains, domestic lighting and so on?
Assume that Capita claim they can develop Van Eck phreaking to work as they suggest, how can their detection equipment differentiate between a TV set displaying a live television programme and one displaying material from a VCR or a DVD? The so-called "light signature" is identical. Should they back-up their claim by adding detection of the local oscillator then again, a lot of VCRs don't have Scart sockets and will furnish their signal via UHF on (usually) Ch36. Not so much of a problem once the digital switchover is completed.
I built my own TV back in the immediate post-war years, many constructors watched the novel green-and-black picture displayed on converted WWII radar sets.
I asked someone about this. I don't know what qualifications he may have or even if he has any but I've seen examples of stuff he's done in the past and he clearly knows what he's talking about.
ReplyDeleteIn reference to para. 5 his reaction was
"Priceless B.S.
Whoever thought that up is probably in politics."
So there you are. I'm sure everyone can see the legal implications.
I just noticed this:
ReplyDelete"3. A response was sent from the TV Licensing (Customer Relations) Department on the 19th August 2010 acknowledging the request but pointing out that TV Licensing reserve the right to use other methods of enquiry to ascertain the licensing situation."
"....acknowledging the request..."?
What request?
As far as I can see the Householder issued an INSTRUCTION!!
I'd like to ask the tech. guys if I was watching a video of a football match on an analogue TV at the same time as a different match was being broadcast would this not confuse the detection equipment? If it works as is suggested.
ReplyDeletei.e. You have a green pitch and players in whatever colour kit so the light signals (or whatever) are going to be very similar.
@Hieronymous, two different football matches (or even the same football match timeshifted a few minutes apart) could easily be distinguished, as it is the colour fluctuations which are being correlated. Imagine a cut to a closeup of a player wearing red. The intensity of red would suddenly increase, only to fall precipitously when the scene cuts away. The same changes, at the same time, would not occur in the other match.
ReplyDeleteThe pattern of jumps in intensity whenever there is a cut between periods of continuous change will be very distinctive signature.
@A Bandit: The system would not need to compare the detected signal with every broadcast channel. It only needs to compare it with one - the channel being watched. In practice, I would expect that a few dozen channels between them account for 99% or more of all watched programs. Why should TVL care if 1% or fewer offenders get away?
I have no doubt that this system outlined by Pat Galea be made to work, and to work very well indeed.
Thank you, "Legally licence-free".
ReplyDeleteI'm not an engineer, technician or whatever but, I'm afraid, I remain unconvinced.
I shall research further.
Well, I asked another acquaintance who seems fairly knowledgeable in this respect and his response was "Loada crap!"
ReplyDeleteSo who knows?
Sure they would be able to tell what programme you're watching through direct line of site - ie with the curtains open looking through your window but to say that with curtains or blinds closed they can give 97% certainty that you are watching a live broadcast programme is absolute bull. I'm not a physicist just an averagely educated joe but common sense dictates that what they can't see can't convict you. Even if such technology were available and in regular use a decent solicitor would destroy it in a court of law. How would any electronic light/colour detector confirm with a required amount of accuracy that you are watching a live broadcast and not watching a dvd, computer screen/game or even coloured lights from a xmas tree or other coloured light based object. Then they laughably claim that they can even analyze the signals that are bounced off walls and furnishings in the room. Light (colour) is absorbed and/or reflected at different rates from different substances yet they can tell through windows with closed curtains what you're watching because there technology is so advanced it takes into account and accommodates for all different colours and types of furnishings in your room and the way that effects any signal......
ReplyDeleteIf I was a magistrate and they claimed a 97% certainty I'd want them to show me proof of their equipment working and when they couldn't then there would be only a 20% chance it was a live signal (based on the fact it could have been a Video game, dvd, laptop, PC or other device) and the search warrant would have been refused.
It can't be on RGB, as then it would have to account not only for everything thats been mentioned, but also for the make, model and individual settings of the television.
ReplyDeleteAny equipment they have may "prove" 97% you have a tv/monitor but it doesnt offer anything towards proof of watching live broadcasts.
Much in the same way that i'm 97% sure my neighbours have very very sharp knives that could be used to kill someone...but theres nothing to suggest that they are actually killing people - by any means.
The detector doesn't need to determine exactly what's on the screen, stand outside a house with a TV in it an you'll generally see a series of sharp brightness changes as the footage changes from scene to scene, by correlating that with live off air footage you can determine whether what is being viewed is likely to be a live transmission. You don't even need to worry about specific RGB fluctuations, simple overall brightness changes would be enough Given that the detector shows a % confidence figure than that seems most likely how it works.
ReplyDeleteAs a recent graduate with an MPhys in Radio Astronomy I would have to agree that the approach detailed here does sound indeed sound feasible.
ReplyDeleteMy concerns would be with regards to interference from other breath light signals, I.e. neighbours' TV's. NO lens (no matter how 'large') can completely block refraction from angles other than where it is pointed. Also, there is potential for other light sources to be reflected in the glass of a window.
- So how long would the detector wait around if your tv was not switched onwhen they arrived?
ReplyDelete- How would it detect any light from the back rooms or through blackout curtains?
- Are they sufficiently directional to pick up signals from tower blocks, or gated properties?
And this is just to obtain a search warrant? to search for what?
The offence is to watch live broadcast TV without a TV Licence, so unless they come crashing like the Spanish Inquisition while you're watching a live broadcast - what evidence can they produce?
From what I gather, the successful prosecutions are based on self incrimination i.e. confessions.
The reason they go to such lengths to perpetuate an anachronistic funding model is that they know that very few people would subscribe to their garbage voluntarily.
As much as it would be nice to have a non-commercial broadcaster I can think of no way to prevent it being hijacked by political movements, as has happened with the BBC (by the Fabians), so voluntary subscription has to be the solution.
I have seen commercial channels overseas where programs are broadcast "without commercial interruption"; with adverts before and after the program - which was fine, especially if the adverts were good (for example - the Cinzano ads with Joan Collins and Leonard Rossiter).
I'd just like to add a brief comment about Pat Galea's post above, "At HQ they are storing a simple file which consists of the R, G and B signatures of all broadcast channels at all times."
ReplyDeleteI have satellite receivers receiving most of the digital channels output from satellites at 13ºE, 19.2ºE and 28.2/5ºE. I have never bothered counting but estimate that allows my TV to display any one of in excess of 1,000 channels.
Are you seriously suggesting that 'HQ' (whatever that means) is storing every one of those at all times?
Pull the other one.
what if your watching a program via sky, which is encrypted . How can this rgb detection compare a live feed if the signals encrypted from sky. What if i paused the channel and they still trying to compare the live feed which for me is now 2 mins behind the live transmission, no comparison can be mind,
ReplyDeleteI'm not a tech guy either but I've had interesting conversations with my Dutch friends who work in hyrdoponics. They said that alot of the crims who used to grow weed switched from plasma lighting to LED due to the Dutch Drug police being able to detect the kind of light being used to grow weed from plasma bulbs. It had a very disticnt frequency. make of that as you will. TVL do say that they can't get a 100% confidence but very close to it, which in the eyes of the law is a high probability. Yet in the above case no guilty prosecution, and like mentioned before no detector evidence has ever been brought to court.
ReplyDeleteThat is not only possible, but pretty damn easy to do.
ReplyDeleteAll they need in the van is a computer to 'watch' TV and compare the average light level of the screen at each frame (possibly as three different levels (red, green and blue) for extra confidence) and compare that with the light fluctuation coming out of the windows, chuck in some VERY simple pattern matching software that any first year computer science student could write and you're good to go.
Even with room lights on there is still a fluctuation in the escaping light, but even if their equipment isn't as sensitive as they claim; all they have to do is park up in the evening and wait for the target to watch a movie in the dark or something.
In fact, even if the target had more than one screen showing different channels/similar, it'd still be easy to find the match, it would just give a lower confidence percentage.
Basically, all you need is a laptop with a webcam and a TV tuner, then you can step through the terrestrial channels one at at time until you get a match, they probably only use a van so they have somewhere to mount a sky dish too (and for somewhere warm and comfy to sit while they watch).
So, yes... that's very believable and your bibble about how ridiculous it is only makes you look like an idiot to anyone with any technical knowledge (or a mere grasp of common sense)
Funny Anon, I don't remember using the term "ridiculous". Dubious and implausible were the words chosen - words that apply to a considerable amount of "evidence" presented by TV Licensing.
ReplyDeleteThanks for dropping by and I hope you continuing reading closely.