I got asked some question by another member of the forum and also needed to do a head to head between the APM and my B&O H95s to decide which one I would keep. So I sat down today and did some listening and less than halfway through my intended testing I had come to a clear conclusion. In case my testing helps anyone else I thought I might as well publish it here.
TLDR: B&O H95 have a mid-bass problem and are going bye bye.
A note about the author: I don’t muck around with sound modes, settings or EQ very often. I think a headphone (any product really) should sound good out of the box on the standard settings. I don’t want to change settings for every third song to get the perfect sound, I don’t want to be forced into installing a buggy app (which they stop supporting after two years) to make something work as intended. Or having to dig into the settings while waving my phone in the air and doing a treble dance followed by a bass jig to get the “correct sound” out if this “excellent product” that just “needs some adjustments”. If it isn’t good out of the box, it isn’t for me. This is also why I stopped having two phones and have stuck with iPhone over Android. I am getting too old to tinker, and my time is too valuable.
Some listening impressions
I start off by listening to my main rig to get a baseline for what it should sound like (to me), quickly followed by the two ANC headphones, alternating which ones goes first to try and not give one advantage over the other. I won’t be giving a view on the Empyrean unless it is for a very specific comparison, that’d be like bringing a tank to a gun-fight.
The ANC headphones:
All songs played were apple lossless from my iPhone, played via Bluetooth, both headphones were set to default tuning (i.e. no accommodations on APM and standard EQ on the B&O) and ANC on (full) since that is one of the main benefits/purposes of these types of headphones.
Main rig:
Apple lossless files played in iTunes (so the source file is the same) over USB to my RME ADI-2 DAC FS connected via an ultra-low capacitance litz cable (from Custom cans UK) to my trusty Empyrean using the leather pads.
The choice of music - I used a few tracks I know fairly well which I regularly have in rotation and which I normally use to test headphones, amps, speakers etc.
Babymetal - Karate
This track is great for listening to issues with artificial treble due to how it is produced, with an inherent artificial sound which comes from overproducing the music. The track also has a lot of bass and recessed mids so a great track to stress-test V-shaped headphones.
On the H95 the treble is artificial, splashy and littered with artifacts. The bass is slightly muddy but with a decent mid-bass kick. Overall it sounds ok except for the treble as the cymbals are completely off and unnatural.
The APM manages to tame the artifical treble and it sounds almost as natural as on my main rig, which was very impressive. The bass lacks a mid-bass kick but the sub-bass is there and not far off competing with my Empy’s (for this element only). The APM did well with this track, keeping it enjoyable and fun.
Alison Krauss – Living Prayer
A great track for testing sibilance control and high notes from female vocals, as well as seeing how it deals with finer details and reverberations.
On the APM I hear more of the guitar reverberations immediately, Alison’s voice is a bit more recessed in the room and some of the immediacy is removed, but not a lot. Her voice is softer than on the H95, the sibilant S’s are still there but lack the sharpness and metallic tone from the H95, although they may be slightly more pronounced. Since her voice isn’t so front and centre the APM manages to move through the highest pats without it feeling like they’ve cut off the highest notes.
On the H95 the overall focus is clearly on Alison’s voice, her voice sounds slightly sibilant with a metallic edge, in particular in the high notes. The guitars are clear and fairly natural sounding but I don’t get the full image of reverberation from the lower guitar-string plucks, and they stay in the background. On the highest notes from Alison the H95 also does not let her voice resonate fully making it feel cut-off and a bit supressed, yet with a metallic edge. Overall it does sound good, but is not my favourite rendition.
Devin Townsend – Deadhead (Live from the Royal Albert Hall)
A great track recoded live in one of the greatest music venues on earth. The mix is difficult to get sounding right, and prefers speakers over headphones, so a good test for enjoyment.
Wow the mid-bass hump on the H95 really comes through here and plasters mud all over the soundstage. The main thing I hear is the kickdrum, and it isn’t sounding good. The cymbals are fairly controlled here, not having the trademark H95 issue, but I really need to listen out for them, and Devin sounds slightly recessed. One positive element I can note is the crowd does sound fairly big and during a lull in the music you do get a bit of a feeling of being there, but then that kick-drum comes in again. I stopped part-way through, I think that says it all.
On the APM the scale of the crowd is still here, and the soundstage might even be slightly better, but not by much. The kickdrum comes in and is very much still a big part of the mix but it now re-enforces the tempo of the song instead of taking it over. The sub-bass adds rumble and the whole recording sounds much more enjoyable, even if the mids feel a tad thin. Devin is still a bit recessed; the cymbals are clearer and might even have a bit more detail, including some which I’m not sure I heard on the B&O. It isn’t quite where I would like it (this track sounds amazing on my HD800S), but it is a good representation.
Eagles – Journey of the Sorcerer
Great instrumental track which lets you listen out for overlapping reverberations and echoes. And a great track to boot!
I’m surprised by the immediacy on the APM, I’d have expected it to be more recessed than the Empys. The reverberations are significantly reduced, all the main elements are here and sound fairly correct, but the reverberation and scale isn’t quite here. The drums lack a little bit of the bite as well, but overall it sounds great and better than expected.
On the H95 it feels like I get a bit more of the scale and echoes than I did on the APM, it feels slightly larger, but the reverberations are missing just like on the APM. The drums don’t sound great however, they are muddy and indistinct. The H95 is doing surprisingly well with the high note guitar plucks, even retaining some reverberation at the cost of a slightly metallic timbre. The cymbals sound excellent here which I wouldn’t have expected considering how poorly it has done on some music in the past. Overall they’ve done a decent job with the drums being the biggest let-down. I think it even managed to eeek out a bit more detail than the APM.
In this moment – In the air tonight
A great cover of the Phil Colins track. Maria’s voice has both strength and frailty which works well with the structure of the song, the significant reliance on bass in the background makes for a good mud test.
The H95 starts out well, with a wide image and scale, but starts stumbling the second the bass kicks in. The balance of the track is way off, it feels immediately closed off and saggy. Normally the bass notes give an oppressive atmosphere which is cut through by Maria’s sharp vocals, but on the H95 all I hear is mud with Maria’s voice hiding somewhere in the middle. In the move to the crescendo of the track it does improve, as it calls for a bit of mid-bass for impact, but it still muddy and indistinct, so not much of an improvement.
Now this is better, on the APM the bass is handled in almost the right way, it is more oppressive and feels like it surrounds you, with Maria’s voice solitary and sharp in the middle of the soundstage. Her voice lacks a little bit of bite overall, but is a clear improvement. The crescendo lacks the proper kick in the bass but gets a fair bit of weight from the sub-bass. Overall a much better showing and clearly helped by the lack of mid-bass hump.
Jinjer – I speak astronomy
A belter of a track with so much speed and varying arrangements, moving from claustrophobic and frenetic to spacious and mellow as well as Tatiana’s switch between clean and growl makes this a great track to test the speed and coherence of your headphones/speakers.
The APM kick off by clearly showing a slight lack in the mid-bass, with a lack of bite from the kickdrums in the opening, but overall it manages to keep the section together fairly well but some details are missing including the normal immediacy and bite. It sounds a little bit like the driver can’t quite keep up with the track in all places, it also lacks some scale in the latter parts, but I would guess the APM is doing fairly well for a closed-back.
The H95 starts off with a lot more thump but also a lot more mud, it’s just everywhere. And it isn’t coming off, I’m starting to think this headphone has a problem with kickdrums. The Cymbals and high hats have a lot of sparkle if you listen for them, and there might be some details from the bass guitar which wasn’t on the APM but it is hard to pick things out of all this mud. It clearly has more scale than the APM in the latter sections, showing off a bit more speed. But oh god the mud!
Katatonia – Unfurl (Live in London)
A nice acoustic set in a church which of course gives enormous scale and so many little details to listen out for.
The H95 starts strong by picking up some minor crowd and ambient noises as well as representing the scale quite well. The bass guitar is given a tad too much space with a very loud strum swallowing up some of the scale, while that is toned down somewhat by the mix after the opening it is still emphasised too much throughout. The guitars have too much of a metallic timbre, while they sound sparkly they are a bit unnaturally so. Jonas’ voice lacks some of the emotion and the track sounds overall lacklustre compared with what I’m used to.
The APM also picks up most of the ambient noises but they are not as clear, and the scale is smaller almost as if you were in just about any room. However, the APMs have the bass guitar well under control, allowing the percussion and guitars to come through. The guitars are nice and sparkly with much less of the metallic edge from the H95. I can hear a lot more details from the instruments and the emotion is back in Jonas’ voice, even if it isn’t getting all the way.
Conclusion
Anyone want to buy a pair of B&O H95?
On a more serious note, the H95 has shown itself to a have real mid-bass issue. It presents itself clearly and matched the first impression I had when I first got the headphones. This was something I tried to justify away, I think in large part due to their extortionate price (for a ANC headphone) and the fact I managed to get one of the numbered pair. But since buying them in September I haven’t actually listened to them that much.
The pandemic has meant I don’t get as much time with my over ear ANC headphones (normally worn when commuting), and I tend to wear my APP on my morning walks as the ANC isn’t wind sensitive unlike the H95. The APMs might change that since trying this morning, not only did they deal with the wind noise, they also fit under my hood when it started raining.
The AMP are clearly a better sounding headphone (to me) than the H95 with standard settings and ANC on. It also has better integration to the Apple ecosystem which over the years I have become fairly invested into. So it is very likely the H95s will go sometime soon, as I can’t justify owning a pair of £700 headphones I’ll never listen to. Even if they are limited edition and uniquely numbered.
The APM are not perfect, I’d like a bit more midrange and they could shed a few grams to be honest (fatty!), and I hate the fact they have a lightning port instead of USB-C. But they are (for me) the better of the two ANC headphones so will get to stay. They will not replace any desktop setups I have, and are not the best things since sliced bread. They are a solid first (over ear)headphone from Apple, and I’m a happy owner who looks forward to trying the next iteration in a couple of years.
Hopefully that was helpful for someone other than just myself.
.
