Comment by rpdillon
13 hours ago
LG's statement:
> ACR uses audio fingerprinting technology using the TV’s internal audio processor (not a speaker) to identify content and does not collect screenshots, screen recordings, video recordings, voice recordings, or other audio recordings from the TV.
So they are claiming that ACR is done entirely through audio processing, and no visual data from the screen itself is used? That...doesn't seem plausible.
This paper from 2024 investigated both Samsung and LG and found that they both capture screen images for ACR, and send the resulting hash (not the raw content) back to the manufacturer for identification.
https://arxiv.org/html/2409.06203v1
We're going to have to get saavy in pulling apart the distinction between "I look at everything and then send a nice summary back to my manufacturer" and "I send everything back to my manufacturer". As AI gets cheaper and more efficient, it'll be baked into everything, and will distill signal at the edge and send the good bits back to central command. I fear the legalese in the ToS will obscure this new way of "not collecting" user information.
> So they are claiming that ACR is done entirely through audio processing, and no visual data from the screen itself is used? That...doesn't seem plausible.
It's probably easier to do and every bit as accurate as just using a screenshot.
And it's still exactly what I don't want them doing. There's literally zero reason for LG to be building an advertising profile on me because I was foolish enough to buy one of their TVs. This isn't something that makes their products better, it's spying.
Here's how it's been done in the past
https://www.cameronmacleod.com/blog/how-does-shazam-work
Audio is a bit easier to turn into a fingerprint for identification vs video. Video has a lot of smaller subtle changes that happen from things like compression which make it a lot harder to identify. It's why youtube still hasn't figured out piracy, but they'll knock you immediately if you play 5 seconds of copyrighted music.
To me, it's not even about why they're doing it (building an advertising profile). I don't want them doing it at all, for any reason.
When I buy something from the store, my relationship is between me and the store, for the 30 seconds it takes for me to pay for it. I don't want an ongoing relationship with the device manufacturer. I don't want to be tethered in any way to the manufacturer. I don't want to have an account with the manufacturer. I don't want the device sending anything to the manufacturer, advertising related, telemetry, or even a single bit "user has used your product." Do you get it, manufacturers?? I don't want any kind of relationship with you! I want to purchase my product and use it by myself not with you.
> I don't want an ongoing relationship with the device manufacturer.
I think it depends on the device. If I buy an iPhone, I have the expectation of installing apps on it after purchase, and that necessarily involves a relationship with the app store provider (ie. Apple), not the store I bought the phone from.
With a TV, I suppose it depends on whether I want to install media apps like Prime Video or Netflix on it. I understand that there will be people who don't want that, in which case I agree with you. But others want to be able to run this stuff somehow, whether it's directly on the TV or via some HDMI device like a Chromecast or a Fire Stick. At that point, they _do_ want a relationship with some hardware manufacturer so that the third party app can be delivered to run on the hardware.
5 replies →
Then what about all the cloud enabled spyware? How would that work? (/sarcasm)
Truly though, smart-home stuff runs the whole gamut of privacy and device manufacturer relationships. eg: I have used three different smart/room AC units:
- one requires its own cloud-enabled app for any remote connectivity, no HomeKit integration - one integrates with HomeKit but then leaves an upgrade hint that can only be done through an app with a login - one integrates with HomeKit and pretty much just works
HomeKit then allows me to remotely control via HomePod/AppleTV acting as a router ... of course this is just a different cloud connectivity but with on-premises devices controlling other on-premises devices.
For each manufacturer of smart devices, there is potentially a separate cloud where data is being funneled through. Some devices support multiple upstreams.
I get that Matter should be a way to fix it but the reality feels more like https://xkcd.com/927/
Part of me just wants dumb devices back. Get rid of buttons on a microwave (just give me a knob for time and maybe another knob for power setting.) Remove touchscreens from cars (mine glares at me with certain sun-angles.) Bring back the desktop/computer hutch and phones that don't live in your pocket. However, having the ability to start cooling my bedroom 45 minutes before I arrive home is ... pretty compelling too.
11 replies →
> There's literally zero reason for LG to be building an advertising profile on me because I was foolish enough to buy one of their TVs.
I feel like the world is bossed by people who read (well, ok listened to a 5 minute summary on a grindlife podcast at 1.5x speed) about the Trojan Horse and thought they smelled a business opporunity.
By and large, that has been my experience.
It’s a lot less data to process as well.
Is it within fair use to surreptitiously record copyrighted music?
No less than Shazam is?
> It's why youtube still hasn't figured out piracy, but they'll knock you immediately if you play 5 seconds of copyrighted music.
Unless the piracy is silent, surely audio fingerprinting would catch it?
As much as I detest ACR this seems more plausible than screenshots but I wouldn't trust LG regardless.
Audio is much easier to fingerprint and match against content.
wouldn't taking screenshots in fact not always be possible depending on the content you are watching? For example netflix with their hardware level drm? (Although its their TV they possibly have ways around that with some custom hardware)
LG sits at exactly the location where screenshots are possible.
Eventually, a digital signal needs to be converted into pixels on a screen. Once that happens, it's trivial to also pipe those pixels into a screenshot.
It's the reason DRM for digital media is pretty dumb IMO. The most sophisticated DRM still has to be decoded at some point for the end user to enjoy it. It can always be broken with a camcorder (though quality suffers). However, there exists devices which decode the HDMI signal so doing a pure digital signals.
I can assure you while out sailing the high seas I have watched many Netflix titles. So their DRM is literally only there to make the experience shittier for the honest customers.
lg’s display controller has to know what to send to panel
> So they are claiming that ACR is done entirely through audio processing, and no visual data from the screen itself is used? That...doesn't seem plausible.
Wrong. They claim it doesn't record video. Recording means saving permanently - live processing is not recording. ACR is processing the video output live.
Notice what they don't say at all. ACR is just one type of spying. A completely separate feature records what you say from the microphone. They only claimed ACR doesn't do that, which we knew. They did not claim the TV doesn't do that.
This is a carefully crafted PR statement, to avoid saying any lies while making it sound like the accusation was wrong, but it does not actually say the accusation was wrong. We can infer that the likely reason it doesn't say the accusation was wrong is that the accusation is right. Instead, it says several other things, which sound like they are refuting the accusation but are actually refuting other things that nobody said.
Exactly, otherwise LG legal would be threatening the authors to publish an apology or be sued.
If they aren't suing they are guilty as hell. Even if they were only most guilty they would sue; their legal dept knows they have no legs to stand on.
As usual, the key to explain the discrepancy between LG's claims and the paper 2024 is in the wording. Read carefully: LG claims it "does not collect screenshots, screen recordings, video recordings, voice recordings, or other audio recordings from the TV.". IOW they are making the technical distinction between "fingerprinting" and "screenshots or recordings".
And specifically, they are not confirming or denying that video fingerprinting are done.
What I think is likely is, as the paper states, that they fingerprint both audio and video. Upload both to their servers. But they end up using only the audio fingerprint for ACR. In this situation, neither LG's statement, nor the paper's statement, nor the GamerNexus folks's claims are incompatible with each other.
It even makes sense from a technical standpoint: it is far easier to uniquely identify content by audio than by video, as a fingerprint of a few milliseconds of audio is pretty darn ideal, compared to fingerprinting even a single video frame which has hundreds of ways it could be compressed depending on codecs, bitrate, resolution, etc.
> So they are claiming that ACR is done entirely through audio processing, and no visual data from the screen itself is used? That...doesn't seem plausible.
It apparently is.
Do you know the Nielsen Ratings? That's exactly how they know what people are watching these days. I've been contacted by them to be one of their subjects. They ship you a device that listens and recognizes based on content. They look for certain "fingerprints" in the audio and match against that - it's not capable of recording our conversations, etc (or so they claimed).
Is that how they do it lately?
I participated once or twice a good number of years ago, and it was kind of a pain: It involved keeping a paper ledger of what was watched, and when. Correspondence was all handled by regular mail.
(As I recall, they included a small amount of cash ~every time they sent more correspondence and this was a primary motivator to keep going with it.)
The Shazam app on my wife’s iPhone can identify TV shows and movies very quickly, sometimes nearly instantly. Not sure how it does that.
It is pretty neat and several resources online explaining it but it chunks the recoding and does a hash of the fft of each frame i think which reduces the data a lot. The server has a pre hashed database https://hackernoon.com/how-shazam-works-in-a-nutshell
1 reply →
I know a guy who made a Shazam-like phone app that can listen to a few seconds of noisy audio and make a fingerprint that can be used to quickly ID those specific few seconds out of a pre-fingerprinted archive of an enormous amount of audio (zillions of hours of TV). Making the fingerprint requires a tiny amount of very smart code (in C with no dependencies). But, the fingerprint is not useful for understanding audio that's not in the archive.
TV shows are completed and in the can well before air date. It is entirely plausible that the interested party in having the content identifiable by Shazam like services submit whatever needs to be submitted for new content
5 replies →
Signals analysis is pretty refined and not THAT compute intense.
This seems like a credulous argument based on no facts you have provided.
If you explained how the box was hooked up, it would be easier to judge whether there was a chance you were correct.
As it is, your description isn’t even clear whether or not the box contains a microphone.
Nielsen is global Company that is used for pricing ads on TV stations around the world. Their way of doing it from what I know personally is that they sign up people to put this ACR boxes voluntarily and then based on extrapolation of data - they can tell you pretty close what was the viewership across demographics. Similar to how election polls are done - you just need a good sample.
It's been done for at least good 15 years.
nielsen's ACR method is embedded into millions of devices and vendors. No need for a microphone, it's processing the audio signal being sent out. A microphone would be a wasted conversion..
(your car infotainment also likely has nielsen tech in it most likely, as they bought arbitron and gracenote)
1 reply →
LMGTFY was made for people like you.
You know you can look it up if you are skeptical, right?
Also, they didn't claim that.
It's how things were always done. The content from broadcasters has/had audio watermarks...frequencies which humans can't hear.
As far as I know, watermarks never took off because the technical complexity and lack of standardization. Nielsen researched and developed ACR technology but it never required watermarks.
When Nielsen first started collecting data (1950s), they handed out diaries to a group of around 5 000 selected households. Nielsen would then collect these diaries and extrapolate viewing patterns across the entire population.
After that came phone surveys. Next step (1980s) was the "people meter", a small box attached to the TV that could automatically detect which channel was tuned and which household members were in the room.
In the 1990s the set-top box made all of that redundant as the cable and satellite providers could directly see which channel every single box was tuned to at all times.
4 replies →
Regardless of it being how it always was done, such a finger print can easily be injected into LG's advertising ecosystem to infer everything about your demographic and what your interests and income might be and hit you with more specialized ads.
What we watch is not OK to be fed ads, especially if I wasn't using their service to being with and its monitoring the audio from any input to do this.
In this way, smart tvs and smart home devices are effectively hostile devices behind your router at home.
Yes. It’s also horrible that they have microphones in your home listening to your conversations. LG knowing what I’m watching is awful, but it seems even worse for LG to listen in to the conversation I have with my loved ones about it after. I don’t want to be under surveillance in my own home.
6 replies →
[dead]
I can’t speak to how LG does this now but I ran an agency that built Smart TV apps and built one of the first ACR powered apps on LG about 10 years ago. We were working with a company called “Cognitive Networks” which later was bought by Vizio and is now their Inscape product. At that time one use case wasn’t just tracking but actual interactive “sync to broadcast” apps (you can still do this, no one does, and I don’t know why)
At that time during testing Cognitive could literally see what we were playing on our TV (not full video but the frames used for fingerprinting)
They don’t use Cognitive anymore for obvious reasons and there are several audio based ACR solutions but they always had issues with sections of video without much audio to work off of so we’re always considered inferior. They are fine for usage data however.
Anyway my $0.02
Does it matter? Whether they are doing it on device or in the cloud, they are still recording what their customers say and collecting data on it without their consent.
To me, this is really an admission.
No, I think they’re saying two separate things:
- ACR uses audio fingerprinting
- ACR does not collect screen recordings
Both can be true if the TV hashes the frames locally and ACR “collects” the hashes.
They're almost certainly being pedantic about what it means to "collect a recording".
Or if they took all this time to respond because they were busy putting all the stuff they are denying in a second program they called "bcr"
This developer did actually find screenshots on their TV of everything displayed on their screen, even when they set every available option that's supposed to stop ACR:
https://www.reddit.com/r/LGOLED/s/x8iBNCPya5
Tom's response: "Sounds good, please don't take offense at me adding https://github.com/furkan-bayrak/lg-tv-blocklist to the PiHole. Can never be too safe, right?"
FWIW, LG will connect to a public DNS and use encryption so a basic PiHole list doesn't help unless you combine it with firewall rules. See the caveats section of that Github link.
Get saavy...? More like "will require trust that manufacturers are honest" -- which I won't believe for a miliesecond.
Yeah, I don't know what I was thinking when I wrote that. No one reads the ToS anyway. Just stop buying TVs, now that Sceptre stopped delivering the last models of consumer dumb TVs.
From reading other responses, it seems there is an assumption the image itself is fingerprinted. The digital content is encoded, there is essentially a barcode on your TV screen with lines running horizontally that you cannot see with the human eye. The encoding is picked up as the signal, it does not need to see or recognize if that is your favorite actress on the screen.
The thing which depresses me the most is all these so-called smart TV's are giant Arm computers that run malicious software. Imagine if we could re-flash your TV with Linux or whatever.
I can imagine that. We'd have a slow SBC with a fast video decoder, a huge screen that is inseparable and also now uncalibrated, some video inputs for other things, maybe a tuner if anyone can figure it out, and some kind of network connectivity.
I could probably come up with a use for that, but it's about as likely as coming up with a plan for one of the Raspberry Pi Zero Ws that are sitting in the drawer.
> That...doesn't seem plausible.
In what way? Dramatically lower processing and signature complexity. Audio signature matching is trivial and extremely well known. I see absolutely nothing implausible about this, and it seems a robust, viable solution.
> This paper from 2024 investigated both Samsung and LG and found that they both capture screen images for ACR, and send the resulting hash (not the raw content) back to the manufacturer for identification.
That paper found no such thing. That paper actually demonstrated that they have no clue how the ACR happened, they just demonstrated that playing content yielded content IDs. It is more likely they were witnessing audio ACR happening, exactly as LG described.
Well, not that Apple is doing it all their fans will claim it's ok
How they spy on people isn't the issue.
Spying on people without explicit consent (in a situation where you can say no without penalty) is the issue.
on-device transcript of audio recording is not audio recording ;)
The key problem is content identification, not how it is done. Seems like they are doing "look over there!" thing.
After doing some investigation into my Bang Olufsen after some tech not working right apparently Dev mode on it allows "Most probable raw grab: ThinQ/SSAP on the LAN after pairing — ~960×540 JPEG. Live Plus/ACR: most probably sends a fingerprint, not a retrievable raw frame."
As per HN thread last week
> ACR uses audio fingerprinting technology
so they aren't even denying listening in on everything?
They're not denying anything about programs other than ACR. That's telling. They didn't say "your TV doesn't listen" or "your TV doesn't collect"
"Everything" would include the microphone, both in the remote and the TV. They're claiming to split off the audio, with one path going to the speakers (and there to your ears), and the other path being hashed (and from there to LG).
They still claim to get your media's audio, but deny using the microphones.
If they don’t use the microphones why do they have a statement in their Terms of Service to not say sensitive things in front of your TV? It says the TV owner is "solely responsible" for obtaining all necessary consents from third parties whose voices may be recorded by the device, and must warn household members and guests that their voices may be recorded, and if someone does not agree, the TV's microphone and voice functions need to be turned off. [0]
[0] https://www.lg.com/us/terms
4 replies →
They deny ACR uses the microphones. Big difference. Nobody thought ACR used the microphones.
They explicitly state that the audio ACR is done through the "TV’s internal audio processor". e.g. rendered audio, not recorded audio. I mean, you could stretch the claim to have it internal include recorded audio, but that seems dubious.
Why are you giving a shady company with shady terms and shady history the benefit of the doubt?
5 replies →
Has that paper been replicated?
this is clearly an "anonymization" claim.
What is anonymous: we record all the hashs from all the programs people are watching to identify people watching the same thing.
What is not anonymous: we hash all the content in the world, along with hashing the content people are watching, and reconstruct what they watched.
> We're going to have to get saavy in pulling apart the distinction between
We probably shouldn't: doing any of this horseshit is a good reason to return a product. We should start doing that en mass until they get the idea.
[dead]