• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Abyss Diana V2 Review (headphone)

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
This is a review and detailed measurements of the Abyss Diana V2 open-back planar headphone. It is on kind loan from a member and costs US $2,995. You get to choose your cable length and type which can add another $50 to the cost.

The Diana V2 is extremely light and has exquisite build and attention to detail:

ABYSS Diana V2 review open back headphone high-end.jpg


It almost melted on my head as soon as I put it on, getting very close to perfect for me. Note that the test sample has an aftermarket pad on the headband:

ABYSS Diana V2 review open back headphone headband pad.jpg


Without it, it may have been less comfortable. I see reports online saying the paint chips easily which if so, would be a serious concern in this price range.

The pads are very soft and attached with powerful magnets and four pins so you can rotate them or change them for another (I have the "DMS" pads which I may test next).

The softness of the pads presented serious challenges in measurements. Mere lifting them and letting good would change bass response by as much as 5 dB! I am not talking about moving the pads. Just pulling and letting go. Not only did the bass change, so did the response above 5 kHz.

Note: The measurements you are about to see are preformed using standardized GRAS 45CA headphone measurement fixture. Headphone measurements require more interpretation than speaker tests and have more of a requirement for subjective testing as a result.

Abyss Diana V2 Measurements
Let's start with the Diana V2 frequency response measurements:

ABYSS Diana V2 frequency response measurements.png


The above is actually the fourth set of measurements I made! I initially made a measurement that looked similar to above. But then noticed one of the pads was rotated 90 degrees relative to the other. I fixed that and upon measurements, the results were quite different. I developed a new EQ based on that and results were quite poor, sounding worse than the no EQ and the first pass EQ. I made another measurement and spent another hour creating an EQ, once again resulting in failure. That is when I discovered what I said at the outset. That while the pads nicely rest on the measurement rig, slight deformation of them radically changed the response especially in the 2 to 6 kHz region.

Since the first result was most agreeable I attempted to replicate that and came close with what you see above. Equalization this time worked even better than my first pass, giving me comfort that this measurement is closest to the "truth." Still, I want to highly caution you to not blindly program EQ based on the graph above. Notice how for example the two channels have different peaks and by good bit.

I even found the filtering I was applying to the relative response was causing me to chase the wrong thing so I reduced that fair bit (1/24th octave now):

ABYSS Diana V2 Measurement relative frequency response.png


Other than the typical loss of sub-bass, the problem we have is slight peaking around 1 to 1.5 kHz and major shortfall between 2 and 10 kHz. You might be tempted to go and fill that large hole with EQ but careful. Look at this distortion graph:

ABYSS Diana V2 Measurement distortion.png


Ouch! The two areas we want to fill in, the bass and the treble are the two with the highest distortion! Any boost in these areas could cause harshness due to harmonic distortion increasing the high frequency energy. This can cause brightness and ear fatigue. Even doing nothing you are faced with the above situation.

To show how bad the distortion is, let's zoom out:
ABYSS Diana V2 Measurement distortion zoom out.png


It doesn't look to me that any effort was spent in analyzing distortion in this driver and attempting to mitigate it. We have seen plenty of headphones with much lower distortion and a number with almost none (relative to bass) in upper frequencies.

The distortion mechanism here is complex as its peak frequency is level sensitive. Notice how the green curve peak that represents 114 dBSPL fore example, is 1 to 2 khz lower than that of 104 dBSPL.

In absolute terms, distortion exceeds our 40 db thresholds in the two areas we already know about:
ABYSS Diana V2 Measurement THD distortion.png


Group delay is quite messy, likely due to internal resonances which combine with the direct sound, creating complex interactions:

ABYSS Diana V2 Measurement Group Delay.png


Impedance is flat but low:
ABYSS Diana V2 Measurement Impedance.png


Finally, sensitivity is the worse of any headphone we have measured so far:

most sensitive headphone.png


I had to really floor my headphone amp to get it to sing so above graph is pretty correct.

Abyss Diana V2 Listening Test and Equalization
In stock configuration, the sound is rather dull, gritty and lacks any deep bass. On the latter, in my tracks with deep bass, you basically would hear nothing. This is how I knew the curve that showed the roll off you see in the measurement section was the correct one. I am going to blame the grittiness on high distortion. Spatial qualities were not there either. Was the sound offensive? Other than the grittiness, no. But you don't spend this much money on a headphone to get inoffensive sound.

As I noted earlier, my first pass EQ went pretty easy last night. I should have left well enough alone but did not so had to start all over this afternoon. After some four hours and listening to ton of tracks, this is the final outcome:

ABYSS Diana V2 Measurement headphone EQ Equalization.png


There is a ton of tuning by ear here. The trickiness was that boosting the 2 to 4 kHz would cause brightness which I don't like. One of the solutions to that was the little dip in yellow.

I was pleasantly surprised that the Diana V2 could handle any amount of bass boost I threw at it! For grins, I pumped that to 12 dB and it kept on playing and producing more low bass without bottoming out, cracking or anything of the sort! Paradoxically, it would cause some brightness due to high harmonic distortion spreading into upper bands but the bass remained under control itself. You would need to tune this for yourself. What I have up there is the best compromise. Anything higher would require massive amount of headroom to avoid digital clipping which in turn require tons more amplification gain. Maybe you want to create multiple profiles and switch between them depending on what you play.

Once this EQ was in place, the bass was naturally improved and I could now hear and enjoy the sub-bass quite well. It was neat to have my ear-lobes resonate with bass even at moderate playback levels. The sound was now more open and spatial ("soundstage") quality improved. Overall the transformation was not large but very much welcome.

Alas, the grittiness, while reduced some, still remained. I did not like strings for example in classical music. There is just too much distortion in this driver to be able to do anything about this. As a result, I could not listen to the headphones for long. It was a relief to take it off and switch to my much larger headphone at the end.

Conclusions
On the physical side, I think you are getting decent offering in the form of "luxury" when it comes to this headphone. Unfortunately, nothing like that comes along on the fidelity front. Both subjectively and objectively there is no there there. I like to see near perfection as far as distortion is concerned even if they thought different tonality was merited than our target. At the same time I can't imagine anyone being able to make a good case that you should have a chunk of your spectrum taken out from the lower to middle treble. This is where a lot of "air" and sense of "hi-fi" exists in music.

Equalization was hampered by difficulty in measuring the headphone and the fact that distortion can work against you, creating more harshness and artifacts as you attempt to fill in the holes. I don't mind distortion if the response is too high in that region. That way, I can pull it down and both fix tonality and reduce distortion. Here we have the reverse situation.

I really liked the feel of having this bit of luxury on my ears and the comfort it brought. And with EQ, good experience for a while at least. But just couldn't go far enough to fix all the ills of the design. Company needs to work on V3 with the goal of killing the distortion in mid frequencies where our hearing is so sensitive.

In summary, I can't recommend the Abyss Diana V2.

Yes, I have read the other reviews where people say they died and gone to heaven after listening to this headphone. Don't ask me to rationalize their experiences. I am here to rationalize mine. :)

------------
As always, questions, comments, recommendations, etc. are welcome.

Appreciate any donations using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
 

Attachments

  • Abyss Diana V2 Frequency Response.zip
    65.3 KB · Views: 347

sweetchaos

Major Contributor
The Curator
Joined
Nov 29, 2019
Messages
3,872
Likes
11,546
Location
BC, Canada
For Equalizer APO users:
Preamp: -8.6 dB
Filter 1: ON LS Fc 35 Hz Gain 8.0 dB Q 1.0
Filter 2: ON PK Fc 1194 Hz Gain -2.0 dB Q 2.0
Filter 3: ON PK Fc 2705 Hz Gain 2.5 dB Q 3.0
Filter 4: ON PK Fc 3700 Hz Gain 5.0 dB Q 9.0
Filter 5: ON PK Fc 5464 Hz Gain -2.0 dB Q 8.0

For more info, see List of Amir's Headphone PEQ filters.
 

digicidal

Major Contributor
Joined
Jul 6, 2019
Messages
1,981
Likes
4,838
Location
Sin City, NV
I'm reasonably sure that no one paying ~$3K for phones (that they've kept beyond the return period) would ever dare to say (or even think) anything but praise for them. I know my criticism for things I've "over-invested" in is far more muted, and I try to be at least somewhat objective on that front. Luckily most of my horrible audio acquisitions are measured in the low hundreds of dollars... so I don't feel bad being more honest about how much they suck in reality. ;)

I can count on one hand all the disgruntled Wilson Audio customers I've encountered... but based on what I've seen measured and what I've heard on occasion in demos... I think there have to be some secrets being held there too.
 

bunkbail

Addicted to Fun and Learning
Forum Donor
Joined
Feb 27, 2018
Messages
520
Likes
662
Truth to be told, this is my dream set of headphones and have been saving up money for awhile now. Thank you for the eye opener.
 

xarkkon

Active Member
Joined
May 26, 2019
Messages
228
Likes
338
Oof, that's ugly. The choice of panther's a pretty kind one given the price and performance.

Head's up on typo: "Other than the typical lass of sub-bass, "
 

the_brunx

Senior Member
Joined
Apr 7, 2020
Messages
341
Likes
859
Maybe they will perform a zillion times better with the expensive cables they sell at their site?

https://eu.abyss-headphones.com/collections/headphone-accessories

“Compared to stock cables (which are very good)-- Greater inner clarity and fuller upper bass and mids without brightness, soundstage and room boundaries open up outside your head, musically detailed from the bottom on up, far easier to touch the source. Did we mention you have to hear these cables...”


how did this product survive the piggy bank pink panther?
 
Last edited:

Archsam

Senior Member
Joined
Apr 8, 2020
Messages
326
Likes
513
Location
London, UK
So which headphone has committed the bigger sin, the Abyss that won't take EQ well and will always sound deficient in some way, or the Audeze LCD-X that has an unfortunate target curve out if the box but sounded great after EQ ?
 

AlexHempel

Member
Joined
Dec 23, 2020
Messages
14
Likes
19
I tried the Abyss Phi and the Dianas at Can Jam two years in a row. I found the Phi to be quite good, but when I auditioned the Dianas, I assumed they were $300 or $400 max based on the sound and the build quality. I was blown away when I heard the price tag, not really sure what they were aiming for with this product. Hopefully we can get measurements of the Phi at some point.
 

vkvedam

Addicted to Fun and Learning
Forum Donor
Joined
Apr 12, 2019
Messages
583
Likes
807
Location
Coventry, UK
Wow! Didn't see this coming at all :eek:, thanks for shedding some light on this :facepalm:
 

xup

Member
Joined
Jun 7, 2020
Messages
32
Likes
61
Very interesting. I've owned These, and currently own the Diana Phi. I actually like the tuning of these (but liked the Phi better).
That said, even though I've spent a lot of money on these I have no problem saying that this distortion is terrible and I would want every company to do much better.
I do wonder if the Phi would measure significantly differently WRT distortion (supposedly a different driver). I do think the Phi resolves the dip in the mids.
Inb4 claims of snake oil, I don't believe in cables, have RME+A90 and had a few other pairs before.

As for Hifiman, I found their build quality lacking and their sound dull personally (had the HE560 and Ediotion X V2), so have no desire to try more of their headphones.
Never owned an Audeze, but their weight alone is something I'm not willing to consider.
 

Nango

Major Contributor
Joined
Aug 6, 2018
Messages
1,463
Likes
985
Location
D:\EU\GER\Rheinhessen
How about that. People on various forums rave about Abyss 'phones. Those measurements - especially the distortion plots - look simply horrible.
If one pays 3k for a headphone you can nothing but endorse it and talk marvelous about it. This is again just another example for the bubble going on regarding head-fi in general. Its time this bubble to burst!!!!
 
Last edited:

Anon4273

Member
Joined
Jul 3, 2020
Messages
5
Likes
4
It is kinda interesting although we just have 2 examples(Heddphone, Diana V2) that some of the modern high end have a rather big amount of distortion and are "loved" by some. It is like AMPs/DACs some call realism liveless/boring but bad meassurements can show how much effort was made to make them because nobody builds something to me meassure poorly(tubes are kinda an exception).
 

YSC

Major Contributor
Joined
Dec 31, 2019
Messages
3,194
Likes
2,570
I think the distortion in mid to high frequencies are those common in early planar designs? I remember once reading somewhere that Hifiman and Audeze alike modern planars are praised because they've found a way to fix those distortions in HF
 
N

nhatlam96

Guest
Truth to be told, this is my dream set of headphones and have been saving up money for awhile now. Thank you for the eye opener.
You could still try it out and fiddle with EQ. If you happen to not like it, then simply send it back.
 
Top Bottom