It may come as a shock to some, but TV used to be a big deal — a very big deal. Sitting down in front of the glowing tube for an evening’s entertainment was pretty much all one had to do after work, and while taking in this content was perhaps not that great for us, it was a goldmine for anyone with the ability to monetize it. And monetize it they did, “they” being the advertisers and marketers who saw the potential of the new medium as it ramped up in early 1950s America.
They faced a bit of a problem, though: proving to their customers exactly how many people they were reaching with their ads. The 1956 film below shows one attempt to answer that question with technology, rather than guesswork. The film features the “Poll-O-Meter System,” a mobile electronic tuning recorder built by the Calbest Electronics Company. Not a lot of technical detail is offered in the film, which appears aimed more at the advertising types, but from a shot of the Poll-O-Meter front panel (at 4:12) and a look at its comically outsized rooftop antenna (12:27), it seems safe to assume that it worked by receiving emissions from the TV set’s local oscillator, which would leak a signal from the TV antenna — perhaps similar to the approach used by the UK’s TV locator vans.
The Poll-O-Meter seems to have supported seven channels; even though there were twelve channels back in the day, licenses were rarely granted for stations on adjacent channels in a given market, so getting a hit on the “2-3” channel would have to be considered in the context of the local market. The Poll-O-Meter had a charming, homebrew look to it, right down to the hand-painted logos and panel lettering. Each channel had an electromechanical totalizing counter, plus a patch panel that looks like it could be used to connect different counters to different channels. There even appears to be a way to subtract counts from a channel, although why that would be necessary is unclear. The whole thing lived in the back of a 1954 VW van, and was driven around neighborhoods turning heads and gathering data about what channels were being watched “without enlisting aid or cooperation of … users.” Or, you know, their consent.
It was a different time, though, which is abundantly clear from watching this film, as well as the bonus ad for Westinghouse TVs at the end. The Poll-O-Meter seems a little silly now, but don’t judge 1956 too hard — after all, our world is regularly prowled by equally intrusive and consent-free Google Street View cars. Still, it’s an interesting glimpse into how one outfit tried to hang a price tag on the eyeballs that were silently taking in the “Vast Wasteland.”
All this talk of privacy and all I can think of is my next door neighbour who watches pornos on his 50” set in the living room which we can all see.
So… what’s the problem? You get free access to modern cinematography, why not join the fun? Better yet, take a couple of beers, knock on the door and have fun together.
it takes one to know one, or, like we say in Dutch, zoals de waard is vertrouwd hij zijn gasten.
Tell your neighbour they’er called blinds for a reason!
@ 5:36 in the video ?
So is there leakage from the RF mixer of the local oscillator which is being broadcast from the the antenna ?
And the frequency of the LO would have correspond to the channel that the antenna is tuned to ?
So it is basically all about leaking electromagnetic emanations, it is a form of TEMPEST.
For me there are credibility issues here.
Old CRT TV’s using valve technology shouldn’t have leaked back to the antenna and if they did then it wouldn’t likely be possible to detected The chosen channel.
The front end was like a tuned tank circuit that fed a valve in what wound be the equivalent of a emitter input common base configuration of a transistor and that has an exceptionally high input impedance. So high that I wouldn’t expect any leakage back to the antenna. Valve equipment had many resistors in the meg ohm or 100’s of k ohm ranges.
TV sets of the early 1950s were very leaky. VHF tuners usually had a pentode amplifier followed by a self-oscillating pentode mixer. Later sets used a triode in a grounded grid circuit as input stage, which greatly reduced oscillator leakage.
The BBC did something similar to verify people were not watching TV without paying their license fee.
And I remember reading that WW2 receivers used shielded RF amplifier stages (especially on ships) to prevent detection of the LO)
Wonder if it works like the sensormatic technology. That is two antennas so two trucks that can detect the oscillation just like the security tag doorway does in target. Its a tuning fork in the little pouches, granted made of thin sheets of stacked metal.
Related, the BBC would drive around looking for similar signals for people watching TV without a license to do so – which HaD also covered
https://hackaday.com/2021/01/18/tv-detector-vans-once-prowled-the-streets-of-england/
I was in TV repair in the early-mid 00’s and would come across sets wired for the Nielsen rating system, a very invasive box that monitored your channel and volume control (ad muting) habits. It was literally hacked into the sets.
As mentioned in the article above giving the same link.
Unlike today, when every cable, streaming and satellite box does the same and LG makes more money streaming pop up ads over your content than it does from TV sales. Try to find a non ‘smart’ tv these days. Not to mention every phone eavesdrops on you and serves targeted ads.
Yup, your smart TV and media boxes spy on you, even on things you’re playing off your local network or off USB sticks etc..
https://en.wikipedia.org/wiki/Automatic_content_recognition
Nielsen was a paid service. Nielsen paid you to use a modified TV so that they could collect the ratings.
Anybody with a Nielsen set knew they were being tracked, and they were paid for being tracked.
That’s a world of difference to today, where all the manufacturers regard their customers as lab rats to be observed in any way they feel.
I love TV but if it’s all you have to do you’re doing it wrong. As an aside, the author has an interesting view on when consent is needed.
If you own or have access to a spectrum analyzer, have a look at devices in your house… learn the LO they would use (10.7 MHz or 21.4 MHz are common), tune the radio to a frequency, and see if you can find the LO and how far away you can still pick it up. WIth a portable Spec A, you can find your clock radio LO from 100′ away (depends on model, etc). Clock radio? OK Boomer…
I could pick up the IF at 21 MHz from a neighbors TV so it was the same for any channel being viewed but I could hear the audio of what they were watching.
I’ve no details, but heard of an even simpler service to detect the stations that AM car radios were tuned to. They just placed a box near a street corner and when cars would stop for a light, they sniff out the leaking Local Oscillator frequency. The vast majority of radios used 455KHz IF, so If you if see a certain L.O. frequency, then you know it is offset by 455KHz from the station they are tuned to. Better yet, the customers drive their radios to you!
Even with limitations, knowing which car radios, in a certain part of town, were tuned to specific stations, had value to advertisers.
I remember hearing that some advertising installed in the mall parking lot boxes with a square antenna on the top and they were used to know which radio stations car radios were tuned on at that moment. This was around 2003-2005 in Florida.
One thing I found interesting is that the antenna on the van is not a wideband yagi like it initially looks but it is a series of simple dipoles properly sized and properly placed in front of the reflector. I am betting this gave them a fair amount of tuning before any amplification was involved. A far cry from today when you look at how SDR’s work.