Hardware Microphone Disconnect in Mac and iPad | Hacker News Reader

Hardware Microphone Disconnect in Mac and iPad

support.apple.com

771 pointsgtufano312 comments

It's weird how the "hardware disconnect" term for the pre-2019 devices was stretched to include a firmware, that is software-only, feature. Kinda like the "end to end encryption" of Zoom. Glad that Apple resolved this for newer devices. I'd still love to see wiring diagrams or explanations how it actually works. Which kind of circuit do they use? Is it the power line that's suppressed or the data line? Is it only an analog line that's suppressed and if you tune up internal amp you might be getting some residual?

About the iPads, what does their "hardware" based microphone disconnect entail? It has to be some electro-magnetic based communication instead of currents so the circuit has to be more complicated. I doubt it's done without using any kind of software but would be glad to hear otherwise.

Overall, I'm glad that they are responding to concerns and working to address them.

For analog, electret microphones a mute switch is very simple. Simply put an R-C pair (R a few MΩ, C 10-20 µF) in parallel to the microphone capsule. To mute it, short the R out. This is pop-free because the R charges the capacitor to the DC operating point of the capsule. When the R is shorted out, the microphone's output is shunted by the capacitor, which attenuates the signal level by >40-60 dB.

For digital microphones it can also be very simple. Consider a PDM microphone, if you disconnect the data line and have it stick to either high or low through a pull-up, the signal becomes DC. This wouldn't be pop-free by itself, but it should be rather easy to make it so through DSP.

(FWIW, you can also easily mute pro-audio microphones by shorting (+) and (-) together. The attenuation is determined by the output impedance of the microphone, usually a few hundred Ohms, divided by the resistance of your switch. Typically you'll get like 60 dB from this, which is perfectly fine for muting in a normal setup, but if you crank the preamp all the way up you can still get a usable, albeit very noisy, signal out of it.)

Shorting the mic signal is not a straightforward solution. For one thing, you can just amplify your signal right back up to intelligibility, since 40-60dB attenuation doesn't go under the noise floor of a well-designed system.

Just send the mic signal through an op-amp with an enable input.

Forward isolation is not necessarily much better than -60 dB depending on the exact op amp.

Most embedded microphones these days are digital MEMS microphones, using PDM. For those, if you gate off the digital signal, it's gone for good. I wouldn't be surprised if this is what Apple is doing.

This is how the Google Home Mini does it. When you flip the mute switch, an AND gate kills the digital signal from both microphones (a stereo microphone pair share a single data line). All the audio hardware gets is a string of 0 bits at that point.

> I'd still love to see wiring diagrams or explanations how it actually works. Which kind of circuit do they use?

I'd like to see that too, but I don't think we will. I suspect it's a magnetic switch and I'm also curious which lines it cuts.

FWIW, you can still get some audio from an accelerometer if you can get it to report data with a high enough frequency.

Not sure what you mean about iPhones, their information does not mention the iPhone (and I would not expect it to).

> Not sure what you mean about iPhones, their information does not mention the iPhone (and I would not expect it to).

I meant iPads but got confused. Thanks for pointing it out.

I’d think the raw pressure-level input from the force sensors on the trackpad would be even more accurate. It’s essentially an array of transducers already.

An accelerometer on a physical object is a force sensor. F=mA

I bet Louis Rossmann and other 3rd party apple repair services could figure it out.

they're probably just shorting the pvpvp3p33vp

... v3 hot

There seems to be a new broad trend of stretching feature names. Shopping for headphones, "noise cancelling" began to be used recently to mean "pretty good seal" and now I see "7.1 Surround Sound" on headphones with two drivers to mean "has some fancy audio processing."

It’s not new, sadly. The phrase “snake oil” goes back to the 19th century but names a phenomenon going back millennia.

The original snake oil, from the Chinese water snake, (Enhydris chinensis), worked.

Ingedients being hard to come by, and the market uninformed but spend-happy, hucksters appeared and the "medicine show‘ emerged. Patent medicines were also amongst the first widely advertised products in mass media beginning in the mid-19th century.

https://www.collectorsweekly.com/articles/how-snake-oil-got-...

http://gen.lib.rus.ec/scimag/?q=10.1007%2Fs12686-014-0200-1

http://www.npr.org/blogs/codeswitch/2013/08/26/215761377/a-h...

http://umm.edu/health/medical/altmed/supplement/omega3-fatty...

I'm actually looking for headphones now. How would I know what headphones have actual noise canceling features?

Try the Sony WHX line. Got a pair last fall and they are great.

Look for details of the audio processing in the tech specs. Even better, check a good reputable review site like headfi.

I miss the days when reference manuals included code listings and schematics.

https://archive.org/details/applerefjan78/

As a maybe 13 or 14 year old, the manual for the Osbourne2 Z80 CPM machine dad bought contained enough information for me to work out how to connect a broken Atari video game joystick to the parallel port, and how to read it by peeking memory from Basic...

Circuit diagrams, bank switched memory details, pretty sure someone dedicated could have built themselves their own machine with access to the manual and the right rom dump and system floppy disks...

My suspicion is there's a hardware method of detecting lid closed - probably detecting a physical magnet (which is used anyway so that lid closes nicely) and that drives a transistor which then does the gating.

I don't see why you assert that is has to be electro-magnetic based, as long as the gating is down by a 'dumb' fixed transistor rather than a programmable chip, then it's a hardware disconnect.

In the same spirit as putting tape over the webcam, I'd love it if someone found a solution for putting a tiny magnet in exactly the right spot to force-disable the microphone while the lid is open. :)

I hope you're right, and I'd agree that would qualify as a "hardware disconnect". It's a little confusing given the mention of the T2 chip; they do differentiate some specific models which are "hardware alone", and at least imply that the T2 chip does not flip the off-switch on those. I'll have to go dig into iFixit's tear-downs to see if they found anything. For the 2020 iPad, they specifically mention "MFI compliant cases", so we can be fairly certain that one's done with magnets.

> I'd love it if someone found a solution for putting a tiny magnet in exactly the right spot to force-disable the microphone while the lid is open. :)

It's the same sensor that's used to detect if the lid is open in general. Triggering it will prevent the display from turning on, and will make the computer go to sleep if it isn't running in "clamshell mode" (running on AC power with an external display attached).

Disappointing, but makes sense, thanks for sharing.

> For the 2020 iPad, they specifically mention "MFI compliant cases", so we can be fairly certain that one's done with magnets.

Why is that? Any case, including not "MFI", could put a magnet at the location if that was the case. Doesn't "MFI" mean that they make it only work with cases that are identified via the software to have paid their certification to Apple?

That's a good question, I honestly don't know. My guess would be that MFI is essentially a checklist of all supported magnetic functionality, and therefore the experience Apple can promise in its marketing (ie, in addition to collecting fees, they don't want a user to blame Apple when a non-approved case doesn't work). I'd wager that a non-MFI case with magnets in the right places would still work; I don't think constraining to pre-approved RFID codes or anything like that.

Interesting, it's magnet based. I've learned something today!

iPad Pro has multiple mics on different sides of the device, so they'd need to have multiple of this.

Talking about stretching terms - how about "zero knowledge" encryption marketed by cloud storage providers? Taking a cool sounding term from cryptography space and misapplying it to describe end to end encryption.

They rely on people with "zero knowledge" of encryption to select them as vendors

Considering firmware as part of the hardware from a higher-level view is not necessarily that weird in itself. It really depends on the details. I.e. a simple microcontroller with a program in ROM is also "firmware", but not all that different from a pure electronic ciruit. For more complex systems it becomes a more difficult question.

Also, why would detecting a closed iPad sleeve require "electro-magnet based communication"? Seems like detecting presence of a magnet would be enough.

As a former firmware developer (not for Apple, for power generation) it has firmware its not a hardware solution.

We have automatic steam cutoffs that are entirely mechanical. I have built a state machine out of hardware logic that would have cost less than a dollar with a microprocessor but was preferred because it could be verified to work and not be remotely “upgradeable”.

The computer free automation of a jukebox from the 1970s is an eye opener: (an in depth teardown and explanation) https://youtu.be/NmGaXEmfTIo

“ It has to be some electro-magnetic based communication instead of currents so the circuit has to be more complicated. I doubt it's done without using any kind of software but would be glad to hear otherwise.”

You can do the whole thing in analog fairly easily. The hardest part would be to get the magnetic sensor integrated into the same chip, but I guess that’s been a solved problem for a while.

I would still like to see Apple implement a Dell-style physical slider to close/block the webcam lens while the laptop is opened.

Despite Apple's anti-competitive ways, I am often impressed with their attention to details such as these. Glad there is still sanity out there in the world of "always listening" devices.

I wish people would qualify their “Apple Hate” every time they mention it. It would make it easier to know who’s opinions to discount.

An alarming number of people still believe Apple slowed down all their old phones for the sole purpose of selling them new ones for example.

They did absolutely slow down their older phones, but it comes down to whether you believe that their explanation of why was the whole truth. https://techcrunch.com/2020/03/02/apple-agrees-to-settlement...

I'm not an Apple hater or a fanboy. I've owned a number of Apple devices in the past. However, they have their issues just like any other tech company, but their devoted following does seem to be more cultish than that of, say, Microsoft or Google.

I know the intricacies of what they did. To summarise: some phones with older batteries would suffer brownouts because the internal resistance of the battery was too high. They updated the OS to detect the brownout and throttle the CPU to prevent it happening again.

So they didn’t slow down all old phones, only ones that had the problem. They actually attempted to fix older models. To me this is the opposite of planned obsolescence and appeals to my environmental view that all manufacturers should be supporting their hardware as least as long as Apple currently does.

Now they didn’t communicate, and were fined. I think that is fair enough, but I feel the size of the fine was quite excessive given they were trying to extend the life of their customers hardware.

Note that the other phone manufacturers that have exactly the same problems did nothing and are actually better off because of it.

Your "summary" curiously omits one of the most important details — That the battery could be replaced and the phone would perform like new. If not, the ever deteriorating battery would result in a never ending arms race of throttling. Inevitably, that phone and battery would be in the landfill instead of just the latter.

Whether or not the whole thing was intentionally nefarious I'm not convinced. But the episode looks way worse than your comment suggests.

Man - almost every other phone brand ends in the the trash MUCH MUCH sooner than Apple, so this is funny to read.

Seriously - can you tell me how long you received updates on your android phone?

Apple is the absolute leader in getting longer life out of their phones, they have much higher resale as a result as well.

Apple not only allows older devices to installer newer software (with bug fixes / security fixes / and supportable features) but they have been backporting stuff to a one step earlier iOS as well in terms of basic fixes. It's actually crazy especially in comparison to their competitors (that get no flak on HN) who literally ship with an old version of android AND DO NOT UPDATE IT!

The statement wasn't about how Apple's phones compare with other phones. It was (pretty clearly I thought) about how long Apple phones with new batteries last against those without new batteries (and throttling applied).

>Seriously - can you tell me how long you received updates on your android phone?

I've used Apple computers exclusively since about 2005 and phones since around 2009. So no, I've never owned an Android phone. In fact, my small apartment has no less than 8 Apple devices in use between my partner and I.

Why do you assume I'm not an Apple user? Because I said something mildly critical of them?

My understanding of the change they made was that it was expressly to deal with a condition with failing batteries where, as they were nearing the end of their life, they would just shut off abruptly when they still claimed to have 40% or more charge left -- the problem was that they could no longer deliver peak power, and if the phone tried to draw what was now too much power for something, it'd fail. The solution to that was to limit how much power the phone could draw.

I don't think it was intentionally nefarious -- this is a real problem with this kind of battery and I've experienced it, and it's not unique to Apple devices. In theory, their solution is actually pretty good! The problem was the way they communicated this to end users, namely, that they didn't. They just did it. And didn't even have iOS tell you it had enabled this "battery management" mode when it was turned on.

This wasn't an engineering fiasco, it was a PR fiasco. Apple has always been guilty of what we could diplomatically call "under-communicating," but this is the sort of change someone -- many someones, arguably -- internally should have flagged and said, "No, look, this isn't something we should just do silently, in part because it's going to create a suboptimal user experience and in part because if we don't communicate what we're doing and why we're doing it, it's going to come across as us just slowing old phones down to make you buy new ones."

A short-lived consumable that can't be field-replaced is always an engineering fiasco. I used to keep a spare battery in my backpack; an external USB battery is not a reliable substitute when (not if) the internal battery gets bad enough.

Are you suggesting it wasn’t known that battery replacements were available for iPhones? I know I personally replaced several batteries over the years in iPhones, I don’t think that was a big secret. There were shops you could drop your phone off at and have anew battery put in for fifty bucks in an hour overseas at least.

> There were shops you could drop your phone off at and have anew battery put in for fifty bucks in an hour overseas at least.

Actually I bought my iPhone 6S in Australia and had the battery replaced for free in Vienna under their battery replacement program. It was done in 45 minutes.

>Are you suggesting it wasn’t known that battery replacements were available for iPhones?

I'm asserting as fact that Apple retail and support employees were kept in the dark about throttling. So, even if you had Applecare, your "genius" would tell you that your phone wasn't slowing down (you were imagining it) or that the slow down was an inevitable result of ever more complex OS upgrades. The end result is that you were told nothing could be done.

So no one was told the simple truth. That poor performance was related to poor battery health and could be rectified by a simple battery replacement.

Are you sure about that? I thought it was pretty much common knowledge that your battery going bad would degrade performance and it was time to get a new battery.

What I believe mthoms is saying that the "battery management" software was the secret thing, because it was. We all have a basic idea that when your battery is going bad you should replace it, I imagine, but the software fix Apple silently pushed out arguably prolonged battery life at the expense of making the phone run slower. But if the Apple Geniuses weren't told this was happening, they wouldn't be able to say, "Oh, yeah, your phone's turned on battery management because your battery isn't doing well."

The phrasing you've chosen ("degraded performance") is purposefully vague in order to move the goal posts. Please don't do that.

We're specifically talking about lower CPU clocking. Now, are you claiming to have known that Apple was throttling Phones with aging batteries before the rest of the world? Because well, that's a pretty amazing feat.

Apple didn't inform users that battery replacements would make their sluggish phones fast again.

But they did, though? I got an email telling me my battery was part of a replacement program before this throttling "scandal" broke. I had it replaced for free in Vienna (phone was bought in Australia) in 45 minutes. I had noticed the throttling taking place but it wasn't until the email that I knew about the fault.

More accurately, Apple didn’t have a UI notification. This was not a secret and support people recommended it if you actually contacted them.

It's almost as if you get ideas to improve products as time goes on. No, that can't be it.

I didn’t leave it out on purpose. In my eyes it was a “limp home” mode so you would expect it to return to full performance. If it didn’t then that definitely would be a scandal.

Yes, they definitely should have communicated better. 100%. It was non obvious to any users that the phone had been throttled and that it needed a new battery. I think the fine they got will remind them to be more explicit in the future.

The feature still exists today, just explained better.

It wasn't non-obvious. It was covert. People had to root phones and observe CPU frequencies to uncover it. I see absolutely no indication that the intent was for the batteries to get replaced.

It's curious that this problem was introduced in the iPhone 6, but not in their earlier phones, wouldn't you say?

I'd assume the solution was introduced with the iPhone 6. The problem was there much earlier, due to the way Li-Po batteries work.

As such, it could even be counted as one of the touches Apple is known for that other OEMs don't bother with. It just needed better communication.

Can confirm. My 5s suffered the random shutdown issue and wasn't part of the throttling fix.

They introduced in direct response to people keeping their phones for longer--an more recent industry trend that has been well documented.

> Your "summary" curiously omits one of the most important details — That the battery could be replaced and the phone would perform like new.

I considered this omitted because it's obvious given the technical details surrounding the throttling to begin with.

Yeah, I had an old, slowed iPhone 5S and I had a third party shop replace the battery and when it can me back, it was significantly faster.

Supposedly, Apple didn't "implement" battery throttling on the iPhone 5S - Apple's discount battery replacement program only applied to the iPhone 6 and up. This suggests that Apple switched to a flawed battery technology in hardware starting in the iPhone 6 that wasn't realized until years later. And, instead of issuing a recall or repair program, they pushed a silent cover up in the software. Whatever the motive, they should have been more open about it.

I had an iPhone 5S that ran ok until I upgraded to iOS 13, IIRC. After that, the frames started skipping and it became frustrating to use.

You could ask any non-Apple industry personnel about what happens to batteries after years of constant usage and all of them will agree that eventually you eventually won't be able to pull the power you need to keep the phone running fast. Getting a new battery always fixed this, Apple just was performing this power management without telling the user before ios 12.1.

But if I'm an average user and my phone is slow, my first thought isn't "hey I should get the battery replaced". It's perfectly reasonable to assume the phone would still be slow, so why not just buy a newer model?

It's a lie of omission, which you could reasonably interpret as a trick to get people to upgrade.

If Apple had done nothing, then that same user's phone would randomly start switching off at non-0 battery levels (the un-throttled behavior) when brownouts occurred. Wouldn't the same user still come to the same conclusion, that their phone has some problem and needs to be replaced?

In general, the mobile phone industry has done a pretty good job encouraging their userbase to forget that their batteries are replaceable, so I don't think that most users would consider that as a remediation step off the top of their head.

I think it's plain that the best solution would be informing the user of the problem and letting them choose between behaviors (or, you know, replace the battery), but between slow usage and random shutdowns, I would personally choose the former.

I wonder if there have been any studies on how many users, once the toggle had been added, switched it from the (default) throttling behavior to the shutdown behavior? I'd be curious to see if I'm in the minority there.

> If Apple had done nothing, then that same user's phone would randomly start switching off at non-0 battery levels (the un-throttled behavior) when brownouts occurred. Wouldn't the same user still come to the same conclusion, that their phone has some problem and needs to be replaced?

No I actually think then the average person would assume it's a battery issue, or at least be able to google it. I'm not saying that this is actually better behavior (I would also choose slow mode over random shutoff mode), but it doesn't obfuscate the fact that the phone has a faulty battery.

The reason I suspect that Apple did this maliciously is that "service battery" notifications are pretty standard behavior (including on their own laptops). It just seems like an intentional omission.

If you're an average user and your phone is spontaneously rebooting, you probably think to replace it even more.

If the battery is dying at <5% I think it's pretty obvious. But I think the real solution is the throttling behavior PLUS a "service battery" notification.

> It's a lie of omission

I got an email well before the throttling "scandal" broke out telling me my battery could be replaced for free. It was, 16,000KMs away from home, for free.

What was it Apple lied about?

I didn't know about this email, but it still seems like a pretty bad way to communicate this no? How about a notification from the OS like in Apple's laptops?

I agree. They deffo dropped the ball on this one.

A big point of this is how user-hostile replacing a battery on a phone has become for flagship mobile devices, which makes this situation a bit worse IMO.

> their devoted following does seem to be more cultish

I don't know about that. I seem to see a lot of irrational Apple hate, but not that much irrational Apple love.

I think they are far from perfect in many, many ways. But so much criticism of them appears overblown, short-sighted, irrelevant to me, or just plain wrong.

I look at it like this: all things considered, balancing various concerns, what's the best option for me for my given needs? Following that philosophy I've ended up with a lot of Apple gear. Am I a fanboy because a better alternative doesn't exist?

Exactly this. In fact, I’d suggest it’s a massive circular argument. A lot of the vehement ‘defence’ of Apple I see is usually in response to some outlandish claim being made against them, which is often made in response to a gushing anecdote of how great Apple are. It has been ever thus.

> I wish people would qualify their “Apple Hate” every time they mention it.

Sure, let me do that: Apple is a corporation and it is stupid to believe that they care more about you than making profits.

This applies to any corporation. I've particularly noticed that the US citizens are increasingly buying into this propaganda that corporations can better protect their rights than a democratically elected government. And that is just sad and dangerous. We should be campaigning for our privacy rights directly with the government, and not just hope that some corporation will act benevolent towards us.

(And it is not a coincidence that this "news" came up in the social media feeds when the [news about exploits that granted unauthorised access to your camera through Safari is going around](https://www.ryanpickren.com/webcam-hacking-overview).)

The issue here is that instead of adding a physical switch to disable the microphone and camera, Apple is again asking its users to blindly trust it. We just have to take their word for it that it will work. They can always blame a bug in the firmware if somebody finds a way to exploit it tomorrow ...

(Note: I mentioned the Safari exploit only to point out that there is some negative news which obviously a corporation would like to bury with some positive publicity. It is commendable that Apple paid the discoverer a nice bounty and will be fixing it soon.)

Edit: And, ofcourse, the downvotes begin as the social media management team steps in.

I don't think caring about the user and making profits are mutually exclusive. A company that is interested in delivering what the user wants and needs will often make profit. It's not a case of 'with the user' or 'against the user'.

I think the OP is trying to say that there is no evidence that apple cares about the user, only its own image in the eyes of users.

There is always an antagonistic relationship between buyers and sellers. Buyers want more money for less features, Sellers want more features for less money.

There is no such thing as "with the user", unless the users are shareholders in the company.

I see opinions like this and then see corporations like Costco that you’d have to try hard to find a reason to hate. Ultimately you’re not going to align the interests of different groups perfectly, but profits and competion align value with consumer value we’ll most of the time. And I think it’s fair taking Apple at its word for now— you can bet a lot of tear down people want to test this.

Apple is easy to hate if you want stuff that you can trust to work over time. Anything that locks you in as much as they do is not a friend of mine (anymore).

They aren't the worst player, or even close really however. It's just there's a lot of power and influence in that company, so you hear a lot about it.

> I wish people would qualify their “Apple Hate” every time they mention it.

Apple hate? I'm a Mac convert, and I admire most of their hardware and design. But it's no secret they purposely refuse to interoperate with other vendors, and purposely develop proprietary technologies to keep you within their ecosystem. The recent acquisition, and subsequent planned shutdown of Dark Sky is just one recent example.

Part of it is just the business model, I guess. For example, Google's apps are widely available across a variety of devices, since it gives Google the juicy data that they're after. For Apple, which is not an ad company, the incentive might not be there for opening their apps up to a wider audience.

Since I cannot justify a $1k+ phone for myself and my family members (I am content with my $500 S10e), and I have a mix of Mac, Linux, and Windows devices, the Apple applications (however well designed) are sadly out.

Apple will always make the right choice—until it messes with their business model. There are uncountable examples of ways in which they have intentionally crippled their own hardware and software to prevent users from using their own purchase.

I wish we lived in a world where people only said things like that they believed.

I think we do. The problem is that people often believe things based on perception and snap judgement, which are often wrong. And once somebody has some "knowledge" even if gained through unreliable means, displacing that "knowledge" is rather difficult.

It's an artifact from evolving in a world where snap judgments were often the difference between life and death, and the person who assumed the rustle in bushes was just the wind, but it was actually a tiger, did not live to propagate. However, the paranoid person who thought it was a tiger every time the wind rustled the bushes, did survive to propagate.

Our current political system is pretty good evidence of people believing strongly enough in ideology to destroy the lives of others. I think they really do believe the stuff, and that's why it's so dangerous.

I don't know. I suspect a lot of people spreading fake news, conspiracy theories and crap like this know perfectly well it's rubbish.

Certainly some people do that, but I would imagine it's a very small subset that are either trolls or consequentialists who feel the end justifies the means who originate it. Most who spread it are useful idiots (I use the word idiot very lightly. It's easy for anyone who is human to get suckered by a fake news story here or there).

I have quite a few family and friends that unintentionally spread fake news because they really thought the story was real. One in particular is extremely hateful toward fake news but made a conclusion based on a headline that turned out wrong (we don't all have time to read every story we see, as much as we'd want to).

I don't hate them, but I find their hardware is typically not as good as the competition, and for a higher price. I also think most people that buy Apple devices don't know this, or just refuse to believe it.

For example, I can't tell you how many times I've heard "I tried a Moto g and it's horrible! My old iPhone was better", or "this $300 Dell laptop is way worse than my MacBook". Of course a $1000 phone is better than a $200 phone, and likewise for the laptops.

But when you start looking at high-end laptops like the surface series, or phones like oppo/Huawei, it's usually in favor of the non-apple.

People don’t buy specs, they buy experiences. If the Android has 12GB ram and some super fast processor, but runs in a GC’d runtime that stutters when collecting and unloads background apps with high frequency, it’s not actually better than whatever’s in the iPhone.

Plenty of videos out there comparing opening a bunch of apps in a cycle on iPhone vs Android. Android wins the first open, iPhone destroys on the second. (iPhone is able to keep all apps in suspense, whereas android unloads them)

Don't make the mistake of assuming 'experience' is synonymous with 'performance' either.

Side-by-side comparisons aren't really of any real-world significance. Very few people do that when forming their opinions. For the average person, performance either detracts from the experience, or it doesn't.

There are many other factors that are more important to experience than performance. Does the interface conflict with their expectations? Does the user find the features to be self-explanatory? Does the device enable the user to conform to social expectation?

> People don’t buy specs, they buy experiences.

More specifically, people don't buy actual experiences, they buy expectations of experience. 99%+ of people who buy a phone haven't had more than a couple of moments of experience with it. They buy it because they expect it to be good based on their first impressions and/or preconceived notions.

I think you will need to be more specific. I have never seen a difference in high-end phones with Apple versus Samsung or any of the other Android phones with the latest Snapdragon.

Is one plus 7 pro on latest snaP dragon? https://youtu.be/Ic8q1kPseVE

One plus isn't the same as Oppo. One plus is the cheaper brand from the same manufacturer.

It is an "Android phones with the latest Snapdragon", is it not? And it was outperformed by iPhone 11. These goalpoasts are sliding...

> iPhone is able to keep all apps in suspense, whereas android unloads them

Android also saves activity state for fast resume and keeps activities running until there is memory pressure. I've heard people complain about iOS devices in this regard because they do the same thing but don't have enough memory to do it effectively.

There is a lot of great hardware out there, no doubt. But the question isn't that simple. Another big component is the software. For the desktop, I find MacOS difficult to beat. I like and use Linux a lot, but there are commercial applications, I can't run on Linux. And I want to avoid Windows if possible.

And once you stepped in the Apple universe, you will find a large number of devices working together. The Mac, the phone, the tablet.

So there are many reasons to choose an Apple product, beyond its specs per dollar. Also, while I think that Apple prices are high, and upgrade prices even outrageously high, if I spec a high quality competitor coparatively, I often am not that far away from the Apple prices.

I don't agree with that. With the advent of WSL on Windows you can happily run Linux whenever you want and have all the normal ubuntu pieces available to you without starting a VM. You also get the windows application catalog, which is essentially every piece of major software ever written. With WSL now in the picture, I don't see how a BSD-based OS could beat it given that many applications that an apt install can give you aren't available as easily or at all there.

As I wrote, I don't like Windows much and are happier not having to use it. I still don't like the overly flat UI style and well, it is still Windows with all its oddities over the ages. I also get the impression, that Windows handles HiDPI way worse than MacOS, which is an important criterium.

WSL certainly has the potential to be a game changer and this has made Windows potentially interesting to me. What would really be a huge step, if Microsoft would built a Wayland server into the Windows UI. Having Linux GUI apps running on the native Windows UI with all acceleration, could make it a premier desktop for running Linux applications.

Well the BSD-based OS OP's talking about doesn't run loads of questionable telemetry (just judging by CPU usage) and does not turn on at night to install an update that will reset whatever telemetry settings the user hacked to disable.

Well, with Oppo and Huawei in particular, its pretty easy to beat the competition with state sponsored industrial espionage subsidizing their R&D costs.

Samsung, too.

> attention to details such as these

Don't a nontrivial amount of people use their macbooks closed in a vertical docking stand? This will require them to use an external microphone.

That said, my experience with similar laptops (XPS 15) has been that closing it just exacerbates the shitty thermal limitations so it's not really practical. Not sure whether this mode of usage is viable on newer macbooks.

Don't a nontrivial amount of people use their macbooks closed in a vertical docking stand? This will require them to use an external microphone.

Wouldn't they have to use an external microphone anyway, since in clamshell mode, the microphone is facing directly into the closed display? AFAIK, MacBooks of any flavor don't have external microphones.

I always wondered why the microphone on my MacBook Pro would stop working whenever I closed the lid when using my external monitor. Glad to hear the "bug" exists for good reasons.

What are users meant to do if they want to use their Mac's microphone while the laptop lid is closed?

Deal with it, frankly. A cheap USB mic attached to the external screen, or for that matter some screens (including fairly high end ones like that LG 5k display Apple was featuring for a while) have built-in mics themselves these days.

But this is one case where there really is a pretty fundamental trade off if the goal is a really simple core "when the lid is shut I'm guaranteed audio privacy from that vector". Anything that allows getting around that in turn is a potential bypass. Anything in software, even if it requires a special boot to access, leaves some room for bugs. And for both software or even a physical hardware switch there would be room for someone to accidentally or maliciously leave it on, and it would be very hard to notice. Given how cheap microphones are and that if someone is using the system with the lid closed they're already committing to a certain amount of much more expensive external hardware, I think it's quite reasonable to stick a straightforward visual heuristic that "if the system is closed and thus the camera cannot see the microphone cannot work either". Easy to verify at a glance for anyone, not just the owner, easy to remember. Like security, good privacy design requires not just a solid technical foundation, but considerations for human UX.

The microphone is located in the bottom area of the speaker grill. It'd work terribly with the lid closed anyway.

I remember years ago visiting a customer shortly after many of their engineering staff had just upgraded to whatever macbook was popular at a time.

Within the week that they had them, two people in the department had broken their screens/hinges from accidents walking in and out of meetings with the lids open because there was no way provided in the software to disable suspend on lid closed which was killing people's SSH sessions. Within the month someone in my office also managed to do the same, and most of us weren't even using macs.

(I understand that this was eventually workaroundable with some power users tools; and I imagine mosh makes the suspends a little less of an issue now).

I'm sad that the popular linux desktops later decided to emulate the bad software culture that brought anti-features like that mandatory suspend.

So I expect that mic off on hinge close will have similar results. Though, ... at least this seems a lot more legitimate to me than a refusal to not suspend.

It's a trade-off. The behavior is pretty easy to grasp for the average user (closed laptop: stop doing everything, enter low-power mode), and there's ways around it for users that want something different (caffeinate command, ssh to tmux/screen sessions, etc).

A macbook works perfectly in clamshell mode. Just provide power and have a display connected. (not sure if you have to have a display connected, but it's how I'm typing this comment right now)

Also, using tmux (or screen) on the other end of the ssh connection helps. And not only for when you close the lid, but also if the connection gets interrupted (power failure, internet failure, routing failure, whatever)

> A macbook works perfectly in clamshell mode. Just provide power and have a display connected.

I can't imagine how much more equipment would be damaged if the people described were carrying around a power source and a monitor with them so they could close their laptop and bring it to a meeting without dropping their sessions.

There are fake display adapters specifically to combat stupid problems with running mac displayless. Maybe we need a fake power source as well...

The vast majority of users expect the microphone and camera to be "off" when they close their laptops while it is sitting on their bedside table. Other users can go buy a product that doesn't take that function very seriously.

The vast majority of users expect the microphone and camera to be "off" when they are not using them regardless of the position of the lid.

Get an external mic. It's a niche enough use case that they don't need to cater for it.

I open it a little bit, just enough for the mic to work, in these cases. Then close the lid completely when im done with the call

- Use headphones that have a mic

- Use a webcam with an inbuilt mic (fairly common)

- Use an dedicated mic

Plenty of options here.

I had to resort to a VM (Parallels) because the external mic I bought does not get recognized by PyCharm - I think it's a permissions issue of some sort. Either way, it's very annoying.

I think having different sources for sound and mic causes echo.

The microphone would probably sound pretty bad when obstructed... I've heard that complaint from a few ThinkPad-owners since everybody is doing homeoffice.

Thinkpad microphones are bad regardless of being obstructed. Almost all (I'm sure there's an exception somewhere) microphones built into laptops are pretty bad, and at least half of that is due to "inside the laptop" is not a good location for a microphone.

They can use the loudspeakers as microphone. Usenix 2017: Turn Speakers to Microphones for Fun and Profit

https://www.usenix.org/system/files/conference/woot17/woot17...

Yet another great reason to always play speed metal really loud :-)

What are users meant to do if they want to use their Mac's microphone while the laptop lid is closed?

USB microphones are like $3. You'd use one anyway if the lid was closed because the microphone would be pressed against the display.

I'm digressing but I've always been curious why people want to use the closed-lid mode. All it does seems to be saving a some desktop space. I place my laptop beneath an external monitor, this way I get a very good second display, a nice keyboard, and a very nice extra touchpad for scrolling and all kinds of multitouch gestures. Probably has benefits cooling the machine as well.

In my office I run lid closed because I have 3 nice large monitors and I don't really have room for a fourth. It's nicer to have three displays that match in size and resolution than 2+1 that doesn't.

I would rather have a physical (manual) switch to disconnect the power to the microphone and camera hardware on all my devices (laptop, phone, ...). This way there would be no need to trust the Security Chip and there would be no attack vectors and possible future zero-day attacks on the security chip...

That’s fair enough but you have to accept you’re in the huge minority there. The average consumer does not want to have to flip a switch for every microphone on their devices when they want to use it. Not to mention the reliability implications of adding more moving parts and dust/water ingress points. There has to be a trade off between usability and security.

My teacher wife and a bunch of the folks she knows (non-nerds) cover their webcams with stickers or whatever. They’d all love (separate!) kill-switches for their cameras and mics. Plus everyone working in tech I know. So that’s... everyone I know who’s aware their laptops have cameras and microphones.

The problem is, as soon as security expert "Barbara from Facebook" posts something about the NSA still watching you because the switches don't really do anything you're back at square one. :)

Not for mics, but you can essentially add the equivalent of a webcam kill switch with something simple like this: https://www.amazon.com/C-Slide-Sliding-Computers-Chromebooks...

My company's IT department has said in passing that they've had many cases of laptop screens being cracked by those sliding webcam obscurers.

I think you're overstating your position.

If all of those people had to flip a hardware switch every time they wanted to use their phones, they'd complain about it.

Not phones, but for tablets and laptops I think a lot more than just privacy geeks would like a way to know for sure they won’t accidentally transmit video or audio, and would put up with a little annoyance for the peace of mind.

On a laptop it may often be once in a lifetime.

Depends on your laptop. My Mac laptops ring when someone calls my phone in another room.

So, the device comes with a slide switch, set to on. A user who cares can slide it off.

It's not that simple. People will accidentally set the switch to off and won't understand why their mic doesn't work anymore. At this point there is a good chance they will call support.

This also happend with the hardware wifi kill switches we used to have on notebooks.

God the number of times people complained about their laptops being broken only to find out they'd turned WiFi off. Either through a separate kill switch or through a function key. It was a pain the arse.

I recently installed linux on a MSI laptop that has a webcam function key switch. For some reason (could have been me), it was off and I just assumed that the camera wasn't working with linux, or that there was a hardware issue. I almost sent it back. I never owned a laptop with that kind of switch before, so I wasn't looking for this, and the keyboard icon [1] looked like some kind of screen contrast button to me...

[1] https://imgur.com/a/2Oe2k4e

Oh man, this happened to me many years ago as well with wifi. The keyboard had a wifi toggle (which in linux translated into rfkill) as a mod to an Fn key. Took me hours to realize that was a thing. I was literally debugging the kernel trying to figure out why the driver was spitting out "rfkill enabled" when "it shouldn't" only to realize a simple hold of Fn + F9 (or whatever it was) fixed the problem right up :facepalm:

That's a little different. Wifi is an invisible feature, and it can become disabled through no action of yours. Cameras are physical, so you can make it really obvious when it's enabled and when it's not. You just look for the physical obstruction. Apple's first iSight camera, for example, had a cool leaf shutter that you opened or closed by twisting it.

I know lots of people who have had trouble getting their wifi to work. I don't know anyone who put a piece of tape across their camera and then couldn't figure out why their camera didn't work.

It is that simple. The computer can prompt "MUTE" just like TVs do

Since we’re integrated between OS and hardware on Apple platforms, they could implement a “Turn on your microphone” dialog. I think this support excuse is too commonly used, when there is obvious value in the result. He’s not a minority, even my grandparents shared the photo of Mark Zuckerberg saying “If this guy hides his camera, so should you”...

You already have dialogs on macOS like "app X wants to use the microphone, allow?"

It’s Apple. If you want switches on your computers, use ThinkPad or Dell.

The latest ThinkPads have a physical/mechanical switch, but just for the camera... The mic can not be turned off with this switch...

Just a warning (I don’t know the specs of that ThinkPad) but just because it’s a physical switch doesn’t mean it actually physically disconnects the device.

It's called ThinkShutter and it physically covers the camera lens.

so it doesn't turn it off, it just puts some plastic in front of it. Plastic that is easier to break then something inside the case that will cut the power. While I am a big fan of the idea of the thinkshutter and friends, a real switch that cuts the loop to the cam/mic is a better option.

I think a physical shutter is superior, if only for peace of mind. If the lens is covered, nothing else matters.

Even the user that cares can sometimes forget to do it. With Apple solution you'll never forget to switch microphone off, but on the downside, you now have to trust Apple.

After months of zooms and many #PoorJennifer moments I'm sure that the number of us with camera covers will increase

Anecdotally even my old 69yo father who thinks Google Chrome is the same as Google.com has covered his webcam with tape and bought a laptop with a mic hardware switch so I really don't see this as a usability problem in 2020.

What do you mean with "flip a switch for every microphone on their devices"? Why would you want to have more than one mic and one switch per device? And turning it on in hardware is way faster than in software so a hardware switch is both faster and more secure and breaking a hardware switch seems a lot less likely than custom OEM made software failing or stop getting updated. Dust and water is no problem as lots of phones already have hardware switches, including my Android phone. I really see no problem in hardware switches at all.

> What do you mean with "flip a switch for every microphone on their devices"? Why would you want to have more than one mic and one switch per device

The ipad has 4 microphones. iPhones have at least two. They’re located in different places to either catch different sounds or implement noise canceling

Clearly anyone sensible would have a single switch for all of them.

Would turning on a single microphone of the array have any meaning?

Since iPhone microphone quality is well below market average maybe it would improve?

Once such compromise is PilferShush Jammer for Android/Windows. There might be a similar iOS version, but I haven't found one yet.

Android. Also on Play Store. https://github.com/kaputnikGo/PilferShushJammer

Windows 10 app. https://appforwin10.com/pilfershush-jammer-app-for-windows-1...

Actually, I have not though about this before, but now I want it.

It totally makes sense, and in the modern world too many people put covers on their cams for this not to be considered a valuable feature.

There used to be such flip switches for Wi-Fi, so it is not too far-fetched.

I doubt this is the case.

Perhaps Apple aren’t doing this so much for you-the-individual-consumer, as for government buyers who don’t want devices that might, if root-kitted by some foreign government, become listening devices even when nominally “off.” Hardware interlocks like this ensure that, at least, the device can only be recording audio while in use, while making it evident by their physical state (lid closed), even from across the room, that they can now no longer possibly be recording. That makes them more interesting to such buyers.

(But if Apple cares about this market, wouldn’t you expect their laptops to have a lot more security features to appease these buyers? Not necessarily. This sort of thing is pretty much “enough”—governments are used to having signals intercepted by foreign intelligence, and so they have mitigations in place, like shutting off all electronic devices before having important conversations, or creating secure rooms known to not leak emissions, and then requiring that nobody bring anything electronic into them. For non-classified but “off-the-record” conversations, the ability to say that there’s no known hardware in the room actively recording the conversation is usually “good enough.”)

I can understand your preference, but the prevailing aesthetic seems to be going in the direction of completely removing physical controls. To be perfectly frank, seeing earlier-than-current iPads and iPhones with physical home buttons strikes me as kind of quaint. I greatly prefer the entirely software-defined interface of the latest devices.

I suppose I can trace that preference all the way back to being fascinated with the Star Trek: The Next Generation LCARS (Library Computer Access & Retrieval System), which (to quote the Star Trek: The Next Generation Technical Manual written by Mike Okuda and Rick Sternbach from teenage memory) was “expressly designed to suggest a highly simplified manner of managing vastly complicated processes” or some such.

If you really want to be sure it isn’t listening to you you can turn it off and/or leave it in another room.

I think it makes a lot sense to have it on devices like the Nest Mini: https://www.theverge.com/circuitbreaker/2019/8/23/20828854/g...

I am not sure if I'd want my phone to have another button just for that though.

Then you should consider Purism Librem laptops and phone.

Based on the linked page this functionality is separate from the security chip and a compromise of the security chip cannot disable it.

My laptop actually has this switch for the camera. It slides a physical cover it behind the glass and tells the software it is disconnected. So even if the software was compromised it is physically unable to see anything. Wish there was an equivalent for the mic

A manual switch is still an attack vector, and I'd bet one of the lightswitches in your home has already been "compromised" where you have to wiggle it a bit to activate it.

> A manual switch is still an attack vector

Like remote targeted attack would force the roomba to sneak up on someone and bump the switch on? Otherwise, i don't see how this is a remote attack.

A manual switch is still an attack vector

So we should chuck all of our electronic devices into the bin in the name of security?

Good security is sometimes better than perfect security.

> All Mac portables with the Apple T2 Security Chip feature a hardware disconnect that ensures the microphone is disabled whenever the lid is closed.

vs.

> On the 13-inch MacBook Pro and MacBook Air computers with the T2 chip, and on the 15-inch MacBook Pro portables from 2019 or later, this disconnect is implemented in hardware alone.

Do these statements not contradict each other for the 15" 2018 MacBook Pro, for example, which includes a T2 chip? This would also contradict earlier documentation provided on the T2 chip by Apple themselves [1].

From [1]:

> All Mac portables with the Apple T2 Security Chip feature a hardware disconnect that ensures that the microphone is disabled whenever the lid is closed. This disconnect is implemented in hardware alone, and therefore prevents any software, even with root or kernel privileges in macOS, and even the software on the T2 chip, from engaging the microphone when the lid is closed.

[1] https://www.apple.com/euro/mac/shared/docs/Apple_T2_Security... (October 2018, page 13)

I don't think they contradict each other. The way I read it is, the former, have hardware disconnects controlled by some software/firmware, such as a relay/MOSFET or something of that kind; an electronic switch.

The latter I read as being hardware _only_; "only" being the key addition to this sentence. I would expect this implementation to be something like a reed switch to magnetically disconnect the lines _physically_ rather than electronically.

Isn't the whole point of software to control hardware, at some level? How is hardware-controlled-by-software different from plain old software-controlled? If a switch can be closed by software, I'm having trouble putting my finger on exactly what security benefit that might offer.

> How is hardware-controlled-by-software different from plain old software-controlled?

Perhaps it’s that the hardware interlock on the microphone can be enabled by software, but can only be disabled by a physical action (i.e. opening the lid.)

The security benefit here is that software has no access to this hardware switch.

I don't think the distinction is a FET vs a reed switch -- the means of blocking electrons, rather, it's in what decides whether that switch is open or closed. I would consider a circuit like this driving the FET/relay/etc. to be a "hardware disconnect" (using HDL to describe the circuit, not suggesting it should be programmable logic):

  module mic_enable (lid_closed, lots, of, signals, mic_enable);
    input lid_closed, lots, of, signals;
    output mic_enable;
    assign mic_enable = !lid_closed & lots & of & other & signals;
  endmodule

I think that there are two aspects to this problem of hardware disconnects: can we really sure that the disconnect actually cuts the microphone and then can we be sure that this disconnect is driven reliably to cut the mic when we want it to be cut.

Having a separate FET/reed switch is about the former: having a discrete component makes it easier to audit and make sure that the microphone is indeed cut off. Technically messing with the audio codec config or pin muxing is probably equally as efficient when it comes to muting the audio input, but that's a lot more difficult to audit.

But then all all that is pointless if the code driving the switch is broken or has a backdoor. Given that most people won't disassemble their phone or laptop to put a probe on the FET/reed/whatever driving signal I feel like this is just a marketing smoke screen.

If I have to trust Apple's firmware to drive the "hardware" switch reliably, why not make things simple and trust Apple's OS to mute the audio codec reliably?

I think the whole point is that there is no code (firmware or otherwise) driving the switch. I guess it depends on your threat model, but if you don't trust apple's hardware or firmware to disconnect the microphone, you can't trust their hardware not to have another microphone somewhere else that isn't advertised and is on all the time.

I'd take the term "hardware disconnect" in this sense to mean that there exists no program that you can run on any of the processors, or no bitstream you could load into any FPGA on the device that would be able to enable the microphone when it shouldn't be enabled, eliminating the threat of malicious code enabling the microphone

Apple in 2020 is such a weird thing. Some of the things they produce - mostly in the software realm, I guess - have really careless and egregious bugs and design problems. But then they still come out and make amazing little features like this one that nobody thought to ask for, and don't get highly publicized, but really speak to that classic detail-oriented Apple mindset.

It's like there's a huge cultural schism running through the middle of the company, or something.

> they still come out and make amazing little features like this one that nobody thought to ask for

Privacy/security-conscious computer users have been asking for user-controlled hardware switches on such peripherals from the moment they were shipped integrated into the device.

> Privacy/security-conscious computer users

So, in the grand scheme of things, "nobody" is an accurate description. Most people don't care about a hardware disconnect for their microphone, even if they should.

Right, and Apple isn't a niche hyper-secure laptop company, they just (sometimes) put good features in their mainstream products that their customers may not even know they want.

Is it too much to ask for a small LED next to each camera indicating if the sensor is powered? Should be standard on every phone.

Ditto. Or even an old-school LED that indicates data transmission like the HDD LED in old computer cases. Make it in hardware so that if any data is sent through the wire, it lights up.

HDD LED in old computer cases

Old? I'll admit I haven't changed my case in a couple years, but mine currently has one, and I can't remember the last time I've used a desktop computer that didn't have one.

old, adj: Having lived or existed for a relatively long time;

https://www.thefreedictionary.com/old

Old doesn't have to mean extinct :)

Unfortunately those were immediately susceptible to tempest attacks, at least at old hard drive speeds.

I think you are confusing this with data tx/rx lights for exfil[0]? I've never heard of a tempest attack (that is, revealing the data written via a side channel) on a hard drive light.

[0] Guri et al, LED-it-GO

https://arxiv.org/pdf/1702.06715.pdf

I think it’s this paper I remember that describes it as an “optical tempest” and talks about storage mechanism leakage: http://applied-math.org/acm_optical_tempest.pdf

Yeah it's type II in the Loughry schema. Still unfortunate if you're operating a high security air gapped network, but not squirting out secrets without being compromised first.

It seems almost quaint to worry about this when so much software has web spyware baked into it -- but maybe we need a virtual spyware led that lights up every time something talks to Facebook/Google/etc tracking. The light would be blinking non-stop for many people.

It would just never go out.

People already complain about the price and battery life of current phones. I can’t imagine either of those things making either situation better.

You can get SMD 0603 LEDs that use ~1mA of power for a few pennies in bulk. Peanuts compared to the power usage and price of the rest of the phone.

yeah that LED cost is going to drive my 1400 phone price up through the roof

Or maybe an icon in the status bar indicating the camera being in use.

Having it in the status bar would defeat the purpose as ill behaving actors could possibly circumvent this, having it in the hardware as a led would make this if not impossible most likely infeasible. I’ve only heard one instance where someone were able to go around the led and if I recall correctly that was on the old PowerBook and required physical access to flash the firmware on the camera hardware module with custom malicious firmware.

You can get that functionality with Microsnitch. It works for the microphone too.

I use MicroSnitch on my mac and love it. I just need it for the iPhone. It doesn't exist, does it?

No but it sounds like a good idea. I suspect it would require a jailbreak though.

I never understood how an expensive device like a camera was added to literally every laptop in existence without any option to remove (save for a few). I can only think manufacturers were pressured.

Supply chain/production complexity. It's more expensive for them to make the option customizable than it is to ship it for everyone. If there was an appreciable market for consumer-level camera-less machines the math to justify it might look different but there isn't. The vast VAST majority of the demand for hardware without cameras comes from the NatSec type people, and they're all doing bulk orders so they get more leverage to have niche production runs done for them. And even then they don't always get devices without the cameras. They end up having to do awkward things like physically disabling the hardware after the fact or locking out the functionality with custom firmware.

Before laptop makers started adding them en masse, there very much was an appreciable market for camera-less machines.

I find that pretty far fetched. I'd need to see some actual data on that that isn't explained by the commodity parts getting cheaper to be convinced.

Because before laptops included cameras we used to buy Logitech USB cameras to attach to our laptops.

I still have mine, and sadly it only produces 640x480 images, and everything looks too dark compared to the camera that is now included in the laptop.

It's a company with multiple people making multiple decisions each day. It shouldn't be surprising that there's variation in quality, just like at any other company.

But what I see is a gargantuan drop-off. A much wider range of variation than you'd normally expect, and mostly partitioned to different segments of the business.

Painting with broad strokes, Microsoft's products are fairly consistently "fine". Lately perhaps they've skewed towards "pretty good". But generally you know what you're going to get. Whereas Apple's products can be anywhere from unparalleled to embarrassing.

I've always attributed that user-tradition as a Jobs byproduct; he was a unifying force for the user experience and that role hasn't been filled since his passing. There are lots of hands involved with many projects, but it feels there's no point-person or champion on this for many years.

From going back to dongle hell of the 90's via USB-C, to the conflicting marketization of the macOS products, the connectors proliferating on the iOS standards(now lightning and USB C!), the conflating of iOS apps into macOS space, dropping magsafe; it doesn't feel very user friendly when you start getting into the details.

> iPad models beginning in 2020 also feature the hardware microphone disconnect. When an MFI compliant case (including those sold by Apple) is attached to the iPad and closed, the microphone is disconnected in hardware

does this mean Hey Siri won't work on an iPad with a closed cover?

https://forums.imore.com/ask-question/394180-hey-siri-featur...

That is precisely the behaviour I see on my iPad Pro from 2018

Doesn't MFI use software to do verification? If the presence of a MFI software command is what disables and enables the microphone, that sure sounds like software control.

Sure, maybe its software on a peripheral that controls it. But it's software never the less.

I expect in this case it's a matter of having magnets in the correct locations as specified by MFI.

Isn't "MFi" just their brand for licensed accessories?

Why doesn’t Apple have an option for buying a MacBook Pro with the camera and mic physically disconnected? They could sell this option at an appropriately priced premium — along with a stylish Bluetooth-connected, 4K webcam with phased array mics for a very reasonably $499. I’m sure they would sell hundreds of thousands of them.

If ANYBODY could monetize “security chic” it’s Apple.

I think you're vastly overestimating the demand.

How hard is it to stick a piece of tape over the camera and super-glue the microphone?

The underlying sensors are still active. I typically just keep my Logitech G430 headphones -- with the physical mute switch active -- plugged into my computer at all times because the system uses that for sound capture... except Siri/Cortana/Ubuntu One hear nothing when those are plugged in.

Camera, easy. Microphone: Hard, and might contribute to thermal issues.

Does it really matter if they have a hardware disconnect, when the hardware disconnect is controlled by software? It's not truly a hardware disconnect unless there's a physical switch that the user can flip on or off.

> On the 13-inch MacBook Pro and MacBook Air computers with the T2 chip, and on the 15-inch MacBook Pro portables from 2019 or later, this disconnect is implemented in hardware alone.

What said its software controlled? I’m assuming it’s a magnetic switch.

Perhaps this is the hardware people are speculating about with magnetic switches?

https://www.ifixit.com/News/33952/apple-put-a-hinge-sensor-i...

Though it seems their "hardware only" solution was on earlier models than that, so maybe not.

> (The camera is not disconnected in hardware, because its field of view is completely obstructed with the lid closed.)

I love this aside xD

Why not a manual switch? I rarely use the microphone on those devices.

This would result in a different UX.

Users would have to remember about turning it on/off, many would forget, leading to frustration.

Also Apple would have to place that manual switch somewhere, this would change the device's interface.

I once had a laptop with a manual switch for the wifi. Its sole purpose, as far as I could tell, was to get accidentally bumped and cause me to waste time figuring out why wifi wasn't working.

I feel you, but there would be a really easy solution to that, just make it clear with the network connectivity icon that the hardware switch is off. Don’t know why it hasn’t been done.

It would require Microsoft to standardise the interface and specifications for manual wifi switches, and certify and police compliance, to make sure the hardware and software worked properly together. For Apple this sort of coordination across teams is routine and simply the way everything works, but for Microsoft and its hardware OEMs it's a serious time consuming and expensive pain.

Thanks, I hadn’t thought of that.

So long as the software could read the setting of the switch it probably wouldn't be too frustrating.

Lenovos used to have a physical wireless disable switch, and I recall it only causing a moderate amount of frustration and I doubt much effort was put into making it problem free.

Mechanical switches are prone to fault.

Compared to what? In my average day, I experience 5-10 software failures/crashes/errors on various web apps, mobile apps, in windows, etc.

The last time I've seen a hardware switch fail was 2 years ago.

Further to that, hardware failures are predictable. Lifetime of switches is well know, and they are easily replaced. Good luck fixing a software glitch by yourself as a user.

Consider that a laptop already comes with a hundred switches in the keyboard, and they can easily last a decade if you dont spill coffee on it.

Software failures are fixed by reboots/patches. Hardware failures are fixed by service. And that service costs much more for vendor. Very few users are ready to replace hardware switch in their laptops.

Keyboard is a different kind of switch and they are major source of pain for many users. My macbook is broken because of keyboard, for example. One reason why new hardware devices like phones and tablets are coming without keyboard is this reason. Apple is not "brave enough" to replace macbook keyboard with touch panel, but they would do that if users would accept it.

Keyboard are a source of pain for many users? Where is this even coming from?

Large tablets have been around for nearly a decade now, and people that write novels and word files all day still use a real keyboard. If what you are saying was true, everyone who types for a living would use iPad pro or equivalent.

Secondly, patches and reboots are a problem, not a solution. Users do not need to replace switches themselves. My father is not an IT nerd, he uses his laptop for work. After the keyboard broke, he took it to a nearby repair store and they replaced it for $20. If in a few years it breaks again, he can do it again (ofc laptop might be worth replacing for performance reasons, but that's different).

To him software crashes are far more disruptive, he cant call Acer or Apple and give them $20, $2000 or even $w0,0000 and get them to fix a sound driver or whatever.

He just has to sit there and hope that some day one of the updates, without proper patch notes, will fix the problem. Leaving feedback on the forums is about as reliable as praying for rain or doing blood sacrifice to cure the plague.

I'd rather have a manual switch for both the microphone and the camera.

I would too, on every device, but I don't know how the industry is going to get past two barriers to achieving this:

1. It costs a little bit of money, and hardware designers love to minimize costs and eliminate parts.

2. It's not idiot-proof enough for a lot of people. You're going to have people who can't find the switch. And who complain and generate support costs. And annoy their coworkers by being muted during video conferences.

Maybe you could fix the second one by having a flashing light on the switch that says, "Please switch this switch! The microphone is needed!" This could also increase awareness of when things are trying to use the microphone. although it would go against the first point because it's yet another part.

I'd rather have a manual switch for both the microphone and the camera.

Then buy a laptop with those features. If any exist.

They do not reveal how this is implemented at the physical level. I'd think such a statement should be accompanied by a circuit diagram.

The simplest and most fool proof method is a physical switch that interrupts the mic lead. If you want it to be 'pop and crackle' free you may have to fiddle a bit or use a make-before-break switch and connect the other line to ground. No amount of software hacking will get around that, and the position of the switch is good feedback that you are not open to eavesdropping. Anything less than that is likely hackable in some way or other.

And yet they’re still not encrypting iCloud Backups in a way that doesn’t allow Apple (and by extension the federal police) unfettered access to spying on iMessage. iCloud Backup is on by default, so this means that iMessage for almost every iPhone user is insecure. Incongruent.

https://www.reuters.com/article/us-apple-fbi-icloud-exclusiv...

I did not realise this feature existed. Am pleasantly surprised.

Would also be nice to have little hardware LEDs that automatically turn on when the microphone or camera are connected. That way you are sure to know if they are recording.

This may be Apple's first real response to the checkm8 vulnerability. It had already been fixed on the current iOS devices as of its disclosure, but the T2 chip is still shipping (and is currently on computers that will last for a decade versus phones that last two years) with no fix AFAIK.

Apple could fix T2s going forward, but it would require a BootROM change not just a software update.

So, the next target for malware creators would be to attack the software that communicates a "lid closed" event to the T2 chip.

I have a suspicion (based on absolutely no evidence, other than that I'm an Engineer, and expect Apple has some brain cells); that the way this feature works is with a magnet and reed switch to physically disconnect the lines with no "communication" via software/firmware or other electronic means;

Based on their "hardware only" wording for the newer devices, I'm inclined to believe this is the case.

There exist easier targets, such as the loudspeakers which can be turned into microphones.

https://www.usenix.org/system/files/conference/woot17/woot17...

I asked my Alexa if this is true, but she didn't know. Anyway, this article makes me feel safer. I'll keep my Mac shut.

Hmmm I wonder if anyone has done a teardown of the MacBook to show how these microphone disconnect mechanisms work.

Feature request: iOS system-wide option to disable all optical cameras. FaceID can continue to use the IR camera.

This is not enough, the sane solution needs physical switches like the Lenovo T400 had for WiFi.

Where are the real laptops? I'm tired of the glossy short screens, lack of ethernet connections and crappy keyboards (this applies both to Lenovo and Apple).

I had the T400 and the physical WIFI switch was more of a bug for me instead of a feature

Purism.

Thou shall follow the FrankenPad route.

Works for the ISS and OpenBSD users

Why is apple so allergic to physical switches, directly switchable by the user?

I’m guessing easier to break, causes more troubleshooting issues, looks cleaner without it. The less moving components there are, the less warranty and tech support they probably have to field too.

Swapping out devices is probably less costly than fixing parts of it.

Depending on your point of view, only the second "reason" is true.

Software switches like the one described in the article (despite its name) are definitely easier to break since they can be hacked by software. No need to have physical access to the computer. A discrete switch definitely looks cleaner than a piece of duct tape, and is much easier to turn on/off. If you want to "switch off" your mic, bad luck.

This is pretty cool. Can I interact with the switch directly? Can I use a little magnet to trip the switch on an iPad? Or a mac? What about iPhones, do those have hardware disconnects too?

I think this is a positive thing and like the idea of being protected even if the machine is compromised. However I would still prefer the convenience of a switch in addition to this.

Any idea does other manufacturers does this? Dell or Lenovo?

I'm impressed with Apple's attention to details like this! ️

My Lenovo Laptop has a little physical shutter you can slide over the camera. It's very handy.

Does other manufacturers does this?

Maybe it's a standard thing that Dell, HP and others do as well?

Purism do hardware switches for mic/camera and wifi on their Librem range. (Wish I could afford one of those machines!)

Begs the question - how do they detect the lid is closed? In software? :P

Magnets!

Hold on. How is this hardware? Is there a physical button/hinge that is triggered? If so, the chip shouldn't matter.

Or is this some piece of logic emebedded in the chip itself or the firmware of some other component, which makes this solution a software one.

My understanding is that the chip is a proxy for when they started including it in the design. Similar to how some older macOS/OS X releases had requirements like, "faster than 867MHz", or "newer than 2012" instead of, "GPU must support Quartz Extreme" or "CPU must support XYZ instructions".

But how am I supposed to use my computer when the lid is closed?

With an external microphone, if you need it.

I think you misunderstand. I want to use my computer normally while the microphone and camera are disabled.

"The camera is not disconnected in hardware, because its field of view is completely obstructed with the lid closed".

How long until someone realizes audio data can be extracted by from the noise generated by a camera in the dark?

Source? Not because I disbelieve that it's possible in practice, but how much data can you really get out of that? A MacBook Pro camera operates at 60 Hz; unless you can interpolate samples some way I'm not thinking of (and that would necessarily degrade your resolution), the highest frequency you can capture is 30 Hz. Technically, this could be extracting audio, but it's not audio that meaningfully overlaps with the human experience.

No source, but there’s just been such a long history of side channel attacks, and similar research (https://nofilmschool.com/2014/08/mit-extract-sound-audio-sil...) that it makes one wonder if a non-hardware disconnect fix is sufficient.

> high-speed camera that captured 2,000 to 6,000 frames per second

Literally the next lines

> In other experiments, however, they used an ordinary digital camera. Because of a quirk in the design of most cameras’ sensors, the researchers were able to infer information about high-frequency vibrations even from video recorded at a standard 60 frames per second.

But I want to hardware disable it at any time...

How about disconnecting it when you're near the laptop and the lid is open because you're .. I don't know... working on it?

Is there any chance this is true? If so, is there a way to trick the hardware into switching microphone off, even while being used?

It’s a physical magnetic switch. The article does explain this.

But magnet does not appear in the article?

I suppose what I meant to ask is, does anyone have any evidence of this being actually true? If so, how would one specifically mimic a closed lid? In lieu of us doing our own experiments, has anyone discovered the appropriate size/strength magnet, and precisely where it could be placed, or whatever other tactic, to activate this hardware switch?

great! now give us back the hardware switch

We only get guaranteed privacy when the lid is closed?

The problem is that otherwise "Hey Siri" wouldn't work on (newer) Macs. Still, you can use something like Micro Snitch to track webcam/mic use.

Also, in macOS Catalina (I don't remember if this was the case in prior versions), applications are not permitted by default to use the mic or camera and have to request permission.

These are perhaps not as good as a hardware disconnect, but I think Apple is trying to balance privacy and usability here. It is clear from what Apple is doing in hardware and software that they do care about privacy. Linux and AFAIK Windows do not provide that level of privacy, since applications have unfettered access to Cameras and Mics.

(Of course, a part of the Linux community is trying to improve this through Pipewire, Flatpak, and portals.)

>The problem is that otherwise "Hey Siri" wouldn't work on (newer) Macs.

I've never heard anyone use Siri on their Mac. I'm surprised they still bother supporting it.

Me neither, but I also wonder how many non-expert users actually know that it is available. There is the initial dialog about Siri when setting up a Mac. After that, it is not that prominent. Moreover, it's a bit embarrassing to use in an office environment anyway.

That reminds me: I knew some people who had been using Macs for years who did not really know that Spotlight existed, let alone that you could use Command + space to start a Spotlight search. They would launch applications through the Finder and put regularly-used applications in the Dock.

I use it all the time. It's a great additional channel for commands in some workflows.

Can you give some examples? The only way I can think it's helpful is for hands free use, but that's more a scenario for a phone or tablet rather than a laptop.

Adding reminders is the only thing I ever use it for. They carry across to the iPhone, of course, so if I need to remind myself to do something later in the day it's quite helpful.

I'd use it to set timers, too (by far the most frequent thing I use Siri for on iPhone) but annoyingly that doesn't work from the Mac.

I got new AirPods and loved it for about an hour. "Hey Siri, open Spotify." "Hey Siri, next track" and so on while I was walking around. I'm spending most of my day on conf calls, so the AirPods are in anyway.

Then they started false triggering about every ten minutes. "Hey Siri, decrease sensitivity" does nothing. "Hey Siri, set a timer": "nope your $4k computer can't count time."

Turned it off. Sorry.

I use it to control my lights and heating when I'm in the same room as my computer (which is also my TV room).

As you say applications have to request permissions to access the mic/camera, here's a good writeup of how that was broken recently:

https://www.ryanpickren.com/webcam-hacking

That was not broken. What was broken was origin validation in Safari. There is a difference between a vulnerability in the OS permissions system (which this was not) to control mic/camera permissions and a vulnerability in an application that has already has OS camera permissions.

If you do not give an application permission to access the camera, then vulnerabilities in those applications do not lead to camera access.

(Unfortunately though, Safari is not controlled through this permission system, probably because it was provided through the OS. Permissions can be controlled for other browsers. IMO this should be fixed by Apple.)

I understand the extent of the security problem there, and as you say it was Safari-specific. But it goes to show that there might be consequences beyond the obvious when you do grant an application access to the camera/mic.

You are never guaranteed privacy.

Especially when slinging IP packets anywhere

Reply on news.ycombinator.com