*NEW* AKG K702 has been measured by Oratory!

[Dazerdoreal]

AKG K702.pdf

Shared with Dropbox

www.dropbox.com

As the topic is up there:
I think the new Oratory1990 EQ contains a typing error. -3,6db@210hz is too much. He probably meant 2,6db.
I entered his measurement & EQ data into Squiglink and you can see that his filter is exactly 1db too strong in that area.

 

[Robbo99999]

@isostasy and others, my 4th unit of K702 arrived today. As you know I was wondering if the newer K702's have been silently revised, based on the fact that Sean Olive's measurement of the K702 somewhat recently showed that it was really quite different to the K702 frequency responses that we're used to looking at. So I've measured my 4th unit of K702 on my miniDSP EARS and quickly compared it to my Unit 3 K702, and as it turns out there's not that much difference, at least not enough difference to describe the missing 5.5kHz in Sean Olive's measurement. Following is Unit 3 vs Unit 4, the blue line is Unit 3 (don't worry that the frequency response looks strange overall, that's because it's measured on a non-standard rig, measured on a miniDSP EARS):

As you can see there's nothing to suggest that the newer K702's are any different to older K702's, at least not in terms of differences at 5.5kHz which is where the peak was missing from Sean Olive's measurement - as you can see there are no differences at 5.5kHz. So I think any differences between Sean Olive's measurement (which mirrored Crinacle's measurement quite well) and Oratory's measurement is down to measurement protocol or just something associated with the measurement process. I don't think it's unit to unit variation as all my K702 have the same general traits in treble around that 5.5kHz area, although I haven't dug through my measurements of my Unit 1 & 2 to put them on the same graph as I'm short on time right now, but I remember (know) them to be similar in this respect - I'll probably put all 4 units on a graph when I've got more time. So I think it comes down to which measurements you want to use - Sean Olives & Crinacle's that both agree with each other in terms of not having a 5.5kHz peak or Oratory's measurement that has a 5.5kHz peak. I may do some other investigations to work out if my miniDSP EARs measurements show evidence of a 5.5kHz peak when a generic "multi-headphone" conversion curve is applied to the minDSP EARS measurement.

And following is some further general information about the performance of my Unit 4 K702: it's got very good channel matching:

And following are all the individual measurements of both left & right channels of my Unit 4 shown all on one graph (as I measured each channel 10 times after reseating the headphone on the rig each time):


Following some Distortion Measurements, just at an SPL level of 85dB (@1kHz):
Right Channel Distortion Measurements (these look ok I'd say):

Left Channel Distortion Measurements (somewhat worrying peak at 2.5kHz):

My Unit 3 doesn't have the distortion peak that is seen above, I'd have to do some more distortion testing on this Unit 4, this was just a quick one, I've run out of time today. Might revisit some parts of this post in the coming days or week to flesh out some more bits and perhaps to add to any of the points I've investigated - I'll either do a seperate post or update this one.

TLDR: Channel Matching is good in my newly bought Unit 4, and it doesn't vary significantly from my others indicating that a silent revision has not taken place. Following are the production numbers for each of my K702, each K702 that is produced is given a number that is simply their running total of the total number of units they've produced so far:
Unit 1 from yr 2015: 123864
Unit 2 from yr 2020: 204975
Unit 3 from yr 2022: 212059
Unit 4 from yr 2023: 224865
(Sean Olive got his sample fresh from the warehouse 241706 was the number, so his is newer than mine eventhough mine was bought 4 months after he posted his measurement. Oratory measured 230909 as his latest one.....maybe a revision has taken place between 230909 and Sean Olive's 241706, but the fact that the Crinacle measurement of even older (old old in comparison) don't show that 5.5kHz peak and agree with Sean Olive's in that respect - makes me think there's not been a revision. I still think it's something related to the "measurer & measurement procedure".)

 

[Robbo99999]

AKG K702.pdf

Shared with Dropbox

www.dropbox.com

As the topic is up there:
I think the new Oratory1990 EQ contains a typing error. -3,6db@210hz is too much. He probably meant 2,6db.
I entered his measurement & EQ data into Squiglink and you can see that his filter is exactly 1db too strong in that area.

Might be just worth checking if the same result is seen when using REW.

 

[Robbo99999]

AKG K702.pdf

Shared with Dropbox

www.dropbox.com

As the topic is up there:
I think the new Oratory1990 EQ contains a typing error. -3,6db@210hz is too much. He probably meant 2,6db.
I entered his measurement & EQ data into Squiglink and you can see that his filter is exactly 1db too strong in that area.

Actually, I just checked this using REW and the EQ is spot on exact with the Harman Curve, so he didn't make any errors in the choice of his filters:

I traced his measurement from his latest pdf using VirtuixCAD, and then applied his filters in REW to see the effect on the measurement. As you can see it's bang on exact to the Harman Curve, so his EQ is accurate at least as far as his measurement is.

 

[Robbo99999]

I'd like to offer you all an alternative EQ to the Oratory EQ, it's still done to the Harman Curve but it's using a different set of measurements. Oratory's measurement for the K702 has changed by a large amount over the various versions of pdf's he's released for it, and for my 4 units of K702 (at least) it's going in the wrong direction, with his measurements getting darker which in turn is resulting in his EQ's being brighter than the earlier versions.

Following is how his published measurement has changed from his first version compared to the latest version, the latest version is the dark highlighted line in the following graph, and you can see how much darker that measurement is in comparison:

Right, so Oratory measured one of my units of K702 (ages ago), and he sent me the measurements for that unit which was the same as his first published measurement (as I was the first person to send him a K702 to measure) - anyway I've measured that unit on my miniDSP EARS (ages ago) and was therefore able to create a conversion curve for turning a miniDSP EARs measurement into GRAS style measurement - that's valid a thing to do, but note that conversion curve is only applicable to that model of headphone (K702). I measured all 4 units of mine on my miniDSP EARS, averaged them & then converted into a GRAS style measurement, then did an EQ to the Harman Curve. I've tested the EQ on all 4 units of my K702 and it sounds good on all of them. The Oratory EQ is just too bright for my 4 units of headphone. In fact, my two newer K702 are brighter headphones than my two older K702, eventhough they've all got new pads on. With this in mind it's interesting that Oratory's measurement has become darker over the years, whereas my K702 units have shown that the newer ones are brighter than the older ones (so the opposite trend). So it's possible that the EQ I will show you will be more applicable for newer K702, it might be worth a try......but even on my older K702 I prefer my EQ. So that was my overall approach and overview, and following are the graphs/data to back it up.

My 4 units of K702 measured on miniDSP EARs (new pads)
You can probably just about make out in the following graph that the two newer K702 (Unit 3 & 4) are brighter than the older Unit 1 & 2.

and the same as above but showing the average line of all of them:


this average line was then converted to a GRAS style measurement and then EQ'd by me to the Harman Curve:

and the EqualiserAPO file for the EQ is attached at the end of this post. Note for those entering the EQ manually, you'll want a -9dB Negative Preamp.

Footnote: Units 1 & 2 have an aftermarket pad attached, which are made of same material as original pad and of same texture and appearance, I did measure to see what the difference was between aftermarket pad and stock pad (on the same unit) and they're practically identical or just a tad tad darker (if perceivable), see following graph:


Anyway, you're welcome to try my EQ based on my measurements which were backed up by an original GRAS measurement, and for my 4 units they are an improvement over the Oratory EQ, and there's a possibility that this EQ is even more relevant for newer K702 units, but I even prefer this one on my older K702 units dating back to 2015.

 

[roladyzator]

My K702 were bought in summer 2020 in Poland and they have less of a treble peak at 5.5 kHz, but instead some extra energy between 6 and 7 kHz. I also think they are the leaner variant. Sennheiser HD700 with Oratory's settings sounds bassier and those are probably more consistent between units than K702.

Another alternative is the Sonarworks EQ (I may post their EQ curve later if anyone's interested), which makes them noticeably darker.
Too dark for me.

I ended up with Oratory1990's settings but with a bit more bass, lesser reduction in 2 kHz and 5.5 kHz filters, lesser gain in 3.5 kHz filter to compensate for the two neighboring filters having less negative gain and an additional notch at 7 kHz.

Surprisingly, the HD700 with Sonarworks sound very similar to how they sound with Oratory's EQ, while other headphones I tried ended up darker in SW (Superlux HD-668B, AKG K371).

 

[Robbo99999]

K702 down to £89 on Amazon today, which is lowest price almost ever, so if you've been thinking you'd like to buy the K702 then today is the perfect time in the UK:

https://www.amazon.co.uk/AKG-K702-Open-Back-Reference-Headphones/dp/B001RCD2DW/ref=sr_1_1?crid=41FJSIDFZ7C6&keywords=B001RCD2DW&qid=1677343458&sprefix=b001rcd2dw%2Caps%2C65&sr=8-1

 

[isostasy]

@Robbo99999 Have you seen this? Confirms it is unit variation rather than silent revisions leading some K702 units to measure better than others.

https://twitter.com/i/web/status/1578484808120176640

 

[Robbo99999]

@Robbo99999 Have you seen this? Confirms it is unit variation rather than silent revisions leading some K702 units to measure better than others.

https://twitter.com/i/web/status/1578484808120176640

I have indeed seen that, and I have my own thoughts on that I'm not gonna talk about. There is for sure unit to unit variation within all headphones, some more than others.

 

[czt]

I like the K702. My first headphone. With Lexicon Alpha: "Headphone Output: (1) 1/8" stereo jack 20 mW per channel at 50 Ohms". Absolutely not hard to drive, I don't understand why so many (esoteric audiophile dac/amp maniac) claim the opposite.

 

[roladyzator]

The low impedance may fool some devices with impedance sensing to switch to lower output voltage limit.

My Samsung USB-C dongle does this, but a Shanling M0 (DAP/LDAC receiver with ESS chip) does not.

I expect laptops like Macbook M1 will also do this.

 

[Robbo99999]

I like the K702. My first headphone. With Lexicon Alpha: "Headphone Output: (1) 1/8" stereo jack 20 mW per channel at 50 Ohms". Absolutely not hard to drive, I don't understand why so many (esoteric audiophile dac/amp maniac) claim the opposite.

View attachment 276029

I wouldn't say K702 is hard to drive, but it's harder to drive than my other headphones: HD560s / HE4XX / NAD HP50 / HD600. If you use an EQ with some large boosts, in the bass for example then it makes them harder to drive, especially as K702 can often need a large bass boost. I think I use a -11dB Negative Preamp with my K702, that's because of something like a 9dB bass boost and me wanting to put in an extra 2dB of headroom to ensure my G6 DAC doesn't clip (a peculiarity with that DAC), although even then I don't need to use anything other than Low Gain on my headphone amp (no amplification of signal required beyond the voltage that the DAC provides).

 

[classy cheapo]

I have encountered many user reviews warning potential buyers about the poor quality control on the solder joints of the Chinese version k702s and warning them not to buy them if not confident with soldering. There are YouTube videos, Reddit posts...addressing this. Do the k612s also share this same issue? Is it better to stay away if I haven't soldered anything in my life and do not have the necessary equipment? Not interested in buying new, only second-hand.

I was recently re-soldering K612 (bought in the middle of 2022) cable for balanced cable and original factory soldering looked solid to me.

 

[Robbo99999]

I was recently re-soldering K612 (bought in the middle of 2022) cable for balanced cable and original factory soldering looked solid to me.

Was there just a blob of solder holding the wire onto the post, or was the solder wicked into the wire?

 

[classy cheapo]

Was there just a blob of solder holding the wire onto the post, or was the solder wicked into the wire?

It wasn't a blob. Looks like nice solid soldering covered by thermal tape. I didn't remove the tape but cut and soldered wires bit further up because factory soldering looked too nice to mess it up by my hack job.
Also, it doesn't really make sense for AKG to not fix such a simple quality control issue. On one hand there is their reputation and repair expenses, on another hand there is an angry message to factory to stop goofing for the next batches. Looks like an easy choice.

 

[Robbo99999]

It wasn't a blob. Looks like nice solid soldering covered by thermal tape. I didn't remove the tape but cut and soldered wires bit further up because factory soldering looked too nice to mess it up by my hack job.
Also, it doesn't really make sense for AKG to not fix such a simple quality control issue. On one hand there is their reputation and repair expenses, on another hand there is an angry message to factory to stop goofing for the next batches. Looks like an easy choice.

You'd have to remove the "thermal tape" (although it's probably shrunk rubber), to see the solder below to really know. Recent K702 now have shrunken rubber sleeves over the binding post soldered wires (I guess in a bid to reduce stress on the solder), but they still have the silly little blob of solder that hasn't been heated up enough to wick into the wire properly - which is the cause of the failures. The problem is that it's almost impossible to solder the wires on properly within the given space without melting a bit of the plastic surround (which is cosmetic), but I suppose they could use some ultra ultra fine soldering nibs to do it properly, although I'm not enough of a soldering expert to know if that would be possible. Earlier K702 didn't have the shrunken rubber sleeves over the binding posts - but my 3rd unit of K702 had the shrunken rubber sleeves and still that solder failed after someone yanked it off their head too aggressively.

 

[czt]

As I compared the unaltered EQ-s of Oratory, AutoEq and Sonarworks, not surprisingly found the last one is the most neutral to me and closest to my calibrated monitoring.

SoundID SR for Headphones: A new Headphone Tuning Standard - Sonarworks Blog

This White Paper covers data-based insights and a proposal for the new headphone calibration standard and its implications for the industry

www.sonarworks.com

 

[roladyzator]

For me the K702 with Sonarworks is too dark.
However, for Sennheiser HD700 the Sonarworks and Oratory1990's preset sound so similar I'd have trouble differentiating those. So one benefit to HD700's exotic construction is lower unit variation and measurement variation.

See my post on reddit, the last paragraph shows the Sonarworks EQ curve.

https://www.reddit.com/r/oratory1990/comments/10abm9k

 

[Robbo99999]

As I compared the unaltered EQ-s of Oratory, AutoEq and Sonarworks, not surprisingly found the last one is the most neutral to me and closest to my calibrated monitoring.

SoundID SR for Headphones: A new Headphone Tuning Standard - Sonarworks Blog

This White Paper covers data-based insights and a proposal for the new headphone calibration standard and its implications for the industry

www.sonarworks.com

It's possible that the Sonarworks sounds most neutral to you, but I wouldn't go as far to use the phrase "not surprisingly", I don't think we should expect Sonarworks to be more neutral than GRAS Harman.

For me the K702 with Sonarworks is too dark.
However, for Sennheiser HD700 the Sonarworks and Oratory1990's preset sound so similar I'd have trouble differentiating those. So one benefit to HD700's exotic construction is lower unit variation and measurement variation.

See my post on reddit, the last paragraph shows the Sonarworks EQ curve.

https://www.reddit.com/r/oratory1990/comments/10abm9k

It's really quite possible that HD700 has lower unit to unit variation that the K702, in fact it's probably quite likely, but I'm not too certain that you can use your Sonarworks experience comparisons as a means of setting that conclusion in stone.

 

[roladyzator]

It's possible that the Sonarworks sounds most neutral to you, but I wouldn't go as far to use the phrase "not surprisingly", I don't think we should expect Sonarworks to be more neutral than GRAS Harman.

It's really quite possible that HD700 has lower unit to unit variation that the K702, in fact it's probably quite likely, but I'm not too certain that you can use your Sonarworks experience comparisons as a means of setting that conclusion in stone.

Hi, you're absolutely right. I'm glad I got lucky with getting great sound of these.

I did read one of Oratory1990's comment where he stated that they do have consistent seal and low unit variation and that's why they make a good candidate for using them with a pre-computed set of filters. But of course I don't know how many units he or Sonarworks measured.