Drop AKG K7XX Review (Headphone)

[Billy Budapest]

Although K701 production shifted from Austria to China about 10 years ago, the only significant difference between K701’s manufactured in the different regions is the number of bumps on the headband.

The K701’s made in China have 7 bumps (instead of 8), although I have heard that some also have smooth, bumpless headbands.

 

[xaviescacs]

The K701’s made in China have 7 bumps (instead of 8), although I have heard that some also have smooth, bumpless headbands.

But in this way the central bump lays directly in the listener head's bump. This 7xx seems to have a bumpless headbump. It's kind of baffling that they can add or subtract a bump instead of fixing the distortion. Can you imagine the meetings with the product manager? xDD

Added quote: "I coulda had class. I coulda been a contender. I coulda been somebody, instead of a bump, which is what I am, let’s face it."

 

[Tks]

"Distortion" covers a very wide range of things. Here we are concerned about non-linear distortions, and even those come in a range of flavours. A simple non-linear transfer function could reasonably be nulled with DSP by simply applying the reverse transfer function. But life is never that simple. It isn't just that the driver may be moving non-linearly, different parts of the driver may be moving differently relative to one another. You can still DSP out some of these distortions (JBL do this for one of their very high end compression drivers where they have a closed form model of the way the dome flexes, and can pre-distort the signal to manage it. This is probably the limit of current technology)

If you see huge nasty spikes of distortion you know you have some sort of resonance/breakup mode involved with the distortion. Drivers can also have different stable modes of breakup, so once you get them distorting, you might reduce the level, below where the distortion set in, and they keep distorting. None of these are likely to be possible to compensate for. Even with feedback/feedforward systems.

Feedback is pretty difficult anyway, you need a feedback sensor that has lower distortion than the target for the system you are trying to correct, and the entire system has a whole range of near impossible stability constraints. Can be made to work for bass drivers, but past this it gets desperate. Adaptive feed-forward is garnering a lot of interest. That tries to predict the way the system will distort from internal models and correct for that (feeding the correction for the predicted distortion forward, hence the name). The prediction system can measure how the system is currently distorting and update the model on the fly. This isn't subject to the same stability constraints. The AI weenies like to imagine they can use some sort of deep learning technique to short circuit the work needed to construct the distortion model. Wish them luck. There are fundamental constraints on how much information is available.

A reasonably well behaved system might well be successfully be further linearised, but one that is just plain badly behaved is unlikely to be a viable candidate. This headphone fits the latter class.

Almost every time you post, I learn something fascinating.

 

[Presently42]

Added quote: "I coulda had class. I coulda been a contender. I coulda been somebody, instead of a bump, which is what I am, let’s face it."

Ooo, On the Waterfront! Excellent film, excellent director, superb actor and superb musical score. Strongly recommended!

 

[jensgk]

Feedback is pretty difficult anyway, you need a feedback sensor that has lower distortion than the target for the system you are trying to correct, and the entire system has a whole range of near impossible stability constraints.

Would it be possible to use some kind of laser directed towards the driver surface as a feedback sensor?

 

[markanini]

I've tried the whole series: K701, K702, Q701, K7XX, K712 and the cheapest ones (K612) sound definitely the best

K612 gang knows what's up!

 

[_thelaughingman]

K612 gang knows what's up!

Shoot I am in the process of buying the K612 Pro because of their rated tonality.

 

[tdockweiler]

AFAIK K702 Anniversary is supposedly the same thing as K712 (Much as Senn's HD 580 Jubilee= HD600). Who knows what you're getting with Drop; their documentation is usually a bunch of 'word salad' whipped up by the sales dept.

I don't think this is true.
I've heard the K702 Anniversary and the K712, but not side by side.
The Anniversary was so bad to my ears that I'd never want to ever listen to it for more than testing.
It actually felt too warm sounding. Everything so thick sounding and very muffled.
Even worse than some of the earlier HD-650s.
Not nearly as bad as a Q701 bass modded with K712 pads though.

Take a look at the Anniversary pads. They are totally different than the K712 pads.
They look more like round doughnuts and are way less deep. The edges more rounded off.
The K712 pads have higher "walls" around the edge of the pads.
It's possible some Anniversary pads are exactly like the K712 pads.

It's possible AKG figured out how to improve those memory foam like pads for the K712.

BTW I compared my Q701 bass modded side by side to the K712 and preferred the Q701 by far.
If the K712 wasn't a total rip off i'd be perfectly happy with the K712 though.
My bass modded Q701 did sound a LOT smoother sounding in the treble somehow.
The K712 did have fuller sounding low mids though.
I listen to a lot of harsh Jpop and the bass modded Q701 would be my pick for that.

I think I paid only $200 for my Q701 and have loved it ever since I got it.
No way i'm ever selling it. Sometimes I prefer the DT-990 Pro due to it having more low bass (and better comfort).

LOL you may think I'm insane, but I just threw away TWO K400s when I moved.
The bass was totally missing and that alone made it not worth listening to.
The early AKGs are IMO severely overrated.
Even the K501 and AKG K240 Sextette were nothing special.
I had the EP and LP K400s.
One of them I found on Amazon and the box was so old that it smelled like it had been in the guy's garage for 20 years.
I think I scored it for under $75.
I disliked both so much that I probably never used either for more than a few hours.

The K501 gave me the impression though of the hugest soundstage i've ever heard (depending on the track).
Playing Fallout 3 with that headphone was just ridiculous. You'd go into Tenpenny Tower and felt like you were listening to surround sound speakers.
But of course there is almost no bass with that headphone.
I traded it away and really kind of regret that.
I think the new owner still loves the K501 and will never part with it!

 

[Ron Texas]

What a shame since they looked so pretty with the red trim and big round pads. Thank you @amirm

 

[theglobalelite]

@amirm I primarily like my K702 for their clarity/detail combined with good soundstage (best of my headphones).

Amir, I notice that the K7XX doesn't have angled pads whilst the K702 does have angled pads, so perhaps you'd find the K702 has better soundstage than the K7XX.

Definitely enjoy the soundstage of my K702s as well, even with their other flaws. I mostly drive HD6XX's and Sundaras, but sometimes I'll thrown the K702s for concert albums, as the 702s's wide soundstage complement that recording style nicely.

 

[TungstenC]

A while back I was EQing the 7XX by ear by using Solderdude's measurement as a guide, tone generator, pink noise, some music and ended up with an profile quite similar to the review.

I have notice similar findings in the 2K region, at first I was confused as to why is sounded "overly bright" when I boosted the region slightly to compensate for the trough. It's apparent now with measurement that's non-linear distortion...

Spoiler: my EQ settings for AKG K7XX

Preamp: -4 dB
Filter: ON PK Fc 26.5 Hz Gain 4.2 dB Q 0.4
Filter: ON PK Fc 281.4 Hz Gain -3 dB Q 0.3333
Filter: ON PK Fc 1000 Hz Gain 1.6 dB Q 0.785
Filter: ON PK Fc 2100 Hz Gain -3 dB Q 2.8764
Filter: ON PK Fc 8500 Hz Gain -2 dB Q 5.5
Filter: ON HS Fc 8000 Hz Gain -4 dB

 

[Helicopter]

What a bummer. They look so cool.

Thanks for the review.

 

[phoenixdogfan]

A truly iconic design. Just a shame it's so badly broken.

 

[Francis Vaughan]

Would it be possible to use some kind of laser directed towards the driver surface as a feedback sensor?

A laser can be used to measure displacement. Indeed Klippel will sell you one for loudspeaker measurement. The problem with feedback is that your feedback loop must respond very quickly. There is always some time lag in the loop, and that lag diminishes the ability of the loop to correct any error. Worse, that delay means the phase of the error signal changes with the frequency of the signal, and eventually, as the frequency rises the phase moves enough that instead of subtracting error you are adding error, and the system goes unstable. The overarching requirement is the Nyquist stability criterion, which states that the loop gain must drop below unity when the phase reaches 180 degrees (or if you plot the loop gain on the complex plain, the loop gain must not enclose the point (-1,0). This is common stuff for amplifier design, and is basic control theory. But for a feedback system on a driver your sensor needs to sufficiently fast, and accurate with it. That is really hard. You need a laser sensor with say 16 bits of accuracy and a sample latency of minimally 10us (or the equivalent in the analog domain). Good luck finding one of those.
That is one of the reasons feedforward gathers so much interest. You don't need the speed, as you can process the measurements at any rate you like, calculate the required error prediction, and apply it. In principle the error prediction could converge and you don't need to update after some time. Depends upon the compexity of the distortion model and what you are trying to correct. Your distortion model could be a phyical model, and include things like magnetic properties of the driver (field strength modulation for instance) magnetic motor linearity, thermal effects in the voice coil, heat transfer modelling, non-linearities of suspension, non-linearities over time of the non-linearities. There is the need for a lot of skill in designing the model and the inputs to the model. (Which is why deep learning and the like is vastly harder to get working than people think.)

 

[Hifihedgehog]

The tonality of the AKG K7XX is wrong out of the box. This is the case for many headphones but what was rather unique here was severe distortion in the 1 to 2 kHz.

Ah, the infamous speaker killer K700 series headphones from AKG. To this day, I still cannot understand why the original K701 got such rave reviews. Exhibit A above, and case in point shows why they shouldn't. I found the K700 series family to have always had a plasticky tonality to them and this is bonafide evidence of the aural anomaly. I would not be the least bit surprised if this spike in distortion is common among that product family.

 

[phoenixsong]

To this day, I still cannot understand why the original K701 got such rave reviews. Exhibit A above, and case in point. I found the K700 series family to have always had a plasticky tonality to them. I would not be the least bit surprised if this spike in distortion appears throughout that product family.

For one, the sense of space is real and enjoyable in the K7-- series. I'm also guessing that DACs and amps in the past were less resolving and came with their own problems, so with certain pairings the series can actually sound really good relatively. I remember at least 6-9 years back (forgot the exact timing) equipment pairing was a major thing

 

[Hifihedgehog]

For one, the sense of space is real and enjoyable in the K7-- series. I'm also guessing that DACs and amps in the past were less resolving and came with their own problems, so with certain pairings the series can actually sound really good relatively. I remember at least 6-9 years back (forgot the exact timing) equipment pairing was a major thing

I'll agree to disagree. I honestly never understood this technique of putting lipstick on a pig, the so-called pairing approach of audiophilia. Premier quality equipment should get out of the way of the signal chain as much as possible, never requiring special distortions or radical shifts from natural response to conceal its flaws. Hearkening back to that era you speak of, the setups I heard that made the K701 sound "good" always sounded fuzzy, bloated and warm to me, obscuring inner recording detail. That to me bespeaks of a severe problem because when I am paying a substantial sum of money, I want to hear the source material sing and not the playback equipment ring. Thankfully, we are confirming here more and more here that retail price and series placement is most certainly not necessarily commensurate with real-world performance. I would gladly take one of AKG's cheaper studio headphones over this overrated mess.

 

[phoenixsong]

I'll agree to disagree. I honestly never understood this technique of putting lipstick on a pig, the so-called pairing approach of audiophilia. Premier quality equipment should get out of the way of the signal chain as much as possible, never requiring special distortions or radical shifts from natural response to conceal its flaws. Hearkening back to that era you speak of, the setups I heard that made the K701 sound "good" always sounded fuzzy, bloated and warm to me, obscuring inner recording detail. That to me bespeaks of a severe problem because when I am paying a substantial sum of money, I want to hear the source material sing and not the playback equipment ring. Thankfully, we are confirming here more and more here that retail price and series placement is most certainly not necessarily commensurate with real-world performance. I would gladly take one of AKG's cheaper studio headphones over this overrated mess.

Ah, do note that I never once said that 2 wrongs make a right! Just coming up with possibilities for the positive reactions to this line of headphones since more than a decade ago

 

[phoenixsong]

For us social workers, context accounting for space-time is paramount after all

 

[Robbo99999]

AFAIK K702 Anniversary is supposedly the same thing as K712 (Much as Senn's HD 580 Jubilee= HD600). Who knows what you're getting with Drop; their documentation is usually a bunch of 'word salad' whipped up by the sales dept.

They're different, have a look here:
https://www.reddit.com/r/oratory1990/wiki/index/list_of_presets