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:
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:
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:
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):
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:
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:
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:
Group delay is quite messy, likely due to internal resonances which combine with the direct sound, creating complex interactions:
Impedance is flat but low:
Finally, sensitivity is the worse of any headphone we have measured so far:
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:
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/
The Diana V2 is extremely light and has exquisite build and attention to detail:
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:
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:
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):
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:
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:
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:
Group delay is quite messy, likely due to internal resonances which combine with the direct sound, creating complex interactions:
Impedance is flat but low:
Finally, sensitivity is the worse of any headphone we have measured so far:
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:
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/
