Sennheiser HD800S Review (Headphone)

[Rock Rabbit]

So after reading 23 pages, is the science gone from ASR here? Are we back to subjective score boards and spaciousness, clarity, thing? :/
Measurements alone do not tell the whole story? Although curves look alike, the he4xx must be worse than HD800?
Please say no

AudioSubjectivereview
AudioEQeverthingreview
AudioDon'tcareFrequencyResponsereview
AudioSoundstageIsEverythingreview
...

 

[somy]

How does HD800S comparing t o Ananda in terms of Bass? I have ananda and sometimes I hope it has a little more bass.

 

[amirm]

Or if Amir measures the HE4xx before then, then I'll post it in there.

I don't have the 4XX but do have the 400i and out of curiosity measured it yesterday and developed EQ for it. I will see if I can squeeze it in today and post the results but have another review to get out first.

 

[PenguinMusic]

I wonder how many audio players there are for linux. I never came across that when I was looking. Perhaps we were looking for different features. I just wanted something which just played music, and could be controlled via Android, ending up with mpd. I'm not even sure how it's possible to no be able to target ALSA; surely there's some dummy null pulseaudio device which passes stuff straight down to ALSA?

something like this:
https://wiki.archlinux.org/index.ph...io_as_a_minimal_unintrusive_dumb_pipe_to_ALSA

Hi,

According to me, the best GUI front end for MPD is Cantata.
It should be controllable with Android but I honestly never managed to get it to connect to anything...

So after a while, I simply gave it up :-(

I've looked about Lollypop to connect to ALSA. It seems that is a "No go"...
But will look at your link.

Regards.

 

[Robbo99999]

I don't have the 4XX but do have the 400i and out of curiosity measured it yesterday and developed EQ for it. I will see if I can squeeze it in today and post the results but have another review to get out first.

That sounds good. 400i 2020 version is most similar to the 4XX, with the 400i 2016 version being less similar to the 4XX. Do you know which version of the 400i you've tested? I think the 2020 version had updated pads IIRC, which might be the reason for the different response.
Here are some frequency responses from Oratory of the 3 different versions, showing the 2020 version is most similar to the 4XX, particularly the bass response:
HE 4XX:

HE 400i (2020 version):

HE 400i (2016 version):

EDIT: scrub my previous sentence, actually the HE4XX has the treble of the 2016 version and the bass of the 2020 version, so it's not really possible to say from frequency response which one the 4XX is most alike.

EDIT #2: the 2016 version looks like the hardest to EQ. I think in order of preference of ownership of those three (assuming EQ), I'd go HE4XX > 2020 version > 2016 version. Without EQ I think I'd choose HE4XX again, and then the 2020 & 2016 version on equal footing in last place.

 

[thefsb]

If you don't use equalization then this is not a headphone for you. But if you do use it (as you must with any headphone)

Ok. How do I know what to set the eq to?

 

[phoenixdogfan]

I bought the original HD800 when it was initially released. I also couldn't stand the treble, and promptly returned them, disappointed. Fast forward a few years later, supposedly, the HD800S solved some of the peaking, so I picked up a pair. They still sounded bright to me, BUT, when I added a -8dB/Q3.0 filter at 6kHz and lots of bass boost, it actually is perhaps the best HP I've heard (and this list includes the SR-009, Focal Utopia, HE1000, etc.). The only issue I have is with bass impact, even with +12dB compensation. The original HD800 supposedly has slightly lower bass-range THD where it matters, so I purchased a pair of original HD800's and applied the Super Dupont Resonator (SDR) mod, which tames some of the FR peaks, and on top of that, applied the 6khz filter and bass boost. These are incredible HP's with a little EQ.

BTW, I've tried the other widely available EQ curves, including the sonarworks tru-fi, and mine's better.

I also have a pair of HD650's that I listened to for many years, and there's just absolutely no comparison with an EQ's HD800. That's my $0.02.

Sonarworks has two filters for the HD 800, I'm using the double secret one that produces flat to 20 hz bass. Combined with the SD mod, these phones make me very happy. Amazing how they make you feel you're in the booth with the musician. Nothing I've ever heard before does that.

 

[Feelas]

Thank you for posting an excellent example of my biggest problem with everything centered around matching the Harman (sic) target curve. I took a sampling of several dozen measurement plots published there. What seems to be the one that most closely matches that target curve? The Samsung Galaxy Buds. Enough of this insanity!

Hopefully, if (insert your most liked headphones here) matched the target curve 100%, it'd be obvious that the curve makes sense. Please! What does Samsung bashing contribute to the discussion?

the whole point of hi-fi is keeping the sound as close as possible to the original, or reproducing it as accurately as possible.

so keep the components single-purpose and the number of components in the chain as low as possible (dac-amp-speakers/headphones currently, integral amps vs pre+power has always been debatable), avoid any tone controls etc.
and rely on each component adding as little as possible of any kind of character to the sound. manufacturers are the ones who have to tune the sound and it should be a single setting, consumer does not get to change it.

as soon as there is knob besides volume for consumer to turn (tone, bass, treble, eq - doesn't matter) all of that goes out of the window.

Absolutely not. Sorry man, but this is some esoteric bollocks. In fact, you'd be really hard pressed to find real, professional studio monitors which don't include DSP crossovers. Anyways, we should also tell audiophiles to stray from rooms coloring the sound. Better not have a house and listen in open field.

Could you explain exactly how "keeping components single-purpose" exactly helps? As for lowering the number of components & avoiding tone controls - remember, that this is all said from analogue sound perspective and the digital noise floor during processing is MUCH, MUCH higher. You can believe what you want, but there's no real grounding in science of DSP to tell people to stray from using tone controls. Not everything can be done during electrical & mechanical engineering, that can be accomplished with DSPs and there's no discussion about that. How much - might be disputed, but whether it actually happens can't.

Reminds me when I looked at Stereophile measurements of some electrostats - were they Martin Logans? - and completely blown away with the absurd FR. Seems the manufacturer always knows what people want. Obviously, it's being told what they should like.

 

[m8o]

Glad I was able to confirm your bias! I also am not a big fan of Harman curve. However, the data provided does give a good idea of which area of frequency response could be best attenuated with EQ for each headphone. After a while, you can develop your own 'curve' taste for your headphones.

I see how my statement could be taken numerous ways. I assure you I am a fan of the site, measurement database, etc. What I am not a fan of is the busy'ing up of the plot with all those additional curves that relate to, differentiate from, and is necessary to achieve, the Harman curve. But I guess that's the whole point, eq to meet a target. Ill shudup now.

 

[m8o]

Please! What does Samsung bashing contribute to the discussion?

What it provides to the discussion is that the consensus from the parties that contributed to settling on the tone curve that became "the Harman Curve" are completely content with the tone of $3 Samsung Earbuds that came supplied with my samsung Note 8 phone. I couldn't have had my dislike & distaste for that target curve supported with a more powerful statement had I begged the gods for it and one of them delivered.

 

[m8o]

But what measurements explain the perceived differences? I don't know.

I tried to touch on that with an answer pages back. When I gave numerous examples of what characteristics makes two headphones with the exact same FR still have distinct personalities, with one being preferred over another.

 

[amirm]

That sounds good. 400i 2020 version is most similar to the 4XX, with the 400i 2016 version being less similar to the 4XX. Do you know which version of the 400i you've tested?

It is my own and I purchased it November 4th, 2017.

 

[amirm]

Ok. How do I know what to set the eq to?

That was in the review:

You need a parametric EQ (PEQ). You then select the type of filter, followed by the center frequency, Gain and Q factor as listed in the table above for bands 1 through 5. The last one is to taste. The first and second ones are as well.

 

[amirm]

What it provides to the discussion is that the consensus from the parties that contributed to settling on the tone curve that became "the Harman Curve" are completely content with the tone of $3 Samsung Earbuds that came supplied with my samsung Note 8 phone.

Is it the AKG one? If so, it is definitely a big step above average IEMs. I use it every night to watch youtube videos. I will measure it at some point.

 

[jhaider]

So after reading 23 pages, is the science gone from ASR here? Are we back to subjective score boards and spaciousness, clarity, thing? :/
Measurements alone do not tell the whole story? Although curves look alike, the he4xx must be worse than HD800?
Please say no

That's not at all fair. If anything this is the most "science-y" thing going on here. As I see it, "science" in audio (as opposed to engineering validation, which is what technical reviews of audio product actually are) is alignment of subjective listening impressions (ideally blinded when possible) with measured performance characteristics. That starts with listening, establishing measurement protocols, and determining durable correlations that can be tested.

To level set a bit, there's a lot more unsettled when it comes to quantifying the performance of headphones. Measurement best practices, for example. At this point with loudspeakers if a given speaker goes to tier 1 measurement setups (e.g. Amir, Harman, NRC, Dynaudio, and the like), tier 1.5 setups (e.g. Erin, James at Audioholics), or even tier 2 quasi-anechoic setups (e.g. people like Napier, John Atkinson, and me, who may have tier 1.5 aspirations but live in more densely populated areas, as well as any number of hobbyists) there's going to be good to excellent agreement for a given sample above the lower boundary of the quasi-anechoic measurements. That's because there are more-or-less consensus best practices for measuring loudspeakers. Headphones measurements by comparison are a wild west. Every fixture works differently, there's no consensus concerning pad alignment, and so on.

None of that changes the fact that bolded recommendations based on doctored product are problematic, but it's where we are. Even for those of us who don't much care about headphones, Amir taking them on is a promising thing.

 

[bobbooo]

No.
I went to download, which requires you to register, which I do, then nothing happens.

What do you mean by nothing happens exactly? After you've filled out the registration form and clicked register for downloads and the page reloads, you have to click the 'Get Sequence' tab and then 'Get Sequence (zip file)' again for the download to start. (I couldn't get past the registration form at first, but figured out that was due to running Pi-Hole on my home network - it worked via mobile data.) If you still can't download the spreadsheets, I can send you them (don't think Listen Inc. would mind as you have actually registered).

 

[m8o]

Is it the AKG one? If so, it is definitely a big step above average IEMs. I use it every night to watch youtube videos. I will measure it at some point.

It is that one. I can't say I've had the same thought. Sure, it's no gen 1 apple iPhone ear bud. But I have so many other fine things to put over or in my ears, I haven't had the inclination to revisit them after a few minutes the 1st time.

 

[bobbooo]

Excellent, this is a very good omen for my HE4xx purchase, as long as it doesn't skimp on soundstage vs the K702 then it's probably gonna be a better headphone, and therefore will be my final headphone.

EDIT: unless smoothness of frequency response is a big factor, K702 is quite smooth without lots of little variations, whereas HE4xx has a bit more "fine grass" action!

Tiny high-Q variations won't be audible. In terms of what matters (low-Q variations) the HE4XX is generally smoother.

 

[hardisj]

As I see it, "science" in audio (as opposed to engineering validation, which is what technical reviews of audio product actually are) is alignment of subjective listening impressions (ideally blinded when possible) with measured performance characteristics.

This. All of this.

 

[bobbooo]

That sounds good. 400i 2020 version is most similar to the 4XX, with the 400i 2016 version being less similar to the 4XX. Do you know which version of the 400i you've tested? I think the 2020 version had updated pads IIRC, which might be the reason for the different response.
Here are some frequency responses from Oratory of the 3 different versions, showing the 2020 version is most similar to the 4XX, particularly the bass response:
HE 4XX:
View attachment 99854
HE 400i (2020 version):
View attachment 99855
HE 400i (2016 version):
View attachment 99856
EDIT: scrub my previous sentence, actually the HE4XX has the treble of the 2016 version and the bass of the 2020 version, so it's not really possible to say from frequency response which one the 4XX is most alike.

EDIT #2: the 2016 version looks like the hardest to EQ. I think in order of preference of ownership of those three (assuming EQ), I'd go HE4XX > 2020 version > 2016 version. Without EQ I think I'd choose HE4XX again, and then the 2020 & 2016 version on equal footing in last place.

Direct comparison:

The HE4XX has the least bass roll-off, generally smoothest response, and highest upper-treble extension (important for good 'resolution' / 'detail').

Below are their predicted preference ratings before (and after) EQ.

HE4XX: 88 (108)
HE400i 2020: 86 (93)
HE400i 2016: 80 (85)

The HE400i models' lower post-EQ scores are likely due to erroneous non-minimum phase parts of their frequency response that are not able to be effectively EQed out.