[tinyman392]

Yes, but generic statments can be made to be interpret as wide range of stuff from something really minor to something very precise. I assume not something precise, but some fudged approximations. They can't be measuring your ear anatomy, which would be the precise way to EQ if they figured out ear geometry to target response correlation. Doesn't Sony have something that asks for an image of your ear?
In addition, they couldn't possibly measure the response as the eardrum so how do they know what precisely should be EQ'd?

Can you explain to me how EQ is done with pressure?

Apple's patent on their Adaptive EQ (HomePods) only talks about using microphones and pressure sensors to create a field of equal acoustic pressure around the speaker. So they have an algorithm that given an input of variable pressure, will apply an EQ to correct the acoustic pressure differences. Using microphones mounted at the ear cups they can pick up different aspects of the ear like resonances (maybe other things as well) to see what's absorbed and what isn't to help deal with those pretty easily.

This sounds very similar to what AKG does with the TruNote auto-calibrating system employed in their N90Q.

Edit: For something real time I could see a microphone picking up what's being played through the speakers of the drivers. It knows what should be being played. "Volume match" at 1k and subtract the difference, then subtract off the target curve and the headphones probably have a good idea of what's different between the sound played and the wave form that was expected to be played. Adjust EQ accordingly. Moving the driver and its location can cause differences as well which would be accounted for in the next round of Adaptive EQ. Note this edit is speculation as Apple could be doing something completely different.

Does somebody have a good technical explanation of what is this and how it works?

See above. That's really all you're going to get is the patent on how it works. The actual algorithms used to do the processing are probably trade secrets. But the idea has been done before (AKG, 2015).

 

[SilverEars]

Apple's patent on their Adaptive EQ (HomePods) only talks about using microphones and pressure sensors to create a field of equal acoustic pressure around the speaker. So they have an algorithm that given an input of variable pressure, will apply an EQ to correct the acoustic pressure differences. Using microphones mounted at the ear cups they can pick up different aspects of the ear like resonances (maybe other things as well) to see what's absorbed and what isn't to help deal with those pretty easily.

This sounds very similar to what AKG does with the TruNote auto-calibrating system employed in their N90Q.

Edit: For something real time I could see a microphone picking up what's being played through the speakers of the drivers. It knows what should be being played. "Volume match" at 1k and subtract the difference, then subtract off the target curve and the headphones probably have a good idea of what's different between the sound played and the wave form that was expected to be played. Adjust EQ accordingly. Moving the driver and its location can cause differences as well which would be accounted for in the next round of Adaptive EQ. Note this edit is speculation as Apple could be doing something completely different.



See above. That's really all you're going to get is the patent on how it works. The actual algorithms used to do the processing are probably trade secrets. But the idea has been done before (AKG, 2015).

Here's one for something like APP, although represented as wired for it's patent. All I see is shelving adjustments, so it's nothing sophisticated. Probably just check for seal and leakage. They probably like to make it sound like it's something sophisticated, but it really can't be. It's basically low-end adjustments with a simple shelving filtering.

https://patents.google.com/patent/US9264823

Here's Harman's

https://patents.google.com/patent/US10219067B2

 

[MayaTlab]

Yes, but generic statments can be made to be interpret as wide range of stuff from something really minor to something very precise. I assume not something precise, but some fudged approximations. They can't be measuring your ear anatomy, which would be the precise way to EQ if they figured out ear geometry to target response correlation. Doesn't Sony have something that asks for an image of your ear?
In addition, they couldn't possibly measure the response as the eardrum so how do they know what precisely should be EQ'd?

Can you explain to me how EQ is done with pressure?

Anatomical variations are mostly relevant above 1000-2000hz, below it's mainly a question of seal. I'm not 100% sure as I don't have access to the full article, but I believe that the following image comes from the study linked below :

https://www.aes.org/e-lib/browse.cfm?elib=16877

I don't know exactly how Adaptive EQ works, only that Apple's own marketing mentions that it adjusts both the bass frequencies and the midrange, and that it works "200 hundred times a second" (on the APP), which at face value would be contradictory statements unless what it means is that it samples and analyses at higher frequencies (we'll assume 1000-2000hz) but only effectively modulates the FR curve 200 times a second. I don't know exactly what the sweeps used for FR curve measurements are like, but perhaps this discrepancy between the sampling rate and the number of times the FR curve is adjusted per second may explain this :

https://www.reddit.com/r/oratory199...1?utm_source=share&utm_medium=web2x&context=3

What's quite certain though is that it isn't quite as simple as the usual feedback / feedforward system used in most ANC headphones at lower frequencies .

 

[Ilomaenkimi]

https://www.whathifi.com/reviews/apple-airpods-max

 

[AlwaysForward]

Guys,

I've been ploughing through this thread and have some thoughts:

1. Firstly a shoutout to tkddans who gave us his painstakingly detailed personal listening comparison between the HD 800S and APM. And to those others who have shared their necessarily subjective listening comparisons with the HD 800S, HD95, etc.
2. The fact that the APM may be close in audio quality to the HD 800S, and in some respects easier to listen to, is surely a testament to the APM, considering that HD 800S/Schiit setup is > 3X the cost of the APM, and considering also the vast range of other benefits (for those that value them): from portability to ACN to spatial audio to beautiful integration with Apple devices, not to mention the aesthetics for those they appeal to.
3. Fact is: these are entirely different product categories, entirely different markets and entirely different companies. Some overlap in the first two categories? Of course. And this in no way invalidates the audio quality comparison.
4. Different product categories: wired, open, bespoke, high-end audiophile kit v. an Apple-level volume market BT closed ACN headset.
5. Different markets: this thread testifies to the Venn overlap, but, without having any inside knowledge, let me suggest that since Apple is thought to have sold an estimated 60 million Airpods in 2019 (i.e. in the range of $12 billion of just Airpods), they are targeting to sell many millions of APM at $545 and its anticipated, lower cost sports version. Rumors suggest $329.
6. Different companies: Sennheisser is a fine 75 year old company with an almost unparalleled reputation for many decades of consistent sonic excellence in both the consumer and commercial space. Their total revenues (2019) are approx. $920MM. I would be surprised if their audiophile headphones sell more than in the tens of thousands in aggregate. Apple did around 13 times more in just AirPod revenues than Sennheisser did as a total company across all their product lines and divisions last year. (They don't publish category revenue share). They spend approx. $70MM total in R&D. Apple spent approx. $19 billion on R&D last fiscal year. If they spent it on AirPod headphone development (I'm leaving out Beats here) according to revenue percentage (total $249 billion), they would have spent over $800MM on AirPods R&D, but since it is a new and fast growing product category for them, I'm guessing it was more. $1 billion or more. Just on headphone development.
7. So what is my point? It is this: the future of audio and the future of imaging is significantly 'computational'. And computational takes incredible resources. Apple is uniquely positioned in this respect, increasingly owning and controlling their entire technology stack, down to the core silicon. They also 'own' the premium end of the consumer market globally, to the tune of over 1 billion of the most affluent and the most aspirational customers. And it can amortize its incredible R&D costs over its incredible sales volume and still be highly profitable. It has unparalleled scale.
8. Does this mean that bespoke brands will cease to exist? Not in the short term, but I'd guess that they are vulnerable where they compete on the edge of mass, typically consumer markets; less so, in Sennheisser's case, for example, in niche professional/commercial/industrial markets that probably represent a significant share of their total revenues. A couple of other observations:
9. Computational allows for improvements in existing products. I'd be surprised if Apple doesn't announce firmware updates to the APM in the future that enhance its sound quality and maybe offer more customization.
10. I would not discount the social signifier aspects of the APM. A surprising percentage of business class passengers, pre-Covid, I observed, who used to typically sport Rolex watches had shifted to Apple watches, especially women. I'm willing to bet that APMs will become de rigeur for front-of-the-bus air travel. At the aspirational end of the market, I'd be surprised if APMs don't also become a core component of street fashion replacing Beats and at a higher price point.

Nothing said here is intended to criticize the audiophile/enthusiast headphone market in any way.

Paul

PS: the photography market is somewhat though not entirely analogous, (substitute Leica for Sennheisser and compare, dare I say, with the camera in an iPhone 12 Pro Max for most use/view cases), but this is an audio forum.

PPS: I am writing this at my desk with AirPods on each side of my MacBook (approx 1 meter apart) listening to Sonny Rollins Way Out West (recorded in 1957, re-mastered in 2010). I'm alternating with the APM. Another significant Apple achievement in computational sound. Breathtaking to my ears. Starting with the incredibly authoritative bass, not to mention the live sax. And you can pick up a pair for under $600. BTW, the desk is made of metal which might improve the resonance.

Outstanding post new head-fier. I completely agree on every single point. Something tells me that you know your gear and you know your tech.

 

[Jackalton]

One thing I noticed is that I could increase the SPL closer to my liking when the volume is set at 50-60% if I use the following setting:

Settings -> Accessibility -> Audio & Visual - Headphone Accommodations -> On -> Tune Audio for: -> Balanced Tone -> Strong. Even the Medium setting would be marginally okay.

But, although this setting is supposed to optimize for a “balanced tone”, I find that it changes the sound signature of the APM/APP, mainly boosting midrange/high frequency sounds, and I’m not sure how much it changes the accuracy of the sound reproduction. For the love of me (), I could not find a technical explanation by Apple about what these settings do. Any idea?

It’s applying multiband compression, so yes, it’s profoundly altering the sound by reducing the dynamic range of certain frequencies, not just boosting them, as EQ would.
If you’re only enjoying your APM with Headphone Accomodations turned on, you should rethink keeping them, cos thats absolutely not the solution to ‘fixing’ the frequency response.

 

[malcbo]

Saw the hearing test results in the previous pages and had to give it a go, using the APM of course.
I’m 42 and have always liked listening at high levels (particularly for more bass impression).

Overall, I’m a huge fan of the APM. To my taste, they perform even better than my TH-900 (driven by Chord Mojo) with less harsh in the highs. They’re extremely fun to listen to.
Also strongly impressed by the ANC, transparency and comfort.
FWIW, I don’t consider myself as an audiophile.

 

[erich6]

Apple's patent on their Adaptive EQ (HomePods) only talks about using microphones and pressure sensors to create a field of equal acoustic pressure around the speaker. So they have an algorithm that given an input of variable pressure, will apply an EQ to correct the acoustic pressure differences. Using microphones mounted at the ear cups they can pick up different aspects of the ear like resonances (maybe other things as well) to see what's absorbed and what isn't to help deal with those pretty easily.

This sounds very similar to what AKG does with the TruNote auto-calibrating system employed in their N90Q.

Edit: For something real time I could see a microphone picking up what's being played through the speakers of the drivers. It knows what should be being played. "Volume match" at 1k and subtract the difference, then subtract off the target curve and the headphones probably have a good idea of what's different between the sound played and the wave form that was expected to be played. Adjust EQ accordingly. Moving the driver and its location can cause differences as well which would be accounted for in the next round of Adaptive EQ. Note this edit is speculation as Apple could be doing something completely different.



See above. That's really all you're going to get is the patent on how it works. The actual algorithms used to do the processing are probably trade secrets. But the idea has been done before (AKG, 2015).

When I saw the announcement of this feature on the APMs I assumed the approach they took was not to directly measure the frequency spectrum of the audio but to detect at certain audio frequency bands combined with pressure levels to then determine how the EQ should be varied. I'm guessing they used machine learning techniques to develop inference models correlating these variables to the EQ spectrum and these models run in the H1 chips in real-time at some high enough update rate (e.g., 200 Hz). Note I used "update rate" to avoid any confusion with "frequency" associated with the audio spectrum.

 

[obsidyen]

Saw the hearing test results in the previous pages and had to give it a go, using the APM of course.
I’m 42 and have always liked listening at high levels (particularly for more bass impression).

Overall, I’m a huge fan of the APM. To my taste, they perform even better than my TH-900 (driven by Chord Mojo) with less harsh in the highs. They’re extremely fun to listen to.
Also strongly impressed by the ANC, transparency and comfort.
FWIW, I don’t consider myself as an audiophile.

Nice, I really can't stand amps and cables anymore. I can finally get rid of my TH-610 for a more practical solution. I exclusively listen to Apple Music streaming anyway. Sadly I have to wait for a few more weeks to get my blue coloured APMs... I'm more excited after reading your comparison to TH-900.

 

[tkddans]

Nice, I really can't stand amps and cables anymore. I can finally get rid of my TH-610 for a more practical solution. I exclusively listen to Apple Music streaming anyway. Sadly I have to wait for a few more weeks to get my blue coloured APMs... I'm more excited after reading your comparison to TH-900.

But hey, wait until you try them haha. Maybe they’ll complement rather than replace what you have.

I feel you though. Convenience is very nice indeed.

 

[obsidyen]

But hey, wait until you try them haha. Maybe they’ll complement rather than replace what you have.

I feel you though. Convenience is very nice indeed.

I will probably keep TH-610s, but won't be using them as much, lol. They are the next best thing after wireless headphones, as they can be driven by anything. I can use them when I miss analogue HPs haha. I move around the house a lot with my MacBook Pro so using wireless HPs can indeed be very convenient. I use my AirPods Pro at the moment for this purpose but they can't really give the satisfaction of full-size over-ear headphones even though they are very good as IEMs.

 

[pnoble]

It’s applying multiband compression, so yes, it’s profoundly altering the sound by reducing the dynamic range of certain frequencies, not just boosting them, as EQ would.
If you’re only enjoying your APM with Headphone Accomodations turned on, you should rethink keeping them, cos thats absolutely not the solution to ‘fixing’ the frequency response.

Jackalton,

But if you are using the audiogram setting, shouldn't that simply be matching the frequency response to your hearing? Assuming the correction required doesn't exceed the headroom, requiring compression? I have to say that to my ears the APM sounds great, and even better when properly matched to my hearing.

 

[tombrisbane]

Serving them up some tunes via my TA-ZH1ES, plugged in using the horrible 3.5mm to lightning cables and they’re sounding better than I thought they would wired... it also seems to have fixed the slightly recessed vocals that I had wireless (the audiogram thing also helped with that but wouldn’t be active when wired).

I assume it’s still converting it back to digital and then processing it with its internals so the difference I’m hearing would be the Sony doing it’s thing before outputting?

 

[Amrit-R]

... I really can't stand amps and cables anymore...

This imo is were all this is going, already probably 99% true for the mass market, and now creeping into the enthusiast market: wireless / amp-less headphones.

The more traditional manufacturers either adept, or they will go the way of the dodo.

They could, like Hifiman, introduce an wireless open headphone. And if the investment is too big, work together on a common platform. For Sennheiser and Beyerdynamic, two familiy owned companies, collaboration should imo be obvious.

 

[SilverEars]

This imo is were all this is going, already probably 99% true for the mass market, and now creeping into the enthusiast market: wireless / amp-less headphones.

The more traditional manufacturers either adept, or they will go the way of the dodo.

They could, like Hifiman, introduce an wireless open headphone. And if the investment is too big, work together on a common platform. For Sennheiser and Beyerdynamic, two familiy owned companies, collaboration should imo be obvious.

If they can do a wireless openback with stable connectivity with a good codec, and actually drive the headphone well, I look forward to it, but the reason why we have wired setup is due to DAC and Amp possibilities that improve the sound in various ways. Can amps and dacs of such high quality be miniaturized to fit inside these wireless headphones? It doesn't seem likely given class A amps with high quality PS parts and discrete components there are that are too large, and that's not even counting a quality DAC. Or for others, the sounds they get from tubes. These wireless devices primary importance is of course miniaturation and power efficiency. This objective causes limitations on SQ.

Hope that I see a day some genius finds a solution that blows me away with wireless. This computational bla will be related to EQ and I never found EQ as good as well acoustically tuned headphone With no EQ. So, I don't think traditional or physical methods of tuning can be beat with computation. You really can't tune a headphone with computation and expect same intigrity as physical tuning.

Acoustics is different from other mediums. It is quite physical and mechanical. It's quite complex to control such medium precisely with computation.

With what I'm experiencing with APM, Apple hasn't done much beyond what was already established, just scratching the surface. I was hoping they would really take it to the next level with their research, but they haven't. What we got was a cheezy lookin magnetic case probably mimicking their ipad cover. It had iWatch and ipad in it. Lol