The case for buying an expensive pair of headphones

Like many of us, I found myself working from home at the start of the pandemic. Initially I set myself up in the sitting room but as things progressed to the semi-permanent, we got rid of the office for my startup Ruffly and I decided to move somewhere that wasn't stealing another space from the rest of the family.

My wonderful wife works from home permanently in a home office and I could have moved in beside her. However, she has a tendendency to break into loud peals of laughter during conference calls which I find endearing through a closed door but has a high possibility of leading to divorce from 2 feet away1.

The place chosen in our house was the upstairs landing. A space that already had a desk and chair that had been set up with a computer in an observable place for the kids to use, back when we niavely thought, pre-pandemic, that we would be rationing screen time for them2. This space is exposed to the comings and goings of 3 children, the dog Dougal and my aforementioned wife. It is open to the main hallway of the house which has a wooden floor. All the doors in our house are Edwardian and make a terrifyingly loud ka-thunk when they close. I have line-of-sight down the stairs into the kitchen, a room which has the capacity to produce an array of metallic sounds that frankly astonish me - more on these later. All this to say, this is not a naturally quiet space.

After the first lock-down ended and the children were packed off back to their incubation pods at school, life got tolerably quiet and I was able to worry about more minor auditory annoyances in life. The cupula above our hallway has an airgap for some reason, and it lets in the dull hubbub of the primary school playground next door. I spent an uncomfortable weekend up a ladder, filling the gaps with a noise insulating expanding foam 3. Despite this, I gradually started to wonder if Active Noise Cancelling (ANC) headphones of some sort might be a good answer for working from this new space.

In December 2020, Apple announced the Airpods Max and reviewers started waxing lyrical about how amazing these (extremely expensive) new ANC headphones from Apple truly were. As an Apple-eco-system stalwart, I considered that if I was going to spend hundreds of pounds on a pair of great ANC headphones, it might as well be a couple more hundreds of pounds for an Apple pair of headphones that would a) give the best performance and b) "just work" with all my Apple kit. As threats of a second school lock-down came to fruition I bit the bullet and ordered a pair.

My requirements

At this point its worth stating what I was hoping for from the Airpods Max, in priority order:

  1. Headphone noise cancellation that would cut out the noise of the house and it's occupants coming to my ears
  2. Microphone voice isolation that would cut out the noise of the house and it's occupants reaching the ears of participants on video calls
  3. Good-to-great sound quality of my voice on video calls
  4. Great sound quality for listening to music while working

Most of what has been written about the Airpods Max is true

The headphones arrived the second full week of January, a week before my wife and I had to start home-schooling the kids again but a time when both of us were back juggling full-time jobs with minding our 6, 8 and 10 year olds, plus the 10 month old puppy.

I'm not going to say much about the good bits of the Airpods Max because plenty has already been written on that. They do sound really great for music and for films. The most noticable thing to me was the way they project the sound in front of me when I'm listening to a normal stereo video on Youtube/Twitter/TikTok etc on the iPhone. It comes across as a very definite sound projection of about 30cms. I wouldn't say I like it especially more, but it's definitely a thing.

The ANC is exceptionally good at cutting out background noise. It fades it when you turn it on and I love the effect - it brings to mind a sound animation. I have subsequently tried a pair of Bose NC 700s and I can categorically say that the ANC on the Airpods Max is more effective at cutting out the normal background noise in my house.

However, as implied by the title, this isn't an article with a happy ending. In the end I had to send the Airpods Max back to Apple for a refund because they were four things that just did not work out for my use case. Without further ado, here they are...

The noise cancellation sucked

Let's tak about that ANC again. The use of 'sucked' in the title isn't (just) click-bait. The very first thing I noticed about the Airpods Max was that the ANC didn't feel great in my ears. I'd describe it as like the feeling you have when you need to repressurise your ears on an airplane. It doesn't bother you particularly at the time but when you swallow you realise it feels better. Taking off the Airpods Max felt like returning my inner ears to their natural, unadorned state. When I posted this on Twitter, @madmax helpfully pointed me to an article about Eardrum Suck3 which I'm inclined to believe is the issue here. The ANC effect is very good at filtering out background noise, but it definitely feels like an effect.

And then there is event noise. Event noise is what I'm calling any sudden loud noise that is over and above the normal background din. In my environment that is raised voices, dog barks, footsteps on the wooden floor, metallic noises from the kitchen and the previously described ka-thunk of closing doors. The ANC on the Airpods Max does not do well with any of these. To be fair it is a tough challenge. But what I was not expecting, what blew me away4 was that low frequency event noise (e.g. a ka-thunk of a door) could actually create an uncomfortable ANC burp right into my ear drum. Like a Q-tip that's gone too deep. It's not pleasant and it happens every 10 minutes in my house. Imagine that - a Q-tip to the ear drum every 10 minutes! It doesn't take much of that to make a man want to take his headphones off his irreplaceable ear drums, no matter how expensive they were!

They do not sound good for wireless calls on the Mac

Believe me when I say I wish I this wasn't true. Apple have done a magnificent job of making the bluetooth wireless experience acceptable when you are listening to music. It's damned near perfect on the iPhone, and given the crapshoot that is bluetooth accessories on the Mac - it worked really quite well there too5.

Without going in to too many details (I have researched the death out of this trying to get it to work), when the Airpods Max are connected as an output device, macOS will negotiate a 48kHz streaming audio session using AAC. It might not convince some music purists but it sounds just peachy to me.

When, however you tell macOS to use the Airpods Max as an input device things start to go downhill. It will negotiate a two-way SCO connection that reduces the audio quality for both the input and the output device. Yes that's right. If the microphone is active, everything sounds worse too. How much worse? Well I'm glad you asked. My desktop Mac6 uses a Bluetooth 4.0 chipset. When the Airpods Max connect to a Bluetooth 4.0 chipset you end up with 8kHz audio. You can search the internet for this - it also happens for the regular Airpods and Airpods Pro. If you've been on a Zoom call with someone using Airpods that sounded like they are on the world's worst telephone - this is probably the reason. It can support 48kHz streaming audio simplex but only 8kHz duplex. I'm not a Bluetooth engineer but something doesn't add up here.

I also have a 2015 era MacBook pro that has a Bluetooth 4.2 chipset. This chipset can support a 16kHz duplex connection to the Airpods Max. It's better... and frankly if my desktop Mac had been able to achieve this I might not have noticed the quality problem until after the return window had closed. But we are still getting a sub-par audio experience if the microphones are in use. The standard response from pretty much everyone on the internet is to turn off the device as an input source and use the internal microphone on your machine. They even have complicated instructions on how to fool the Mac into doing this automatically and one of the most popular apps for Airpods management ToothFairy has the ability to force this built-in.

Hell no.

I did not spend all this money on headphones with noise-isolating microphones just to turn it off.

On my iPhone 11 Pro things are better still. It is able to support a 24kHz duplex connection to the Airpods Max, which (if I'm reading the debug messages correctly) might even be using the AAC codec. I'm not sure if this is a Bluetooth 5.0 bandwidth capability or something else. But suffice to say that if you are getting 24kHz duplex then you can probably stop worrying about wireless quality problems. I wish I had a modern Mac with Bluetooth 5 built in to verify that the same is acheiveable on a modern chipset Mac. I'd be willing to believe so and it leaves me confident that Apple are never going to invest the time in making this better for older hardware. Heck, if it's great on the iPhone and not great on the Mac, that's probably ok by them too.

What 8kHz audio sounds like

I recorded some samples to drive this home. You can make your own mind up at which point you think it's good enough to make voice calls. As a benchmark against wired audio I've included the iPhone recording from the internal microphone and both a cheap Logitech USB headset and an expensive Rode Procaster microphone that are wired into the Mac.

8kHz Wireless to Mac

This is what I sound like on calls and don't forget, what all audio sounds like to me when I have the Airpods Max set as an input device. Admit it... this is worse than you expected right? Just to be totally fair - this is the exact same outcome I get from a pair of Bose NC 700 headphones too.

16kHz Wireless to Mac

This is probably ok for most people. I'd probably live with this if everything else was fine.

24kHz Wireless to iPhone

This sounds good to me, if a little whistle-y. There is a small amount of background noise being picked up but I think the voice is totally clear. If you live in a quieter house than I do, and you have a machine that will give you 24kHz recordings over Bluetooth, you'd probably be very happy with these headphones.

48kHz Built-in Mic on iPhone

At this stage there is no loss of quality from the transmission but the microphone is picking up more background noise.

48kHz Logitech Headset on Mac

The cheap old Logitech Headset can also do 48kHz without breaking a sweat over it's USB connection. The microphone doesn't sound great but it naturally does a decent job of blocking out background noise purely based on the dynamics of the setup.

48kHz Procaster Mic on Mac

This is why audio professionals recommend proper microphones. It does sound a lot better. I have it a bit quiet here to try to cut out background noise. Despite this being a dynamic microphone it will easily pick up a tonne of the noise from the space I'm in unless I turn the gain down and stick my face an inch from the microphone.

You cannot make wired calls with the built-in microphones

You've probably heard that you can buy a £35 lightning to 3.5mm audio jack cable from Apple for using the Airpods Max wired. It's not at all obvious to me why this cable is so expensive. I've heard it said that it has to contain a digital to audio converter in it because Lighning cannot transmit analog audio which I suppose might be true but a) that sure must be a tiny DAC to fit into those end-pieces and b) why the hell didn't they use a USB-C connector instead then.

Either way I begrudgingly ordered this cable when it became clear that I was having Bluetooth quality issues. The standard use case you hear mentioned is that this cable allows you to use the headphones for direct monitoring without any delay7. When the cable arrived I fully expected to be able to plug in the cable and get beautiful 48kHz duplex functionality without a hitch. Not so fast. It turns out that the cable is TRS, not TRRS, and so it does not carry the microphone input over the wire8.

This is insane.

In fact because it cuts off the wireless communication while in use, you lose the ability to use the headphones as an input device altogether. At this point I started to wonder if I'd missed a trick. Perhaps this cable really was just for direct monitoring by pros? Surely a much more sensible approach, and justification for that lightning connector decision, would be to allow a bog-standard USB to lightning cable to transmit audio to and from the headset. This setup had the potential to deliver the holy trifecta of 48kHz output + 48kHz input + charging, from a cable that costs only a few quid.

Nope.

It only works for charging. At this point I was definitely losing faith, but I wasn't ready to give up yet. I was seriously thinking of ditching my current desktop Mac and getting a new Mac Mini in the hope that Bluetooth 5.0 would solve all my woes and give me that sweet, sweet 24kHz duplex with perfect voice isolation from the headset microphones.

They do not isolate your voice from event noise

Which brings me to the final nail. I thought I'd better test how well the headset microphones isolated my voice from event noise when I was connected to my iPhone. Without the potential for a wired connection, this was the very best I could hope to experience from these headphones in this environment. I only ran this test a couple of times. It's not a great sample, my wife was preparing some cheese9 and oatcakes, along with plates and cutlery being placed onto a hard kitchen counter. You need to listen until at least the 33 second mark for the event noise to start in earnest.

Airpods Max mics event noise

The headphones do not reduce the noise of the plates and cutlery at all. This noise is probably 10-15 metres away from where I am through an open door but without direct line-of-sight to the actual noisemakers. It sounds through the headsets pretty much how it sounds in real life. It's not as loud as if you were standing beside it, but it is certainly not appreciably cancelled.

Conclusion

I honestly do believe that part of the value proposition of buying the Airpods Max is that Apple will continue to support and improve them over time. It's not impossible that some of these limitations could be improved by a software update. The £35 cable will never transmit the microphone but they might come out with a different £35 cable that does, or heaven forbid, support audio over USB10. In more likelihood we might continue to see improvements in voice isolation from event noise. They do such a terrific job with ANC that I do believe all the bits of hardware are already in the Airpods Max to make big strides here.

However, for me, right now, in the way I want to use them, the Airpods Max kinda suck. I initiated a refund from Apple yesterday and sent them back. In a fit of pique I ordered a pair of Bose NC 700 headphones for overnight delivery and they turned up today. As you can see above, they also failed to work for me in a similar set of ways. Having been through a lot of trial and error with the Airpods I've already admitted defeat on them too and sent them back. I've now ordered a pair of Jabra Evolve2 85s. They support audio over USB. They have their own bluetooth USB dongle to work around chipset limitations. They have a little boom that swings down in front of your face which I'm assuming is the very best possible aproach for isolating a voice from event noise. They arrive tomorrow and they could finally be the solution to working in the hall in a household with 3 kids and a dog call Dougal. I live in hope.


  1. This is definitely my problem, not hers - I am, shall we say, quite particular about noise when I'm trying to work :/
  2. LOL
  3. This worked, in the sense that while it only cut out about 10% of the noise, it made me feel 100% better for having tried it
  4. Or at least, blew my eardums away
  5. I'm hedging because it's not perfect, but its better than anything else
  6. Yes, I have a hackintosh. No, that is not the reason.
  7. As an aside, I do not think that transparency mode quite fits the bill as a replacement for direct monitoring for me but its close. If they let you adjust the volume of your own voice in your ears I think it would go some way to convincing me that direct monitoring wasn't required.
  8. Again, the Bose NC 700s are the same. Why would you do this when there are so many compromises with Bluetooth and you've already included a bloody socket on the headphones!?!
  9. It was, as she remarks, smelly cheese
  10. It was via a firmware update don't you know!