Sennheiser HD800S Review (Headphone)

[xykreinov]

This is ridiculous.
I still listen to real life with the same ears I use to listen to my hifi.
I have seen this sort of comment before about compensating for age related hearing loss and there is not even a tiny glimmer of logic to it.
My reference is the outside world and the many, many more concerts I have been to since I was young and, hence, a much keener knowledge of what real acoustic music, as opposed to music from speakers, actually sounds like.
Any relationship with my age is more knowledge and experience. When I was young and had pretty well only heard music over speakers, and almost exclusively pop music at that, I am sure I would have liked the preference curve Harman have assembled from a lot of listeners.
Now I am very much more knowledgeable and experienced I definitely do not.
Talking preference back in the day nearly everybody I knew had the "loudness" contour on and/or the bass turned up. "Preference" IME has little to do with accurate reproduction.

Why do you think how well you hear and quantify sounds in "real life" outside of headphones proves anything about your hearing range? Do you think it's easy to figure out on your own when you have recesses from 15kHz up? Yes, you know what real, live music sounds like from concerts. But, what does that have to do with your hearing range? In relation to the above, everything is relative.
Also, though I don't know what alternative you have in mind, due to the nature of the human ear's pina, we currently have no better way to quantify accurate reproduction in headphones in headphones than with target preference curves.

 

[magicscreen]

In general yes. So the approach in research is to take one headphone and EQ it to the other headphones under test. Then listeners wear that one headphone and EQ is changed to simulate other headphones and preference score measured. Testing against the real though has shown something like 85% similarity so the method is thought to be reliable enough.

Cannot we record it using a microphone? Then the recorded files can be ABX tested.

 

[DavidMcRoy]

I don’t envy you for diving into headphone measurement. I used headphones in broadcasting as required when speakers were inadequate or impractical to use. As for subjective evaluation, when randomly listening testing headphones at BestBuy, I struggle to find a pair with a natural bass balance. They’re mostly unbearably bass-heavy and boomy.
This is true even when you defeat any bass boost feature that may be present. How about a “neutral bass“ switch? I also struggle to find adequately low midrange distortion. Some highend Bang & Olufsens I tried once were nice. At home I have used some Plantronics and most recently HiFiman HE400i 2020s which I like very much.

 

[Erispedia]

Hi,

Cool

The Qudelix 5k : that is the BT receiver that allows to turn a wired HP in a wireless one is it not ?
Or can it also be used as a DAC and pass over the sound to something else ?
Because with those small things I am always somewhat short of power... Even though I do not listen to music very lound and even though my HP aren't the most difficult to drive...

But I was interested in a thing like that indeed.

So may give it a try

You can always chain it to another headphone amp if you want. The DAC section of Qudelix is not that special though.

Hi amirm,

Thanks for taking time to reply

But if you can simulate the other headphone and have 85% similarity with the other headphone, then there's no real reason to choose a headphone for other reasons than comfort and features and build quality... and price maybe.
Provided of course, you can EQ your headphone properly, but that's another story

Or did I misunderstood what you wrote ?

If you only care about tonality in the way a headphone sounds, then yes.

 

[xykreinov]

I think you're wrong for two reasons:

1. The members here like them and there is plenty of good, informed discussion. We are likely to get more, not less, discussion of this kind because of the community and the ideas supported here. That Amir's headphone reviews are less firm and require more introspection on his part is not a sign that he or we are lapsing from the established approach. It's just new territory.
2. Related, just like the SINAD and preference score debates, we can't keep arguing on the part of imaginary masses who will, apparently, see no difference in the approach here and elsewhere, and will further jump to conclusions without understanding or looking further at the data. If we do see evidence of that, I would think our goal is to clarify, set the proper context, provide references and so forth. In other words, we should not conclude that the overall approach is inherently wrong because there is a possibility of misinterpretation.

Personally, I'd get into headphone testing out of curiosity, no matter the state of the field, for the very reason because that is somewhat of an unknown. It means there are still things to discover.

On top of that it's possible that someone might offer up gear like the Smyth Realizer, and Amir could test both the electronics and headphones. Maybe even extending this to plug-ins. In headphone listening these spatializers are as thoroughly part of the discourse and market as EQ.

I wanted to say something similar to this. Wanting to forgo reviewing headphones on grounds that headphone measurements & targets are less scientific than speaker measuremens & targets is in poor faith. It's like saying the study of physics should have stopped after the apple fell on Newton's head because whatever lied beyond was much more incomprehensible at the time.

 

[Tks]

Interesting findings @amirm, thanks for the effort. Is there a way to measure that extra layer of transparency which sets it apart from the 650s for instance? I believe your subjective assessments though.

Yes, massive sized cups, and massive highs native from the awful driver tuning.

Take the HD600 (or any of its brothers), stick it into the shape of an HD800, I'd guarantee people would say the same as they've been saying all along.

I've tried this headphone out for a while. Comfort wise, every other headphone feels like it's made by idiots (aside from the Meze at double the price). The sound is absolutely idiotic though. Without EQ, it's yuck, with EQ and some volume, the thing distorts so bad in the lows as these measurements confirm. The entire Sennheiser line is riddled with this issue of just weak bass that takes the headphone to shit the moment you try and compensate for loud listening. (yes yes, I keep my "pre amp gain" lower than the highest equalized frequency, so it's not user error sparking this)

The problem with this headphone extends to even the highs, where you have that disgustingly sharp rise.

Must be all 90+ year old golden ears at the company. There's no way something like this gets through QC, without purposefully going ahead with it.

Luckily for Sennheiser, most people seems can't hear highs well it seems.

And also luckily for them PEQ exists.

This just shows how good the 600 line is. And how not would entirely replace the 800 line out of existance, if they ever dared to make a cup size/shape similar to the 800.

 

[amirm]

Or did I misunderstood what you wrote ?

You did not. It is something we need to replicate and evaluate.

 

[wasnotwasnotwas]

Hi,

One question (and it is serious).

As amirm says, the HD800s should be EQ'ed as any other headphone.
Let's assume that I get a headphone that is decently designed and engineered.
If I proceed with nice EQ'ing, will it not sound more or less the same as the HD800s ?

So my question is : will 2 properly EQ'ed headphones still be distinguishable in sound or not ?
Like : will you be able to tell them apart in a properly performed double blind test ?

Regards.

Hi,

So if I have a headphone that suits my listening tastes without EQ'ing it, then I'm good to go with that one ?

That is cool as I think I just found THE headphone that I like over all others I've listened to.

Regards.

What is THE headphone?

 

[pavuol]

It is a matter of design philosophy. I read that Sennheiser calibrates their headphones against the "diffused field." They set up 8 speakers in anechoic chamber and measure and match what arrives at the "ears" of the measurement fixture. No real room has such a sound field yet no one questioned this until Harman came around and made the reference a good speaker playing in a room.

That aside, I can't understand why their HD650 and HD800S measure differently. Surely if they are all designed to comply with that diffused field, they would have similar response but they do not. So it is a puzzle. Likely it is the "sound they like" as they designed it.

Anechoic? Cannot find a web source about Sennheiser but Beyer states echoic (with ceiling treatment only) for their diffuse field method.

 

[BrokenEnglishGuy]

What? Did you read the review I wrote? I said it is only recommended with equalization. It is NOT recommended otherwise.

And no, there are plenty of speakers that are worse than this. Find me one that goes down to 20 Hz with equalization as this headphone does.

Come on, the 20hz from hd800/800s doesnt sound good because the driver doesnt move air! You need a very nice driver because the drivers are small, planar magnetic are much better in bass... The driver of 800s can be big but planar magnetic are easy much bigger and by nature can reproduce better bass, look at the hi end planar magnetic they are fast and clear as hell
For example the focal clear have surround wich allow to have more movement, in headphones the measurements doesnt say the whole story, if you do the eq to hd800 you will lose the details and speed because you boost a lot the bass and turn the highs down so the perception of details gonna turn down, even the 800 Is better than 800s if you only gonna pick them with eq



There is so many in ear can reproduce 20hz flat but you need to move air! A 20hz -10dB gonna sound better than a 20hz from headphone in most cases, the bass from speakers are far more dense..
In many headphone you can hear the 20hz notes but you dont gonna feel the bass, planar magnetic do this task much better, also planars are more clean

 

[amirm]

Come on, the 20hz from hd800/800s doesnt sound good because the driver doesnt move air! You need a very nice driver because the drivers are small, planar magnetic are much better in bass... The driver of 800s can be big but planar magnetic are easy much bigger and by nature can reproduce better bass, look at the hi end planar magnetic they are fast and clear as hell

Please watch your tone. Comparison was against speakers, not other headphones. Whether planar magnetics do well or not will be subject of future review.

For example the focal clear have surround wich allow to have more movement, in headphones the measurements doesnt say the whole story, if you do the eq to hd800 you will lose the details and speed because you boost a lot the bass and turn the highs down so the perception of details gonna turn down, even the 800 Is better than 800s if you only gonna pick them with eq

In my testing it is easy to get the Clear to severely clip resulting in ticks and pops which the HD800s did not suffer from.

Anyway, your comments are off-topic. This review is about Hd800S. As I noted, until I test others there is no other data to argue about. Your personal impressions are of no value because there are thousands of such impressions. This is why we professionally measure and I listen and evaluate headphones. If all we wanted was your comments then you would be writing the reviews, not me.

 

[bigjacko]

I thought HD800 and HD800s have very low distortion and probably one of the lowest. HD650 also got low distortion but I think HD800s should still win in lower mid to bass. In DIY audio heaven HD800 has better distortion compare to HD650, although HD800s has a bit more distortion at bass compare to HD800. @solderdude what was the SPL you measured those and what is your opinion on this?

 

[bobbooo]

Yep, even when EQ'ed properly in tonal balance the HD800(S) is simply a step above most other headphones. The imaging and detail retrieval have a sense of realism that the vast majority of headphones simply doesn't reach. I like to call this 'effortless' sound.

I like to call this large, angled earcups that maximize pinna activation (as Rtings have impressively attempted to quantify), in combination with listener pricing bias, and leave the nonsense Head-Fi / YouTuber vocabulary where it belongs. You don't need to be robbed of over a thousand dollars for a 'TOTL' headphone to get this 'effortless' sound (read: high pinna activation) - plenty of 'entry level' headphones offer this, such as the AKG K702, or the HifiMan HE4XX at almost a tenth of the price, the latter not only having lower distortion (beneficial when EQing up the bass to the Harman target) but even following the $60,000 Sennheiser HE-1's frequency response more closely (in fact the closest I've seen in the mids and treble):

This gives the HE4XX a Harman predicted preference rating of 88/100, just two points below the HE-1's 90, and five ahead of the HD800S's 83. I don't know how many more times it needs to be shown that price is not a good indicator of sound quality when it comes to audio reproduction equipment.

 

[gatucho]

Hello everyone,

I just registered to post to this very polemic thread. I've been in this hobby for around two decades now, I enjoy both headphones and speakers as well as some amateurish headphone measurement.

1) It is true that headphones may be more difficult to objectively assess than speakers; however, even for speakers there is a lot which is relevant in addition to FR. In headphones this is also true, you can have two headphones with very similar FR and they can sound clearly different in other aspects. Many suspects are possible such as transients, driver technology (BA drivers just sound different to dynamic drivers for instance), distortion, etc

2)Regarding EQ: I have found that even the best headphones can benefit from an SLIGHT level of EQ. In this regard, interestingly I was looking at my own HD800 EQ settings and found them to be very similar to those of @amirm I find some consistency in this. The level of EQ that a headphone can accept before sounding "bad" (which I wouldnt know how can be measured) is also frequency dependent. Typically higher levels (say 6dB) are acceptable for low frequencies while mid/high frequencies respond better to lower levels (up to 3dB)

3)If someone (such as all of us) is willing to pay thousands of bucks for the best sound available, it is just unjustifiable not paying a little extra for EQ software. Purists may disagree and it would be their loss. This is a performance race, not a purity test!

Considering all of that, at least to me, a good headphone would be one which can achieve an acceptable FR with a SLIGHT level of EQ and which has all those other "nice" (non-measurable?) properties such as sound-stage, layering, etc. In this regard I also consider the HD800s as a very good headphone.

The truth is that good headphone review, to me, has to be a mix of these two (objective and subjective) aspects. Believe me, I would prefer that subjectivity wouldn't show it's ugly head (my day-work is as a scientist!), but I believe that even if the data is there in the measurements its interpretation is still lacking in many aspects.

I applaud @amirm for his efforts and his openness to critic and change. He is different in this regard than most in this hobby, and I believe that's the reason why he has gained so much respect in so little time.

 

[preload]

In general yes. So the approach in research is to take one headphone and EQ it to the other headphones under test. Then listeners wear that one headphone and EQ is changed to simulate other headphones and preference score measured. Testing against the real though has shown something like 85% similarity so the method is thought to be reliable enough.

Hi Amir, thanks for the review. I looked up the experiment where the Harman group demonstrated an r=0.85 correlation between virtual and actual headphones (Olive et al, 2013). My personal opinion is that the work isn't quite as compelling as I had hoped. It looks like only 6 headphones were tested. The mean virtual vs actual preference ratings were "okay," but 1 out of the 6 had marked score differences (HP5), and a couple of pairs had flipped preferences.

What's worse is that looking at individual listener performance, 3 out of 7 had pretty poor performance. Apparently Listener 58 didn't undersatnd the instructions, so that means for 2 out of 6 listeners, the correlation between virtual and actual HP was pretty bad.

Honestly, I think this matches everyone's own experience - in that, if it were possible to simply EQ a pair of headphones to sound like a TOTL model, why not turn a pair of HD650's into HD800's with software?

 

[Tachyon88]

I listen to both stock and EQ. Sometimes treble can be a bit hot on certain tracks. I just flip the stock OT1990 EQ settings and it takes the edge off the hot treble and ads a bit a warmth in the lower end. To me the detail is still there and the changes are more nuanced than dramatic. For speech though, I do have to use EQ, the TSSS can come in sharp.

I think these are over priced though, like all TOTL HP. I like them more than my HD6xx, which I listen to daily as well.

 

[Joachim Herbert]

Probably a ton of people but I am here to change that. We need to move people to usage of EQ.

I beg to differ. Each and every eq setting points to a design flaw, essentially. We need to move vendors to deliver proper designs.

Thats how it works with electronics (see schiit), and how its supposed to work for speakers.

 

[Doodski]

I beg to differ. Each and every eq setting points to a design flaw, essentially. We need to move vendors to deliver proper designs.

Thats how it works with electronics (see schiit), and how its supposed to work for speakers.

I'm not sure that transducer technology is good enough to create flat a frequency response output. Imperfections are the norm across the board. Otherwise they would all sound the same.

 

[Erispedia]

I beg to differ. Each and every eq setting points to a design flaw, essentially. We need to move vendors to deliver proper designs.

Thats how it works with electronics (see schiit), and how its supposed to work for speakers.

What is proper tonality or tuning? And why it's called proper?

 

[solderdude]

I like to call this large, angled earcups that maximize pinna activation (as Rtings have impressively attempted to quantify), in combination with suffering from pricing bias, and leave the nonsense Head-Fi / YouTuber vocabulary where it belongs. You don't need to be robbed of over a thousand dollars for a 'TOTL' headphone to get this 'effortless' sound (read: high pinna activation) - plenty of 'entry level' headphones offer this, such as the AKG K702, or the HifiMan HE4XX at almost a tenth of the price, the latter not only having lower distortion (beneficial when EQing up the bass to the Harman target) but even following the $60,000 Sennheiser HE-1's frequency response more closely (in fact the closest I've seen in the mids and treble):

View attachment 99577

This gives the HE4XX a Harman preference rating of 88/100, an insignificant two points below the HE-1's 90, and five ahead of the HD800S's 83.

Thanks... you're totally correct.. what was I thinking ?
I'll sell my HD800 and from now on will listen to my (Oratory EQ'ed of course) K702 that sits in its box, obviously for no apparent reason.

After all I have to get rid of my pricing biased, elitist, youtube/headfi line of thinking. I might even buy the HE4XX believing it is very close to the HE1 at 1/250th of the price because it only is 2 insignificant points below it... without EQ even.. go figure.