We need physical audio kill switches

Nothing will ever come close to convenience of analogue volume knobs. They give both immediate action and infinite granular control. Same goes for kitchen stoves. All the modern touch controls with step-based control is the absolute nightmare in terms of UI.

I&#x27;ve been thinking the same about power switches lately. If I turn a flashlight, or an old radio on or off, I flip a switch and get the result I want. With my 65 EUR gamepad, or 300 EUR headphones, I hold a button and wait several seconds for the result.<p>Why has UX regressed so much in these areas?

&gt;&gt; &quot;mute buttons on keyboards shouldn’t need to go to software, they should immediately send a signal to the motherboard’s DAC&quot;<p>I don&#x27;t have a lot of knowledge about this, but audio is often streamed via USB or HDMI or Bluetooth or WLAN to another device, like headphones, an AV receiver, or wireless speakers and then uses the DAC in that device.<p>So the blog posts wish would result in the mute button not doing anything for many users, because it controls only a very specific hardware, but the users expect the function to work against all sound emitted by the operating system on any connected hardware, right? This guy actually wants a switch on his headphones. Or am i wrong and all sound goes through the motherboards sound card at least once, even if it is sent somewhere else?

Since the user is complaining about the way Mac OS handles the audio buttons, I&#x27;m surprised that he doesn&#x27;t mention the problems with HDMI.<p>I don&#x27;t understand why the OS disables the audio controls of HDMI outputs. If I press the volume down button, I want the volume to go down. Instead, I get a &quot;I can&#x27;t let you do that&quot; warning!<p>This is particularly bad if I have my headphones connected to the display (I have both my desktop and laptop connected to the same display, keyboard and mouse, being able to share the same headphones would be pretty convenient).

I let my spouse, who likes the novelty of a touchscreen, talk me into a Whirlpool oven a couple years ago that is controlled by touchscreen. I had no idea it was possible to hate an oven so much. The only saving grace is that the stovetop elements are still controlled by knobs.

&gt; <i>I hit the mute button on my MacBook Pro Touchbar, and it took a solid two seconds for it to register. My ears were ringing throughout the whole call. This is unacceptable. Well-engineered mute buttons on keyboards shouldn’t need to go to software, they should immediately send a signal to the motherboard’s DAC—ideally on a separate wire or connection—to say terminate this signal. Then it’s less of a concern if it takes the OS a few seconds to react to the change, because our ears have been spared.</i><p>&quot;Should not need to go to software?&quot; I&#x27;m not sure.<p>I have a Yamaha AV receiver on my desk. The huge volume knob has the best user experience than any software-based solution, unlike menus or hotkeys, it&#x27;s always in my reach and instantaneously responses to my inputs. It has smooth mechanical damping, giving a satisfying touch and feel. I can immediately turn the volume down in a second if there&#x27;s misbehaving, non-normalized loud audio. I did it from time to time.<p>On the other hand, I clearly know it&#x27;s software-driven. The knob is nothing but an user interface. It&#x27;s an optical rotary encoder, not a real potentiometer. When the knob is turned, it chops a beam of light, and an optical sensor converts the optical pulse to a digital signal, which is then sent to a GPIO pin on the application processor in the AV receiver. The firmware on the application processor receives an IRQ, then commands the DAC via the system bus (could be I2C) to increase or decrease the volume in a 0.5 dB step. It&#x27;s 100% software driven, not hardware driven. The software can fail, yet the software never let me down.<p>We can get another conclusion from this: Special-purpose computing is more reliable than general-purpose computing. The operating system handles everything, and has numerous failure modes due to its unpredictability. Furthermore, it&#x27;s in a constant state of change to bring the &quot;Next Great Things&quot;, who cares if the audio mixer or the TouchBar becomes a bit slow? An AV receiver only needs to do one thing, so it can do this well.

&gt;I hit the mute button on my MacBook Pro Touchbar, and it took a solid two seconds for it to register. My ears were ringing throughout the whole call. This is unacceptable.<p>The USB-C era MacBook Pros had some serious degradation in the audio department. The previous much loved generation, you plug headphones in the audio system switches to them so fast it&#x27;s perceivable as instant, the USB-C era there is at best a 0.5 delay and at worst a 2 second delay.<p>During this delay anything you change volume wise will actually apply to the previous setting not the new setting, so the usual scenario where you&#x27;re in the office you have the speakers on mute, plug in headphones then you don&#x27;t hear anything because of the delay so you hit volume up will actually set your speakers off mute. Later on you unplug your headphones and your laptops speakers are then blaring out whatever you were playing instead of remaining on mute.<p>This situation would never happen in the pre-USB-C MBP because the headphone switch was instant, it felt like a hardwired connection. Like a lot of things on the USB-C MacBooks interactions like sound and power feel more like interfacing with a what people may call a &quot;device&quot; rather than a computer, there are tangible layers that have to handshake between your real world action and the hardware catching up sometimes seconds later.<p>(This is all the non-touchbar model, so it&#x27;s not just that)

I have accomplished this by using an external audio mixer at my desk. I have physical volume controls for three PC&#x27;s, two ham radios, and a police scanner. Then there is a master volume control for everything mixed.

I can continue this rant for making physical buttons mandatory in many places, MAINLY because they do only one thing(mostly) but they do it pretty damn well AND its reliable as hell!<p>like a knob to adjust the shower temperature or a knob&#x2F;button to adjust the car volume (in some cases these avoid accidents too).. I know where those keys are, eyes closed!!!<p>Physical buttons are becoming a luxury now a days :( , and don&#x27;t shove in the &quot;hey its software age &#x2F; touch screen age&#x2F;minimalistic age&quot; here :|<p>Minimalism doesn&#x27;t mean , remove ultra necessary stuff..

My old IBM ThinkPads operate exactly as he described, the embedded controller immediately instructs the hardware amplifier to adjust its volume or power down.<p>My slightly less old Lenovo X201t does everything in software, and while my computer is still groggy after waking from suspend it often takes 10-15 seconds for volume changes to take effect.

Author should just use a proper browser with ad-block and mute add-on. Regardless of audio levels or hardware switches, a website should never play audio on its own without permission.

A master and a student are sitting in a forest, reflecting. Two men walk by, vehemently discussing whether a household item&#x27;s design is acceptable.<p>After they pass, the student tells the master: &quot;The opinionated user-experience enthusiast is right. Every wrong decision is not just a single day ruined, time wasted, or train of thought lost, but thousands. However, the pragmatician is also right: it is not reasonable to expect every object we use to attain UX perfection. They are built by flawed people with real constraints. Tell me master, which one of those men is closer to the Way?&quot;<p>The master thinks for a while, then gives an answer. But the student does not hear it, because he forgot to turn off his Bose™ noise-cancelling headphones.

The real problem he has is different levels of audio and missing normalization of audio levels. You shouldn&#x27;t get into the situation of wanting to quickly adjust audio levels in the first place (although I agree that any latency in adjusting the volume is inacceptable).<p>Normalization (especially the real-time case) is an interesting topic. I&#x27;ve experimented with crude algorithms in the past to take all audio samples in a fixed window and then normalize them all to the same level [0]. But it&#x27;s harder than you&#x27;d think. First there is a difference for humans in how loud they perceive the same sound to be at different frequencies. So any measurement function needs to take that into account. ReplayGain does this btw [1]. Then, there is the problem that if you make this window too short, you turn the audio into a garbled mess. If you make the window too long, you increase latency. This is a big deal for streaming settings. And last, you have the problem that silent periods aren&#x27;t silent. E.g. if the person takes a breath, the level is increased to amplify the muffled sounds of the road nearby to levels to make you think they stand in the middle of a highway :). It&#x27;s an interesting problem and while I doubt that it&#x27;ll be solvable by simple hardcoded algorithms, ML might solve it. Then we can finally enjoy audio without having to manually press +++ and --- all the time :). But you know maybe we&#x27;ll have different problems similar to the inability of phones to photograph the orange sky over SF.<p>[0]: <a href="https:&#x2F;&#x2F;github.com&#x2F;est31&#x2F;js-audio-normalizer" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;est31&#x2F;js-audio-normalizer</a><p>[1]: <a href="http:&#x2F;&#x2F;wiki.hydrogenaud.io&#x2F;index.php?title=ReplayGain_specification#Loudness_filter" rel="nofollow">http:&#x2F;&#x2F;wiki.hydrogenaud.io&#x2F;index.php?title=ReplayGain_specif...</a>

I use a dynamic range compressor when dealing with unpredictable output (watching Youtube videos, Zoom calls).<p>On Linux, I use this: <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=typM_AQUzi4" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=typM_AQUzi4</a><p>On Windows, I tick the &quot;Loudness Equalization&quot; option in the output device&#x27;s menu.<p>I disable it when I listen to music.

More important than speaker kill switches, we need physical&#x2F;hardware microphone switches that disconnect the circuit. These are built into Amazon Echo devices, but it&#x27;s unclear if Google Home or other Apple HomePod have this (a quick revealed they do not).

I have encountered the opposite problem.<p>Some speakers turn themselves off automatically when they don&#x27;t detect any sound on the line for a certain amount of time.<p>So when you want to have your monitoring application ring the alarm on the speaker, you need to be aware of it so you don&#x27;t miss it.<p>Depending on the speaker you may need to be continuously playing some very low volume audio to prevent them from auto-turning off.

The 2 seconds would be a bug or maybe a design fault in his computer. I use a big USB volume knob [1] which is also a mute button and it always responds instantaneously as far as I can tell. I just bash it when an unexpected video starts playing. It was hard to find but these things do exist and are refreshingly convenient compared to whatever fiddly volume controls laptops always have.<p>[1] <a href="https:&#x2F;&#x2F;www.aliexpress.com&#x2F;item&#x2F;32649442146.html" rel="nofollow">https:&#x2F;&#x2F;www.aliexpress.com&#x2F;item&#x2F;32649442146.html</a> Added link to answer reply question

I’ve got a microphone with a physical kill switch and when turned off it still leaks like a sieve. Be careful with those switches.

I wanted to love the touch bar, but the current implementation is a failure -- it needs to be pressure sensitive. A light brushing of a physical key doesn&#x27;t register and a good implementation of the touch bar would behave the same way.<p>In the meantime, I would pay extra to the replace the touch bar with the physical buttons we used to have. I&#x27;ve gotten so tired of accidentally hitting the touch bar that I have disabled it as much as possible -- I have it configured so holding the fn button brings it back, making it much more difficult to accidentally trigger something. Of course, as the article mentions, this makes it even more difficult to trigger the sound controls when they&#x27;re uncomfortably loud, but that&#x27;s still a trade off I&#x27;m willing to make to avoid so many false positive interactions with the touch bar.

The real design flaw here is not conveniently being able to set the volume per application. Most systems are able to do this, but don&#x27;t expose that functionality conveniently.<p>In my opinion the volume keys should only be adjusting the volume for the application&#x2F;website currently in focus.

ive been working remotely for a few years and do a little recording and mixing on the side. I like using plain old analog headsets since they make it easy to have control over the sidetone level from the headset mic, and they&#x27;re usable in many situations where a usb headset would not be.<p>To simplify using one headset with a few different devices (pc, desk phone, ipod&#x2F;phone), I wound up buying a little Behringer Xenyx 302USB bus-powered mixer. If you have a bunch of analog devices to mix, you&#x27;d want something larger, but for a PC with a headset and one or two other analog sources or outs (like a pair of desk speakers), these things are just about perfect. They use off the shelf TI PCM2902C converters, have excellent audio quality, work driver-free on Linux, OSX, and windows, and seem to last pretty well.<p>The real beauty of these little mixers (or the myriad other comparable ones) is that all the common adjustments I need to make- my own mic monitor level, the other party&#x27;s output level, and the overall headphone level, each have their own rotary fader on the mixer. There&#x27;s also some minimal eq available. The mixer is small enough that it doesn&#x27;t waste much desk space, and in the event of a surprise volume spike, the master volume fader is right by my hand, and I&#x27;ve saved myself from being blasted many times.<p>If you pair something like this together with an effects host (I use pulseeffects on Linux and love it), its easy to apply a compressor or limiter to the stream from the VC app that I use to reduce transients, or to add an expander to help on multi-party calls where one callers mic gain is too low.<p>The 302USBs have gotten a little tricky to find lately since they seem to be popular with new podcasters. Several companies make similar models but nothing has quite the same set of connectors and routing to be as convenient as these, although everyone&#x27;s needs are different- if you&#x27;re using a headset with 1&#x2F;4&quot; connectors, the Yamaha AG03 looks really promising, although its nearly twice the price of the Behringer.<p>There aren&#x27;t a lot of gizmos that have simplified remote work for me as much as these things; they&#x27;re probably overkill if you don&#x27;t work from a fixed location (although they don&#x27;t require any power connection), but if you have a typical desktop setup, they&#x27;re great.

I&#x27;m not sure I get this article - if you want a physical kill switch for audio, get a volume controller and set that directly under your monitor? Now you have a physical kill switch in the form of a GIANT near-zero-friction volume knob as well as dedicated hardware true-instant-mute button.<p>No reason computers should come with that out of the box, in the same way there&#x27;s no reason they should come with a DVD drive or a vertical mouse or brand-name webcam, but if that&#x27;s what you need in your life, any of these are &lt;$100 peripherals.<p>Because you don&#x27;t need a $2000 audio interface with twenty XLR inputs with individually controlled phantom power and built in EQ, or even a $299 one that podcasters and youtubers like to keep recommending: if you <i>just</i> want control over your audio, get something like a Behringer MONITOR1 for $60 [1], or a slightly nicer Mackie &quot;Big Knob&quot; passive 2x2 for $70 [2] and go &quot;ahh, exactly what I needed&quot;.<p>[1] <a href="https:&#x2F;&#x2F;www.behringer.com&#x2F;product.html?modelCode=P0C9G" rel="nofollow">https:&#x2F;&#x2F;www.behringer.com&#x2F;product.html?modelCode=P0C9G</a><p>[2] <a href="https:&#x2F;&#x2F;mackie.com&#x2F;products&#x2F;big-knob-series-monitor-controllers-interfaces" rel="nofollow">https:&#x2F;&#x2F;mackie.com&#x2F;products&#x2F;big-knob-series-monitor-controll...</a>

&quot;I was in a conference call last Friday where I’d adjusted the volume up to compensate for the client’s quiet microphone, only to be audibly shot in the ears by an auto-playing video on a website. There is a lot of problematic stuff to unpack there, much of which is not the fault of the audio hardware or OS.&quot;<p>If you&#x27;re suddenly reaching for the mute button because you&#x27;re being blasted then there has already been a significant failure. It just happens to be something <i>everybody</i> is failing at and nobody can be bothered to fix.<p>We suck at quantifying (simply) the loudness of audio content. If you set volume according to average SPL, high transients (e.g. Cannon fire in a too-faithful recording of Tchaikovsky&#x27;s 1812 overture) can still be shattering. If you set it according to the maximum SPL, an action movie with lots of explosions might have a lot of dialogue that&#x27;s inaudible. Audio that has been highly compressed to the point of distortion may seem uncomfortably loud even at moderate SPL&#x27;s.<p>The thing is, we probably shouldn&#x27;t even try to use simple rules for setting volume. We need to build systems that observe when people adjust the volume and <i>learn</i> from that. No OS should be dumb enough to suddenly let rip with a new audio stream whose first second is, on average, 30 dB higher than what was previously playing. Yet, they practically <i>all</i> do.<p>A hardware volume knob is an important thing for any audio device to have but, if the device has the capacity to run programs and it&#x27;s users are still reaching for that knob in startlement, there has been a failure.

I&#x27;ve been thinking about the &quot;you&#x27;re muted&quot; problem in video calls, and imagining a big physical button to mute&#x2F;unmute without having to stuff around with the particular UI for this particular video product you are using right now. One solution is to mute&#x2F;unmute at the hardware or system level (some computers already have this, or sometime close), but this then requires you to be unmuted in the software (Zoom), which isn&#x27;t always under your control. So unless the system-level mute can send a signal to Zoom, or Meet, or Teams, or X, then it defeats the purpose. What would be nice would be an agreed-upon standard API for muting&#x2F;unmuting within an application, but that both seems unlikely (as that&#x27;s not how companies like to do things anymore), and would add to the complexity of things to a point where &#x27;ordinary users&#x27; would get confused or stuff things up. I&#x27;m at the point where I think a dedicated device for video calls is necessary.

We also need hardware on&#x2F;off switches for cameras and microphones.

Absolutely.<p>We&#x27;re rapidly headed toward a world in which not only is volume controlled by software, but where apps and <i>even websites</i> will be able to tell whether your audio is muted. Imagine what Youtube would be like if not only are videos delayed by ads, but <i>the ads themselves refuse to play</i> if they&#x27;re muted.<p>Madness.

How about a kill switch in the gym?<p>Every time I walk into a business I&#x27;m assaulted with their choice of music. No music is as bad as gym music: unrecognizable songs by unrecognizable artists chosen by God-knows-who (all no doubt via the lowest bid).<p>Where is &quot;The Sound of Silence&quot;? It&#x27;s cheaper (free even) and better.

I don&#x27;t know what I&#x27;d do if my iPhone didn&#x27;t have physical volume buttons. When I watch videos on YouTube now, as soon as an ad starts playing, I just mute it by using the volume button and look away for a few seconds as I can&#x27;t stand ads (sadly there isn&#x27;t a dedicated mute button).

After glancing over the title I first thought the topic is physical <i></i>microphone<i></i> kill switches. That being said<p>We need physical microphone kill switches.<p>Even if you are super careful about getting your laptop infected or hacked, you have probably tons of legit software installed on your laptop that can be hacked, or spy on you for purpose, there are OS and vendor backdoors, zero days, and many of us use laptops managed by our corporate overlords.<p>I wish that microphones (and cameras) in your private space that are connected to a network and a battery would be taken more seriously.

When working with audio for production (work) stuff which you need to separate from non-work, maybe use something like Voicemeeter Banana and route your audio signals separately (and adjust loudness separately.) That way you can keep things separate so that nobody is picking up things you don&#x27;t want them to pick up (and the same for your end.) There&#x27;s also apps like Krisp which filter background noise from incoming calls. I don&#x27;t know if it sets loudness, but something like that would probably be good for saving your ears.

Good point, I know the librem 5 has cut off switches for the inputs--camera, mic, cellular. I&#x27;m not sure anyone ever thought of the output side.<p><pre><code> Kill Switches 3 – WiFi, Cellular, Microphone&#x2F;Cameras (all 3 will turn off GPS)</code></pre> <a href="https:&#x2F;&#x2F;puri.sm&#x2F;products&#x2F;librem-5&#x2F;" rel="nofollow">https:&#x2F;&#x2F;puri.sm&#x2F;products&#x2F;librem-5&#x2F;</a>

This sounds like an XY problem. What you want is for your ears to not be hurt by loud sounds. Physical audio kill switches is one solution but I can think of many others that I personally would prefer. One thing that comes to mind is a system that monitors the audio output and keeps that from exceeding some predefined maximum.

I agree with him. Also, why is there no microphone mute button (or shortcut) integrated in windows? Seems like a must-have nowadays.

This is one of the reasons why all the amplifiers I build for my personal use, be it power audio amps or headphone amps, don&#x27;t use any fancy audio switch ICs. I prefer analog exactly for this reason. A simple relay with a push button. Leaves me more room for other stuff inside the cabinet, while getting the job done in the simplest way possible.

Physical audio kill switches are a nice idea, but if you didn&#x27;t wire them yourself, how do you know they aren&#x27;t merely claimed to be physical but are actually just setting a flag in the operating system to ignore the physical input from the speaker?

Anything that goes into or comes out of the computer needs a physical kill switch. I have no idea why this is even a question.<p>Best I can come up with, the changes were so incremental and advertised as features. There was no natural stopping point, so vendors kept going.

After a volume shock it&#x27;s too late. I think there should be a limiter&#x2F;compressor by default or something, especially sensitive to when there is an initial loudness jump.

The Bose Companion 20 multimedia speaker system for computers has a desktop control pod to plug in your headphones that is for volume control and an on&#x2F;off button.

&gt; proclaim that people could just unplug their headphones<p>Actually I found plugging <i>in</i> headphones is the easiest and quickest way to mute mobile phones.

If you physically cut a audio signal, one end of the cable turns into a broadcasting radio antenna. Reality isn&#x27;t as simple as presented.

you could make the same argument about a lot of things on your PC, or elsewhere. simpler is always better, until you need to do a lot of things.

What we really need is for engineers who never wanted to remove them to begin with to be in charge.<p>Starting long enough ago to make the difference.

I have certainly experienced this on Android. Random apps seem to be able to override my &quot;mute the phone&quot; command.

It&#x27;s much more of an issue without headphones. Anyone who ever their laptop start playing audio in a meeting knows.

That why whatever conferencing software your using should have audio processing to control level of different audio.

Also need physical write-enable switches for ROMs, and physical mike and camera enabling switches.

it basically boils down to either buying a logitech craft or any decent audio interface and move on...<p>while such knob would be nice on a laptop I don&#x27;t think it&#x27;s downsides would justify adding one...

more importantly. we need physical video kill switches as well! i have dealt with so many apps just opening the camera out of nowhere and it’s also mostly unexpected.

Did he forget about pulling the cord out? Lol