Sennheiser HD 820 Review (headphone)

[TulseLuper]

Do they leak much sound?

More than expected, but not much. At moderate volume, my wife can hear from 8-10ft away that I'm listening to music.

 

[617]

More than expected, but not much. At moderate volume, my wife can hear from 8-10ft away that I'm listening to music.

Does she complain about the midrange dip?

 

[TulseLuper]

The adage of "you get what you pay for" only applies in the face of technical evidence.

Otherwise you are only paying for somebody's marketing budget...Which is "most of the time".

Have we seen a $1K+ headphone reviewed here for which the price was justified by the out-of-the-box frequency response graph? Not suggesting the $2.4K list on these is justified, it just seems to me that in this segment, we hope for something unique in the trimmings.

 

[TulseLuper]

Does she complain about the midrange dip?

In her words, "What the fuck happened at 400Hz? I hope you only paid $500 for those." She also said they'd look nicer with a goldfish in the earcup.

 

[solderdude]

She also said they'd look nicer with a goldfish in the earcup.

 

[LTig]

Just a personal subjective option, so take it with a bucket of salt or stop reading right now.

About 2 years ago I audioned a few headphones at my local dealer: the HD820, some Focal and other brands. A HD800s served as my reference (I own a HD800). I auditioned them connected to my smartphone, a Samsung S5 neo.

The HD820 was the only one which came close to the HD800s wrt tonality, soundstage and bass quality. The soundstage was not as wide and the bass was stronger but not as clear (more quantity, less quality). Compared to the HD820 all other headphones had even more bass with even less clarity. I need to state that I'm not a fan of closed headphones at all (don't like overblown bass) and at that time the H820 would have been the only one I could live with, but not for this price. Now I own a K371 ...

 

[Robbo99999]

Just a personal subjective option, so take it with a bucket of salt or stop reading right now.

About 2 years ago I audioned a few headphones at my local dealer: the HD820, some Focal and other brands. A HD800s served as my reference (I own a HD800). I auditioned them connected to my smartphone, a Samsung S5 neo.

The HD820 was the only one which came close to the HD800s wrt tonality, soundstage and bass quality. The soundstage was not as wide and the bass was stronger but not as clear (more quantity, less quality). Compared to the HD820 all other headphones had even more bass with even less clarity. I need to state that I'm not a fan of closed headphones at all (don't like overblown bass) and at that time the H820 would have been the only one I could live with, but not for this price. Now I own a K371 ...

And K371 is probably a better headphone than HD820 as long as you can live with less soundstage! It's certainly better at stock, and it has lower distortion in the bass, but slightly more distortion in the treble. So if you think that increased bass distortion and a better soundstage are worth an extra ~$2200, then HD820 is your headphone, or get the Dan Clark Aeon RT closed for $500 for both soundstage and fantastically low distortion!

 

[maverickronin]

Dan Clark Aeon RT closed for $500 for both soundstage and fantastically low distortion!

I have an Aeon 2 Closed with the perforated pads (so it's basically rouge instead of a noire) and it's certainly one of the top current production headphones, but it doesn't touch the 820's soundstage.

 

[Robbo99999]

I have an Aeon 2 Closed with the perforated pads (so it's basically rouge instead of a noire) and it's certainly one of the top current production headphones, but it doesn't touch the 820's soundstage.

That's not the same headphone that I mentioned, Aeon 2 Closed is different to Aeon RT closed that I mentioned, so I don't think your soundstage comparison is valid when trying to extrapolate to the Aeon RT Closed.

 

[Jimbob54]

That's not the same headphone that I mentioned, Aeon 2 Closed is different to Aeon RT closed that I mentioned, so I don't think your soundstage comparison is valid when trying to extrapolate to the Aeon RT Closed.

I don't think any of the DCA models are going to win awards for soundstage tbh. Neither my RT open or cx closed are great for that. Doesn't extend out much from my head.

 

[Robbo99999]

I don't think any of the DCA models are going to win awards for soundstage tbh. Neither my RT open or cx closed are great for that. Doesn't extend out much from my head.

Well I'm mostly going from Amir's review, he said soundstage was very good (or he said it in one of his comments in that thread.). I won't recommend that headphone so often then if the soundstage is not remarkable for a closed back headphone......I had gotton the impression that it was.
EDIT: he said soundstage wasn't as good as HD800s, but it seemed it did have "good" soundstage, but hard to tell, this is Amir's post:
https://www.audiosciencereview.com/...-rt-review-closed-headphone.19393/post-639500

I know that some closed back headphones can have better soundstage than open backed headphones, for example my NAD HP50 has better soundstage than my Sennheiser HD600, after EQ of both to Harman Curve. Perhaps I should recommend NAD HP50 more than the Dan Clarke given I've not listened to the Dan Clarke, and I think the soundstage of the HP50 is good. Pity the HP50 is no longer in production.

 

[peniku8]

You can be 100% sure you won't be playing music with 4kHz at 114dB SPL. Consider that you would be around 94dB when playing loud music.
2nd harm. being 0.4% in a very narrow band is not going to be audible at those levels.

That didn't answer my question. I have no interest in these headphones, I was just wondering about the general logic behind this statement (the relation between THD and our loudness perception) and why Amir chose to word it like this.

 

[solderdude]

Distortion at 94dB is below Amirs own 'limit' and 60dB below the signal and only 2nd harmonic as well. I don't see it as an issue. Just like the 60Hz artifact is pad bounce the issue at around 5kHz may well come from the driver or the resonator.

I would not call -60dB 2nd harmonic an issue.

Furthermore the 4kHz peak does not cause a higher 8kHz peak but rather a 2kHz peak will have a higher harmonic (4kHz). As higher harmonics aren't shown we can't say if the 4kHz also is 'triggered' by lower fundamentals.

 

[maverickronin]

That's not the same headphone that I mentioned, Aeon 2 Closed is different to Aeon RT closed that I mentioned, so I don't think your soundstage comparison is valid when trying to extrapolate to the Aeon RT Closed.

No it's not the same, but it's a very similar design.

I haven't heard the RT, but my remarks apply to original Aeon closed which I have head and are extrapolated to the RT with good reason.

IME tweaks to the FR may change the shape of the soundstage, but the biggest changes in overall size and spaciousness come from changing positioning of the driver relative to the outer ear. With closed headphones another concern for soundstage is diffusing or absorbing the driver's backwave but all the closed DCA's I've heard are already excellent in this regard. As all the Aeons have the same size driver held in the same relationship to the outer ear so no major changes are to be expected.

The soundstage of the Aeons falls into the "good for a closed back" category. The HD820's soundstage is bigger that most open backs.

 

[MayaTlab]

IME tweaks to the FR may change the shape of the soundstage, but the biggest changes in overall size and spaciousness come from changing positioning of the driver relative to the outer ear.

But that's basically tweaking the FR... .
Albeit, presumably, in a way that could possibly better match how your ear's anatomy modulates the FR it receives from natural sound sources or speakers. That's the whole presumption underlying Rtings' PRTF tests.
I'd like to see more tests done to confirm this hypothesis but it's highly impractical to do so I guess.
Also, it's only concerned about the FR curve differential between pinna vs. no pinna and doesn't say anything about the "basal" FR curve, if it's rubbish to start with I'm not sure that the end results would amount to anything.
It's interesting to note that angled drivers don't always necessarily perform better in their PRTF tests than non angled drivers (for example HyperX Alpha vs Alpha S). According to them "As a rule of thumb, headphones with angled drivers and large and deep enclosures have the best PRTF responses.", but I'm not sure that they've actually made a statistical correlation yet and there are plenty of non-angled drivers that perform well as well. https://www.rtings.com/headphones/tests/sound-quality/passive-soundstage

 

[maverickronin]

But that's basically tweaking the FR... .

Yes it mostly comes down to FR, but...

Albeit, presumably, in a way that could possibly better match how your ear's anatomy modulates the FR it receives from natural sound sources or speakers.

It's nothing you can replicate without taking personalized measurements of your own HRTF. Even then, you'd basically need convolution between a given headphone's response on your head and your natural HRTF.

I've never actually heard something like that, but the theory is sound, and for example, people report nearly universal praise for the Smyth Realiser.

Personally, I use crossfeed and EQ to a diffuse field target. I used to use a HRTF simulation VST plugin which was better, but I got tired the sound changing with sources I couldn't run through a VST host so I switched to something which isn't quite as good, but is more consistent. Between my ADI-2 DAC and Rockbox on my DAPs, most of my sources sound the same now.

I'd like to see more tests done to confirm this hypothesis but it's highly impractical to do so I guess.

Definitely. All my comments on the subject are based on my personal modding and DIY experience. There seems to be very little research or commercial interest in the soundstage of stereo mixes over headphones. In another thread Sean Olive basically dismissed it out of hand saying one should get an Atmos system instead. As if people are using headphones in situations where they could just take them off and switch to a 12+ channel surround system.

Even then we'd need a revolution in upmixers since 99% of music is stereo to begin with.

 

[Robbo99999]

No it's not the same, but it's a very similar design.

I haven't heard the RT, but my remarks apply to original Aeon closed which I have head and are extrapolated to the RT with good reason.

IME tweaks to the FR may change the shape of the soundstage, but the biggest changes in overall size and spaciousness come from changing positioning of the driver relative to the outer ear. With closed headphones another concern for soundstage is diffusing or absorbing the driver's backwave but all the closed DCA's I've heard are already excellent in this regard. As all the Aeons have the same size driver held in the same relationship to the outer ear so no major changes are to be expected.

The soundstage of the Aeons falls into the "good for a closed back" category. The HD820's soundstage is bigger that most open backs.

You sound very certain of what creates soundstage. My understanding is that it's not really known what creates soundstage as it seems to be a mix of factors, frequency response certainly influences it heavily, and anecdotally angled drivers and/or angled pads along with large earcups. But no, your assertion that you can say that Aeon RT Closed has the same soundstage as your Aeon 2 Closed is nonsense because small changes make big differences, they're not the same headphone, it's as simple as that. You'd have to directly compare Aeon RT Closed with HD820 to make valid comments on soundstage differences, not listen to the Aeon 2 Closed and then infer it sounds the same as the Aeon RT Closed, that's just nonsense.

 

[MayaTlab]

It's nothing you can replicate without taking personalized measurements of your own HRTF. Even then, you'd basically need convolution between a given headphone's response on your head and your natural HRTF.

Is a full-blown HRTF profile necessary though for music stereo recordings ? My feeling would be that if one simply wants to hear their stereo record in a way that is closer to the experience one has with speakers, then perhaps only trying to match how a pair of stereo speakers would measure at one's own eardrum would be enough ?
Interestingly, Rtings's PRTF measurements are based on a speaker at a 30° angle relative to the ear : "We picked the loudspeaker PRTF at 30 degrees as the reference, because that’s where loudspeakers are positioned in a stereo setup".
But that brings other questions to my mind if the ultimate goal is to re-create a speaker experience with headphones : it seems incomplete to only measure the right speaker with the right ear. Since it also hears the left speakers, shouldn't how the right ear hears it also be characterised ?

Personally, I use crossfeed and EQ to a diffuse field target. I used to use a HRTF simulation VST plugin which was better, but I got tired the sound changing with sources I couldn't run through a VST host so I switched to something which isn't quite as good, but is more consistent. Between my ADI-2 DAC and Rockbox on my DAPs, most of my sources sound the same now.

I'd like to experiment more with crossfeed but every time I've done so it simply seemed to collapse the stereo width and failed to push forward the performance. Well, that's sort of what it's designed to do, ie collapse the stereo width to some degree, and might simply be not enough on its own.
My feeling / thinking would be that if I manage to EQ well enough a pair of headphones to something that would be more akin to how I perceive my near-fields, then perhaps crossfeed would be more convincing ?
Going back to the question of measuring the two R/L channels for each ear, wouldn't the ideal crossfeed be able to use some form of data on what the right ear's FR curve with the left speaker is ?
Perhaps we'd need four EQ profiles, two for each ears with the two L/R speakers characterised ?

Anyway, for now I'm taking baby steps, I'm already quite content to feel that I'm getting somewhere with reliably and efficiently EQing headphones below 1khz thanks to in-concha mics and a few other techniques.

 

[maverickronin]

Is a full-blown HRTF profile necessary though for music stereo recordings ? My feeling would be that if one simply wants to hear their stereo record in a way that is closer to the experience one has with speakers, then perhaps only trying to match how a pair of stereo speakers would measure at one's own eardrum would be enough ?

You just need one measurement cycle for each channel so obviously stereo is just two.

Taking the measurements themselves isn't hard, but I think for most people having access to a reference level speaker system and room is. Otherwise you'd just the worst of both worlds.

I'd like to experiment more with crossfeed but every time I've done so it simply seemed to collapse the stereo width and failed to push forward the performance. Well, that's sort of what it's designed to do, ie collapse the stereo width to some degree, and might simply be not enough on its own.
My feeling / thinking would be that if I manage to EQ well enough a pair of headphones to something that would be more akin to how I perceive my near-fields, then perhaps crossfeed would be more convincing ?

How crossfeed is perceived seems to vary from person to person. I find it almost always moves the stage forward to some degree and that degree is usually improved by EQing to diffuse field-ish target curve. Even if it didn't, most headphones and music give me headaches without it, so for me it's an absolute requirement.

Going back to the question of measuring the two R/L channels for each ear, wouldn't the ideal crossfeed be able to use some form of data on what the right ear's FR curve with the left speaker is ?
Perhaps we'd need four EQ profiles, two for each ears with the two L/R speakers characterised ?

Yes. "Traditional" Meier/Cmoy crossfeed is a comparatively blunt instrument. Something like the Smythe Realiser or Impulcifer will play from one speaker at a time and then record the impulse response for that speaker in each ear. Each input channel is split, convolved for each ear, and then all left ear responses and all right ear responses are mixed together to make the 2 channel binaural.

 

[maverickronin]

You sound very certain of what creates soundstage. My understanding is that it's not really known what creates soundstage as it seems to be a mix of factors, frequency response certainly influences it heavily, and anecdotally angled drivers and/or angled pads along with large earcups. But no, your assertion that you can say that Aeon RT Closed has the same soundstage as your Aeon 2 Closed is nonsense because small changes make big differences, they're not the same headphone, it's as simple as that. You'd have to directly compare Aeon RT Closed with HD820 to make valid comments on soundstage differences, not listen to the Aeon 2 Closed and then infer it sounds the same as the Aeon RT Closed, that's just nonsense.

I know what changes are likely to make what kind of differences.

You might as well say that one could not infer anything useful about the HD650's soundstage from listening to an HD600.