[MayaTlab]

Make no mistake, Apple's only desire is to get more money off you. All BT products (or anything with non-user replaceable batteries) are throw away products. A first generation HD-650 is still perfectly usable (a is used) today. How many BT products will be usable after just 5 years of usage? 10 years?

None but that's a by-product of them not being fully passive. That's all. Same applies to any of Sennheiser's active headphones. If them being active means a better audio rendition (something that's very theoretical at the moment but will happen), I don't see why that's necessary a problem.

 

[angelom]

How many BT products will be usable after just 5 years of usage? 10 years

B&O H95 more than 10 years.

 

[SparkOnShore]

Trebles is the one area where measurements are the least reliable, show the highest amount of variation, even with similar tests rigs, and are the least relevant to your own ears as it's where anatomical influence on FR variation at your ear drum is the most important.

Rtings' measurements show a -10dB response at 4800hz. Of all the measurements we've seen so far for the APM they are the outliers in that area. That doesn't make their measurements invalid, but it just shows that particularly above 1000khz you need to be very careful about how you interpret FR curve measurements. And you're over-interpreting them and probably relying too much on Rtings' score (which is a useless, over-interpretation of their own measurements).

Given the eight or so measurement sets we've seen so far from websites whose methodology is reasonably known, we can see constant trends across measurements :
- An excellent response below 1000hz. It isn't because of the exact magnitude of the bass response, where individual preferences vary anyway. It's because it's very smooth with no significant peaks or valleys, something fairly rare with closed headphones.
- a fairly conservative response around 2000-4000hz, but with a fairly benign response : there is no high Q (sharp) peak or dip.
- a lack of constant tendencies in the trebles response above that : we have a rather poor match between measurements in terms of where the peaks or dips are located, let alone their magnitude. This is not surprising as most headphones will show a lot of variance in measurements above 5000-6000hz or so. The only thing that I personally look for in that area with measurements is whether I get constant features or not, such as a 8000hz peak with most Beyedynamics headphones across various measurements. In the case of the APM it seems that peaks and dips in the trebles are less a product of the headphones' intrinsic design and rather a case of interaction with the test rig / listener's ears.

The first two points mean that at least up to 4000hz the Airpods' Max are very easy to EQ to a satisfying degree in that range. In the realm of closed back BT headphones, let alone closed back full stop, this is exceedingly rare. Hence why I think that their curve is very decent to start with.
Above 4000-5000hz it's difficult to know, but that applies to most headphones in general. Most will show peaks or dips that are at least just as significant as the ones measured on the APM, regardless of whether they're closed or open, wireless or wired. If Rtings's measurements above are not decent to you, then I wonder what you'd think of that :
https://www.rtings.com/headphones/graph#12049/4012/1659

As far as I'm concerned, when they're on my own head :
- trebles is probably the one area where I subjectively find the APM's most objectionable, but at least the peaks aren't painful to my ears unlike most other BT over-ears I've listened to last or this year. headphones such as the A50, M3, Bose 700, PX7, etc. I simply can't listen to and certainly can't EQ to my satisfaction given the very high Q peaks they exhibited when mounted on my ears.
- on my head at least, Rting's dip at 4800hz simply isn't there at all, something that's easy to check simply by adding 10dB with a fairly high Q. The result is ear-rape to me. I've listened to headphones with a built-in null, such as the HD350BT at around 4200hz or so. You can hear such nulls by playing single tones and shifting up or down the tonality. No such null on the APM at 4800hz.

All headphones in rtings have been tested under the same terms and the same test rig. So if one is showing better curve than the other it is good to remind that they both have been tested with the same exact rig and methods. I don’t look at their scores, I look at the curve. Below curve is the result of the same rig that tested APM, but look at it. Elevated bass for bass lovers, meanwhile with clear mids and balanced highs. YMMV though since they may not sound good to your ears, but it has nothing to do with their original FR:
https://www.rtings.com/headphones/1-4/graph#1585/3992
Edit: I mean, if you say that treble is difficult to catch on APM, it is difficult also for this headphone, but still, check how more balanced it looks:
https://www.rtings.com/headphones/1-4/graph#1585/4046

 

[MayaTlab]

All headphones in rtings have been tested under the same terms and the same test rig. So if one is showing better curve than the other it is good to remind that they both have been tested with the same exact rig and methods. I don’t look at their scores, I look at the curve. Below curve is the result of the same rig that tested APM, but look at it. Elevated bass for bass lovers, meanwhile with clear mids and balanced highs. YMMV though since they may not sound good to your ears, but it has nothing to do with their original FR:
https://www.rtings.com/headphones/1-4/graph#1585/3992

OK so I'll repeat it again in other words : even on a Harman target compliant rig, you'll find significant frequency response variations relative to the target across various measurements. In other words, Rtings' measurements alone aren't that useful, particularly past a certain frequency. Best use laymen like you and me can make of measurements is to preferably look at several of them, extract the common tendencies, regard with a lot less importance the deviations, and take it all with a pinch of salt, particularly above 1000hz, and even more so above 4000-6000hz or so.

That's the M3 vs APM on Crinacle's rig BTW, vs Rtings, normalised at around 700hz since that's where the curves meet on Rtings' graph :

See how the delta between their FR curve has nothing in common, particularly above 1000hz ? For example at 6300hz or so : Rtings' measurements show a minimal difference in dBs, while You have a difference of around 10dBs with Crinacle's measurements. Even when normalising the curve in Crinacle's graphs to reduce the maximum delta you still get significant variations in the delta between the M3 and APM's curves. So, who's right ?
In other words relying on one single set of measurements, particularly above 1000hz, is not a good idea, and you're over-interpreting Rting's measurements.

 

[alidark3000]

if they do anything like that I hope they upgrade the current model with software... I don’t want to buy another pair for a decade.

 

[SparkOnShore]

OK so I'll repeat it again in other words : even on a Harman target compliant rig, you'll find significant frequency response variations relative to the target across various measurements. In other words, Rtings' measurements alone aren't that useful, particularly past a certain frequency. Best use laymen like you and me can make of measurements is to preferably look at several of them, extract the common tendencies, regard with a lot less importance the deviations, and take it all with a pinch of salt, particularly above 1000hz, and even more so above 4000-6000hz or so.

That's the M3 vs APM on Crinacle's rig BTW, vs Rtings, normalised at around 700hz since that's where the curves meet on Rtings' graph :



See how the delta between their FR curve has nothing in common, particularly above 1000hz ? For example at 6300hz or so : Rtings' measurements show a minimal difference in dBs, while You have a difference of around 10dBs with Crinacle's measurements. Even when normalising the curve in Crinacle's graphs to reduce the maximum delta you still get significant variations in the delta between the M3 and APM's curves. So, who's right ?
In other words relying on one single set of measurements, particularly above 1000hz, is not a good idea, and you're over-interpreting Rting's measurements.

Sorry to ask, I am really not aware, but what makes Crinacle’s measurements as reliable as rtings? Asking since I see rtings references all over but Crinacle not that much. Thanks for your info.

 

[MayaTlab]

Sorry to ask, I am really not aware, but what makes Crinacle’s measurements as reliable as rtings? Asking since I see rtings references all over but Crinacle not that much. Thanks for your info.

I didn't say that they're as reliable. You'll have to ask the experts for that. What I do know however is that he publishes his methodology, has frequently engaged with people to discuss it, and his rig is Harman target compliant (ie you can compare his measurements against Harman's target, which would suggest to me that it adheres well enough to some standard - unlike Rtings BTW).
What I'm saying is that even when people publish and document their measurement methodology, we see increasingly significant differences the higher the frequency and reliability suffers as a result.

The way I tend to look at measurements is that I chop up the frequency range into different areas (the frequencies below are only a rough idea, I'm not that dogmatic about it) :
- below 250hz or so : here we both have significant variations across measurements (because of sealing issues) and user preferences vary quite a lot (according to Harman's research). What I expect to see here, provided the headphones can extend low enough, is something that falls within Harman's research or thereabout, associated with, for headphones with a built-in DSP, a sensible EQ capability, ie one that adjusts the bass shelf so that it encompasses the variation in preferences Harman's research tends to highlight, and that starts at the right frequency.
- between 250hz or so and 1000-1500hz : That's probably where we have the lowest amount of variation both across measurements and also when it comes to user preferences. What I think we can all expect to see in that range, particularly for DSP enabled headphones, even closed ones, is a smooth curve. Perhaps tilted slightly warmer, perhaps slightly brighter, but a smooth curve. There is no justification for weird peaks or valleys.
- between 1000-1500hz and 5000hz or so : here we start to enter the range where an individual's anatomy and variations in testing equipment and methodology start to make quite a bit of difference. I prefer to look at an aggregate of various measurements (as long as their methodology is reasonably detailed), particularly when I can compare the headphones in question to a very well known and widely measured model (let's say the HD650 for example).
- above 5000hz : here humans' anatomy varies enough that variation in dB at your eardrum can be very significant, same for testing rigs. The only thing I care to look for here are either extreme responses or very constant findings across a multitude of different measurements (such as Beyerdynamic's traditional 8000hz peak).
- above 9000hz : I don't even care about measurements.
The last three are made even more difficult because of hearing loss.

When I say that the APM measure decently, it's because in the range we can be most confident with, they measure very smoothly, follow reasonably well Harman's target, and are very easy to EQ. Even slightly above, in the 1000-1500 to 5000hz or so, the aggregate of the measurements we've seen so far with them seem to suggest a perfectible response, but no dramatic peak or null that would make EQing them difficult (with the exception of Rtings' measurements, which are a very clear outlier so far). Above 4000-6000hz ? I don't know. And I'm not confident in measurements to give me a particularly clear picture.

Jude's B&K 5128 presumably models better the human ear than previous test rigs. I'm very curious to see additional measurements made with it, and particularly whether the results above 4000-6000hz could be more reliable, but in that range your own anatomy is likely to make these measurements less relevant anyway, at least to you.

One thing that I'm eager to see from smart, active headphones, is the capacity from them to adjust their FR curve above 1000hz based on your own anatomy.

 

[tinyman392]

Regarding that goal I'm wondering if the ultra-wideband technology Apple has started to implement in their products could be used for audio over wireless communication. I have no idea so if someone knowledgeable reads this please feel free to chime in.
Bluetooth suffers from higher latency than would be desirable, at least for most codecs, and lower data rates than optimal.
Wifi's data rate is alright but latency even higher and power consumption too high for wearable devices.
I'm not sure I understand how ultra-wideband technology works but from what I gather we could obtain Wifi-like data rates at Bluetooth-like levels of power consumption with a decent range. And perhaps low latency when it comes to wireless audio ?

I'm not sure if Apple has plans to go about this. They have their solution for video already (if I'm not mistaken iOS delays video slightly to sync up with the headphones). For something like live gaming (or gaming in general) this would be a lot harder to do though and something that you'll likely never get rid of.

https://www.rtings.com/headphones/1-4/graph#16092/4046

Sorry, but this is not a treble’s decent default curve. Even taken at first ”sitting” without Jude’s special care, it is still not decent at all.

@jude talked about this already, at length... You were there, the fact that you keep talking about this single dip...

Yes, the APM treble is peaky, that's well established. But that dip and spike has been talked about at a very long length (headphone placement). Most other graphs of the APM (Crinacle, Head-Fi, Orathotory or however you spell it) really don't have that huge dip in there, it's likely either an issue with headphone placement or that rig on its own.

Treble readings up high typically will also have weird dips and spikes due to rigs in general (each rig having different ones). I tend to kind of average out huge dips and spikes to get an idea of tonality. But that's about all I do when "reading" treble off a graph as reliability of reading this area is quite difficult in general.

 

[tinyman392]

Somebody discounted someone’s entire impressions between a few headphones (with respect to heat) because of the fact that they just assumed they had sweaty ears (when they in fact didn’t).

Well, you got it all wrong. I wrote that plastic is warmer than leather, as it is well known in science. Plastic is not absorbent, leather is, thus it is cooler. Leather is natural thus adjustable to human skin, plastic is not. So either his info was wrong, or he just sweats on every material as millions other humans. What is wrong with that?? My point is proven, a leather ear pad can be cooler and more comfortable for most humans (not all, but most) than plastic. It is the truth, as simple as that. Common sense.

Wrong, but to each his own! Most probably your are generally a sweaty type...

That's your whole quote for the post... You specifically said he was wrong just blamed the dude saying his ears were sweaty because his impressions did not conform to your world view. This ended up being false after he clarified that wasn't the issue.

 

[yticolev]

Everything else was a master/slave with the right bud being the master.

This was rather jarring to read. As it is when I hear it said out loud amongst workplace peers who are actual descendants of slaves and still suffer from our shared history. I know this idiom is commonplace in many settings, but it really needs to be banished. In my work, I make an effort to say "primary" monitor and "secondary" monitor. Very easy to slip even after some years especially right after a coworker uses the historical idiom. But I try. In some similar instances, "daisy chain" works well and is highly descriptive.

I'm not the PC police (first such comment online in thirty years of forum posting going back to UseNet) but some simple language choices can help make a difference.

 

[Author]

A lower hanging fruit to improve SQ for most people would be to make a few subtle corrections to the APM’s FR curve.
Higher up the tree a properly and rationally designed consumer-friendly EQ UI that takes into account research into consumer preferences variance to provide the right adjustments (and not necessarily a lot) would be even better. Given the APM’s rather decent default curve I think that a bass shelf adjustment starting at the right frequency + a fixed anchor point at 1000hz + two anchor points higher at around let’s say 3000 or so and another one higher up would already be a good start.

Is there an App out there that allows for EQ? I feel happy with the audiogram (and the other functions there), but I wouldn’t mind more flexibility just to play around.

 

[justanut]

And there’s a company who’s desire isn’t to make money? >..<

Apple usually does it while trying to give the best possible balance of user experience, user support and profit for the company.

I can easily bring my old i-products to the Apple store to have the battery replaced reliably.. and if they have reached EOL, Apple products are usually popular enough that I can easily order my own battery from AliExpress or Taobao. Compare that to other brands who may not even have a service provider in your area, plus you may or may not be able to find replacement parts / batteries (depending on their popularity?).

My iMod is still being kept alive because of this. My iPad Gen 1 is still useable (although on the software front it’s GG... I can only web browse...). The APM’s neat design, plus only having the battery on one side makes it easier to self-replace it if ever the Apple Store stops servicing them.

Still, that’s the pitfall of every battery dependent product like portable Amps, DACs, DAPs... I don’t see what this has to do with how good the APMs are or how good of a value they are to the average user. Prob no one on the streets will be able to play the HD 650 if you pass them one today...

Make no mistake, Apple's only desire is to get more money off you. All BT products (or anything with non-user replaceable batteries) are throw away products. A first generation HD-650 is still perfectly usable (a is used) today. How many BT products will be usable after just 5 years of usage? 10 years?

 

[MayaTlab]

Is there an App out there that allows for EQ? I feel happy with the audiogram (and the other functions there), but I wouldn’t mind more flexibility just to play around.

Not one you can save on the headphones themselves . Which is personally what I'd like to see. I use headphones with too many sources including ones I can't EQ.

 

[justanut]

And you know this because they were out before 2010?

B&O H95 more than 10 years.

 

[Author]

As far as I'm concerned, when they're on my own head :
- trebles is probably the one area where I subjectively find the APM's most objectionable, but at least the peaks aren't painful to my ears unlike most other BT over-ears I've listened to last or this year. headphones such as the A50, M3, Bose 700, PX7, etc. I simply can't listen to and certainly can't EQ to my satisfaction given the very high Q peaks they exhibited when mounted on my ears.
- on my head at least, Rting's dip at 4800hz simply isn't there at all, something that's easy to check simply by adding 10dB with a fairly high Q. The result is ear-rape to me. I've listened to headphones with a built-in null, such as the HD350BT at around 4200hz or so. You can hear such nulls by playing single tones and shifting up or down the tonality. No such null on the APM at 4800hz.
- EDIT : that being said, I certainly would love to be able to EQ them up a little bit somewhere around 3000hz.

This could be a gross oversimplification, or really obvious to most readers here, but I wonder if Sony’s heavily veiled XM signature is a response to grating treble/potential for listening fatigue.