Sennheiser HD800

REVIEW: Sennheiser HD 800



Overview

Reviewing a statement product is tough – especially one that was as highly anticipated as the HD800. There is already a huge buzz on head-fi about them, and many have already declared them the best dynamic headphone on the market. But most of these proclamations have been from owners, One potential difference between my listening to the HD800 and other people's so far is that I am fortunate to have been loaned a pair for a fairly lengthy period (1 month) - I did not buy them. So I do not have to deal with either buyer's euphoria, or worry about remorse, coloring my view on the headphone's performance. While I admit I *want* to like these, as I would like another pair of really great headphones, and I could afford to buy them if I liked them enough, my intent in listening to them is to analyze them with no pre-conceived notions, and since I have nothing to lose either way, it allows me a certain freedom to be scrutinize them very carefully, as I believe a "statement" product from a company like Sennheiser should be.

Please note that I am NOT saying that the views of the people who own them are not valid, or any less valid than my opinion. ALL opinions are equally valid. But mine does come from a slightly different point of view - not as an owner, but purely as an analyzer. I do want to add that initially I was considering buying them if I liked them enough. I'm not sure whether that has any bearing on my comments, but I am a big believer in full disclosure, so there it is.


Review Methodology

I spent a LOT of time with the HD800 in a wide variety of contexts. All my comparisons were done at my calibrated listening level of 80dBA (using the Rives Audio Test CD and my SPL meter for calibration).

And just so folks do not think that what I am hearing and have described below is either a factor of my amps, my cables, my sources, or my recordings, let me say this: I used 6 different sources (all with their own set of high-quality cables), 9 different amps, and lots and lots of the world's finest recordings . 3 of my sources are one-time Stereophile Class-A rated - two digital, one analog (the cartridge). I have amps from $400 - $1200 that I played them on, both tube and solid state. Sources included the Denon DVD-5900, Sony SCD-555ES, and my Vinyl rig of Benz-Micro Wood L2 on Denon DP-59L > Audio Electronics PH1 DJH. Amps used were SinglePower Extreme and MPX3, Decware CSP-2, Graham Slee SRG w/PSU1, and Audio-GD C-2-C.

Further, and unlike other headphone reviews, I also used my speakers for reference for this review. I also decided, in addition to other headphones, will that the very, very natural sounding but highly detailed B&W Nautilus 800 Signatures would be valuable in getting a good handle on the HD800’s performance. So I compared the sound of the HD 800 to the B&W N800’s as well as the JVC DX1000 and the Beyer DT880.

Also, I used truly excellent recordings in my evaluations, of wide variety. I've listened to my audiophile standards, which include a lot of jazz, some folk and bluegrass, and some extremely well recorded pop/rock records. I have developed a list of such recordings for evaluation use over the years, as many of you have.

I find Patricia Barber records, for example, excellent evaluation tools, as they are extremely well recorded, have both female vocals and piano, and have both quiet and explosive (for jazz) parts, and wide dynamics. Some of my initial concern about the HD800's treble was a result of listening to Patricia Barber's new record.

I believe this is more than sufficient to eliminate the other variables, and to decide what characteristics belong solely to the headphones themselves.



The Sound


The Midrange

The midrange on the HD800 is fantastically good – the best I have ever heard from a headphone. It is just as full-bodied as the DX1000’s, and yet lacks the DX1000’s very slight coloration. The mids are incredibly open and transparent. They are, to me, the “rightest” mids I have ever heard in a headphone. They make the Beyer DT990’s slightly recessed mids sound just silly by direct comparison (and I like the DT990 overall – a lot).

Mary Fahl and Julie Flanders’ harmonizing in October Project’s “Ariel” (from their eponymous record) is just as engaging as can possibly be. The mids are just slightly forward, but not ever congested. Male vocals are very natural and not overly chesty, or dry. The midrange has great body, but it does not have any noticeable coloration that I can hear. Many cans with body in the midrange accomplish this with a coloration which benefits some music more than others. To an extent the JVC DX1000 are like this, although over time I have grown to love the DX1000’s mids. But the HD800’s are better. The midrange performance is just beguiling. For some, it may be worth the price of admission all by itself.

Just listen to Joan Osborne’s vocals on her cover of The Grateful Dead’s “Brokedown Palace” (from “Pretty Little Stranger”) – I have never heard her voice sound so natural – it’s effortless, unstrained, and has a presence to it – it sounds like a read voice coming from a living, breathing woman, not a 2-D facsimile.


Soundstaging

There is also no doubt in my mind that the HD800 are the imaging champs of the dynamic headphone world. I have owned or heard almost every significant dynamic headphone there is – Sony R10, At W5000 and L3000, Senn HD650/600, Grado RS1 and GS1000, all the ones I currently own, and many, many more I have owned and sold. And I have never heard a headphone image like the HD800. Depth is phenomenal, and somehow width is even better. I really feel like the sound is IN FRONT of me, not stuck “inside my head” like with so many headphones. The soundstaging is better with the HD800 than with the DX1000, and that is saying something – I think the DX1000 image sensationally. But the HD800 is terrific in this regard.

With the HD800, it is easy, not “work”, to imagine real performers in real space. It’s very tempting to keep closing my eyes when listening to them, because the imaging is so convincing. For anyone who has avoided headphones because they don’t image like speakers, spend some time with the HD800. No, they don’t image like speakers – but I can’t imagine anyone who would CARE, because the way they do image is just spectacularly enjoyable.

Simple Minds’ “See the Lights” presents the band in a perfect half-circle around you. I saw them live in their heyday, and hearing them on the HD800’s really brought me back to that show in a way that was almost startling.

The HD800’s have the LEAST amount of isolation of any headphone I own. I wonder if this is a factor in their great soundstaging? Not sure, but however they did it, this is quite an achievement, and here again, maybe worth the price of admission by itself.


The Treble

So we have established that the midrange is terrific, and the soundstaging is amazing. But the treble...ummmm...well...this is more problematic. IMHO, the HD800 has a slightly hot treble. Hundreds of hours of break-in has not changed this. I hear it on some vocals as a pronounced sibilance that I don't think should be there, and I hear it on cymbals that have a slightly over-exaggerated splashiness.

I want to make clear that the treble is extraordinarily clean. There is no grain, spit, grit, or (god forbid) “veil” of any kind here. The treble is very pure in nature, and it has no texture that detracts from transparency. In fact, I would classify the treble as being astonishingly transparent (which is different from being neutral).

When I began to listen to the HD800, and I began to examine the transparency, versus the linearity, and I spent a lot of time trying to determine what was going on, because this is very difficult to accurately assess. How is one to know if the HD800 is just being accurate about the recording, or over-emphasizing it? So I had to listen very carefully to a LOT of music, and do a lot of comparison, to nail it down. Some examples:

> “Burn Down This Town", by Roseanne Cash, from Black Cadillac. The HD800 are over-emphasizing the sibilants on this recording, IMO. An over-emphasis on sibilance, it is very important to understand, is the symptom, but not the "problem".
>“Out of the Woods” by Nickel Creek, from their eponymous album, there are some additional emphasis on Sara Watkins’ sibilants that I do not believe to be an accurate reflection of what is on the recording – this is a very smooth SACD, and it should not have sharp sibilants on it when played back, but they are there with the HD800.
>“So Far Away” by Dire Straits from Brothers in Arms - there is a crispness to the attack of the snare that I think is artificially crisp, and the song in general has an aggressiveness that I don’t think is an intrinsic part of the recording, based on how it sounds on other transducers.

These are just a few examples. It's easy to say "well it's just on the recording" if one does not have any other high resolution transducers with which to determine whether that is the case. But in direct comparisons I have done with the ultra-high resolution B&W N800's, I come to the same conclusion that there is some extra treble energy on the HD800's. This might be something people like – heck maybe even the majority of people, and that is fine, but in the context of a very expensive headphone, I would be happier if the treble were more neutral even than it is.


I don't want to make too big of an issue with this, but it is a real phenomenon that multiple people have noticed. At a minimum, it points to the need for very careful system matching to get the most of out the HD800's, although anyone buying $1,400 headphones should be carefully considering system matching anyway. But if you have a bright source and/or amp, you could be in trouble.

I definitely prefer the treble of the JVC DX1000, which I find to be smoother. The DX1000 are very slightly rolled off in the extreme upper treble - they have a roll-off above 10kHz. That is the VERY upper end of the treble range, and does not affect the area where the HD800 is hot to me, which is more in the lower-to-mid treble. The DX1000 are not rolled off there, but they are flatter versus the reference 1kHz level in the mid-treble than the HD800. If the HD800 were only tipped up where the DX1000 are rolled off, this would just lead to a little more "air". But that is not what I hear going on.

The HD800 has even a little hotter sounding mid-treble versus the DT880, which are *far* from rolled off in the treble - in fact they are sometimes criticized for having too much treble themselves. But the DT880 does not have quite the same problem that the HD800 has with treble (although the DT880 is a little hot in the treble for sure). Their treble flavor is different. This can be seen in the HR Frequency response graphs:



The DT880 actually has less of a treble peak, versus the level at 2 or 3kHz than the HD800 does, and the DT880’s peak is above 8kHz, versus the HD800’s being at 6. So not surprising that they sound different in the treble. Interesting that the HD800 frequency response, in the Headroom graph, certainly doers correlate very strongly with what I seem to be hearing. That is a very noticeable peak of almost 8db at 6 kHz versus 3kHz. I cannot see how anyone can argue that such a peak won't have SOME audible consequence. Of course headphones like the Grados have even peakier treble, but there is no argument that Grados have a hot treble.

It’s important to note that measure headphones is tricky, and while Headroom knows this and does a lot to minimize the issues, one cannot make judgments about a headphone’s sound based on looking just at frequency response charts. In this case, though, there does seem to be a correlation.

So on this issue, I have concluded that, at least as I hear it (but also seemingly supported by the frequency response and by the experience of some others), there is a treble coloration with the HD800, in that more than being just "very revealing”, the HD800 has an elevated treble response that is a departure from neutral . Yes, this may periodically enhance sibilance, but it has other effects, like a little extra sizzle on cymbals (which I have also noticed), and generally delivers a slightly “tipped-up” sound. For many, this will come across as just a “revealing” nature, but for others, it will be bothersome. I found it to be enough of an issue that it has dissuaded me from buying them, given the asking price.

And again, to try to minimize the ruffled feathers - this is all in the context of a $1,400 headphone, which I believe deserves microscopic scrutiny given the asking price. I'm already willing to state that the midrange and soundstaging are the best I have ever heard in a dynamic headphone. But the treble performance is still a slight disappointment for me given this is an expensive, "statement" headphone.’

None of what I am hearing will mean that there won't be LOTS of people who will LOVE what the HD800 does in the treble. Heck, there are some headphones that are KNOWN to be terrifically bright (the AT W5000 come to mind), and yet they have a lot of fans. Given that what we are discussing is such a small thing, relatively, there will be some people who won't be at all bothered by it, and some who will indeed LIKE it. But from what I can hear, it's there, and it should be known.


The Bass

I find the bass quantity to be slightly on the light side. The QUALITY of the bass is stupendous – very well defined, tight, and with great definition and attack. But the DX1000 manage this trick at least as well as the HD800, and manage to have bass weight as well. I find the bass weight of the HD800 to be similar to that of the DT880, which I personally always wish had more bass. I am willing to accept that this is may be just a preference for me, but it nonetheless is something to be aware of – if you like a punchy sound, the HD800 are not always going to please you.

Again let’s look at en example. Listen to “A Secret Place” from Megadeth’s “Cryptic Writings” – the kick drum simply lacks impact. Listen to this track on even $500 speakers – it will sound more like at least I think a kick-drum should. It isn’t imactful on the HD800 – it’s overdamped, and subdued, versus what it sounds like through the N800’s, and the DX1000. It only sounds similar on the DT880. Yes, the impact of the striking of the drum is VERY clean, and very tight, but the drum itself seems to lack body.

The bass is also definitely less than the HD650, and based on my brief comparison, also less than the HD600. Some people may prefer the HD800’s bass performance. It may be "better" bass on the HD800 than the 650, but it is less for sure.

I think it’s very possible that the overdamped bass is what actually makes the treble stand out a little. Or maybe it’s the other way around. But there is a well known trade off there. The famous loudspeaker designer Henry Kloss once did a demonstration where he took a speaker that people thought had too little bass, and he padded the tweeter down by 2db, and all of the sudden – WOW – bass. No change was made to the woofer or the woofer level.

The HD800 have a very tight-fisted punch in the bass – but they punch at something like bantam-weight or junior welter-weight at most.


On the issue of amplifying the HD800

Much has been made about the inductance of the HD800, and the need for certain types of amps. Some people swear they sound great on tube amps, and some insist that due to the inductance, any amp with a high-ish output impedance will not sound as good as they should, and that this includes many OTL tube amps. It's a complex issue, and that doesn't mean the HD800’s *won't* sound good on any given OTL amp. But on the SinglePower Extreme, the treble emphasis was slightly worse that it is on the SS Audio C2C, which I would not have expected. I have never considered the Extreme to be at all rolled off, and of course I cannot measure its output impedance, although SP claims it to be 20 ohms, which isn’t very high, so it's likely IMO that the amp isn't contributing to the treble emphasis I hear, and that the slightly laid-back C-2-C might be masking the HD800’s treble list slightly.

I also tried the HD800 with the Darkvoice 337 ( Tung-Sol mesh-plate 6SJ7GT + GEC 6AS7G) and the 336SE (Shuguang Treasure CV181 + Sylvania 6AS7G). In both cases the treble emphasis was just a tiny bit more obvious, and just slightly more pronounced, than with the SS Audio-GD C2C. I tried the Graham Slee SRG+PSU1, and the treble was between the Extreme and the C-2-C, but on all of these amps, the feeling of a bit too much treble energy never went away. On the other hand, I felt the treble performance on the Decware CSP-2 was basically identical to the C-2-C. So in the end, I think too much was made of this issue of amp type (as is almost always the case).

The HD800 do not, IMO, change their own character radically from amp to amp. Being as revealing as they are, the HD800 certainly will let you know the difference in the sound of different amps, or different tubes. And of course it is difficult to generalize about tube amps, since you can often get the exact sound you're looking for by tube rolling. For example, if I took the 5998's out of the Extreme, and put in RCA grey-plate 6AS7G's, the sound would be very different, and I probably wouldn't notice any treble elevation in the HD800, since the RCA is a soft-sounding tube.

So the point is that with OTL designs there is a potential issue that should be noted - nothing more than that. And since one can alter the sound by using different tubes, tube amps still offer nice flexibility in this regard, and a great tube amp can still sound great with the HD800. My favorite combo ended up being the Decware CSP-2 and the HD800. The Decware, with Amperex 6DJ8’s and a Sophia Princess mesh-plate 274b, when combined with the HD800, provided enough lushness that the sound was just beguiling, and I didn’t worry much about the tipped-up response – the music was enthralling.


Build Quality and Fit

I found the HD800’s very comfortable to wear for long periods of time. They are big, and they were right at the limit of earcup size for my head, but they were within the limit. I of course heard the “pinging” of the spring that people have commented on, but I wasn’t bothered by this. The cable isn’t at all microphonic, which sure is nice.


Conclusions

I've been living with the HD800 for a couple weeks, and I while I think they are truly excellent in many ways, there are some things that bother me in the context of a $1,400 headphone. After several hundred hours of burn in, and hours and hours of listening and comparing, on a very wide variety of amps, I still find the treble to be a bit too aggressive, and the bass to be a bit polite. This is a flavoring, and many will like the flavor. But it is a flavor, and it's not going to be for everyone.

That said, the soundstaging and midrange are so good that the slightly tipped-up response, for many people, will be worth overlooking, and some people will of course LIKE the slight tip-up. And with the right system matching, it might be a non-issue. For me, they are not more enjoyable than the DX1000, and so I’m not going to be buying a pair, given that they cost $1,400. But for you, dear reader, they might be just exactly what you are looking for, and they are certainly world-class performers that have some incredibly special qualities.

I’m sure I’m going to suffer many slings and arrows for criticizing these, even though anyone who actually reads the whole review will see that I’ve actually RAVED about the HD800’s in MANY ways – best mids and imaging of maybe any headphone currently made; crystal clear treble; super-tightly define bass. I simply don’t think they are completely free from personality. As always, this is only my opinion, but it was very carefully arrived at, if nothing else. In the case of a headphone that costs $1,400, I think we have the right to expect a lot; I don’t think we have the right to expect universal perfection, as there is no such thing. So what I have done is tried to describe the flavor of the HD800 as I hear them. Perhaps at least to a few people, this will be of some small help.