AKG K371 Impressions

[markanini]

The K612 are better.

View attachment 35412View attachment 35413

You are correct, they sound better too, even without EQ. The treble range is especially clear present and non-fatiguing.

 

[Erik]

oratory1990 has published AKG K361 and AKG K371 measurements.

 

[pwjazz]

oratory1990 has published AKG K361 and AKG K371 measurements.

View attachment 37162

View attachment 37163

Just tried the adjustments with my K371, and as usual don't love them. It sounds a bit artificial and thin to me. I know that I'm not 100% sold on the Harman bass response, and for whatever reason I think oratory's measurements always show a bigger dip at around 4 KHz than I hear so when I apply his adjustments it always sounds overly boosted to me.

 

[LeftCoastTim]

oratory1990 has published AKG K361 and AKG K371 measurements.

...

This is why I only trust Oratory1990 measurements.

 

[Spech]

This is why I only trust Oratory1990 measurements.

I've taken a look over Oratory's measurements and they seem pretty strange to say the least, best measurements I've seen so far are RTINGS's ones

 

[Fregly]

Are the Oratory measurements reliable or not? From those graphs the 371 seems the one to get. The sub tones are lifted, yet would that not just add ambient weight instead of the typical bloat pushng into the mids? I was going to get the 361 and then I see these charts.

 

[LeftCoastTim]

Are the Oratory measurements reliable or not? From those graphs the 371 seems the one to get. The sub tones are lifted, yet would that not just add ambient weight instead of the typical bloat pushng into the mids? I was going to get the 361 and then I see these charts.

Oratory is an audio engineer with access to state of the art measurement equipment (see bottom of the measurement sheet) and knowledge to use it properly. I haven’t seen anyone else on the Internet with that kind of equipment.

 

[solderdude]

Consider the following:

Measuring headphones is NOT remotely as accurate as measuring electronics and even measuring speakers.
It can provide some clues, that's all.

One measures the same headphone on 4 different test rigs.
Pick one headphone you know how it sounds and is well documented.
Overlay these measurements.

Largely, from 100Hz to 1kHz they all measure roughly the same. Small differences can be found. And with small I mean <3dB.
Look at the plots above 1kHz... compare those.
They all differ and can do so substantially (10dB is no exception).
Pick one... EQ so it measures 'flat' and be in awe how great those measurements look.
That MUST be correct as measurements show it is 'near perfect'.... right ?

Now... measure the same headphone with the same EQ on the other rigs.... will it be as flat as measured on the first rig.
Hell no ... it will be off in certain places while other frequencies may be closer or even further removed from the uncorrected plot.

Now comes the difficult question... which one to trust ?

Assuming Oratory is the only correct one and then Tyll, Rtings, Sonarworks, SBAF, many others and my measurements are all wrong.
yet some of the above also know what they are doing and some also use expensive HATS, others use the (flawed) EARS or DIY rigs.

The following question is why all rigs measure differently and who's rig is closest to the truth.
Where would one EQ on ? even on peaks and dips that clearly appear to be from poor correctrion/compensation/resonances that aren't corrected properly ?

 

[JJB70]

When it comes to EQ surely there is no "right" or "wrong" in that you adjust to your own preferred sound signature. My own preference is for relatively flat but with a bit of brightness in the treble most of the time, but I'll admit that when I'm in the right mood I kind of like the stock tuning of my Sony MDR Z7's which have a "Beats lite" type of sound.

 

[LeftCoastTim]

Consider the following:

Measuring headphones is NOT remotely as accurate as measuring electronics and even measuring speakers.
It can provide some clues, that's all.
....

Now comes the difficult question... which one to trust ?

Since I trust Sean Olive’s research, that means measurements done on the same equipment with the same protocol.

 

[solderdude]

So all others are considerably flawed and those 2 are not ?
In any case when EQ is based on one rig it will be fairly similar on the other... but what about ones ears ?
Are they the same as those rigs and why aren't other test rigs accurate ?

 

[nhunt]

I’ve corresponded with oratory1990 a bit and actually sent him this K371! This is his part of his day job and he does the measurement on a GRAS setup similar to Sean Olive’s research. Sean Olive told me the K371 measure within a dB of the target and this measurement looks like it just barely agrees

Since measurement rigs can vary, as solderdude points out, getting apples to apples measurements seem important if you are trying to EQ it to the same measured curve from the research. But I don’t think they are the only measurements that can be used for analysis. I often check Sonarworks too since they will measure many samples and average them. And solderdude’s site for more measurements. Rtings for soundstage, consistency, various other comfort things.

 

[markanini]

Oratory is an audio engineer with access to state of the art measurement equipment (see bottom of the measurement sheet) and knowledge to use it properly. I haven’t seen anyone else on the Internet with that kind of equipment.

He actually designs transducer systems for products that get sold under big brand names.

 

[nhunt]

So all others are considerably flawed and those 2 are not ?
In any case when EQ is based on one rig it will be fairly similar on the other... but what about ones ears ?
Are they the same as those rigs and why aren't other test rigs accurate ?

Yeah, personally even after EQing to the curve I then do some sweeps and fussing around to get it right for me and that particular set of headphones. oratory1990 does, too. The curves that oratory1990 and Sonarworks use just seem like a good consistent starting point. The target doesn’t say anything above 10k, either, so you are on your own there.

Something I have become increasingly paranoid about is unit to unit variation. I have two Oppo pm-3, the black pair and the white and blue pair, and my wife and I are convinced they have different FR and the white and blue sound nicer. This is actually something we can practically blind test. Could be pads or clamping.

I‘m interested in sending a pair of headphones in to Sonarworks to get them individually calibrated and see the results.

 

[markanini]

So all others are considerably flawed and those 2 are not ?
In any case when EQ is based on one rig it will be fairly similar on the other... but what about ones ears ?
Are they the same as those rigs and why aren't other test rigs accurate ?

If a set of measurements is done on a newer generation rig, a piece of hardware specifically prescribed by the main researcher in the field, then yeah. It stands to reason it should be trusted more.

 

[Severian]

I've never much liked the sound of the full battery of oratory1990's filters on my headphones. It tends to be a little... tinny? Unnatural? And the filters often have large amplitudes that call for substantial negative preamp gain that can necessitate driving a headphone amp awfully hard. I would be curious to try out his filters on a headphone like the K371 that needs less dramatic adjustment than many headphones to hit the target curve.

My manual approach to EQing headphones is to tame any bad treble peaks that I see across multiple measurements (I'm not great at hearing these myself) and a low shelf of a few dB at 200Hz. A few slight adjustments to the mids to tune vocals to my liking.

 

[solderdude]

Something I have become increasingly paranoid about is unit to unit variation. I have two Oppo pm-3, the black pair and the white and blue pair, and my wife and I are convinced they have different FR and the white and blue sound nicer. This is actually something we can practically blind test. Could be pads or clamping.

Swap pads on them and test truly blind. Could also be one has slightly more squished pads or somewhat higher efficiency or a silent revision.... or production spread.

Sonarworks does not appear to use an expensive HATS . His measurements (apart from bass correction) largely are similar to mine.

 

[Spech]

In my experience using EQ is pretty tricky sometimes a headphone just sounds 'better' even tho it sounds flawed as in EQ-ing it to make it flat makes it worse

And also using Equalizer APO I find that only using subtractive EQ works, if you add you have to either lower the pre-amp which then in turn lowers the sound level and makes the headphones sound weak and subdued(and turning the volume up doesn't work in making them sound the same it's weird) or dont lower the pre amp and have idk what to call it its not compression or clipping its something weird.

But some EQ's (like my phone's EQ) seem to deal with this problem as in you can add as much as you want and it won't have that issue, quite strange.

 

[pwjazz]

1. I'm super grateful for the work that oratory1990 does. Heck, he's even sent me some raw data on occasion when I've asked for it, and he doesn't get a dime for what he's doing (afaik).

2. oratory1990 uses an industry standard rig and knows what he's doing

All that said, IMHO even industry standard measurements of headphones have inaccuracies in the treble region (say 4 KHz and above). Whenever I strictly correct treble to those measurements, whether they're from oratory1990, @solderdude , my own MiniDSP E.A.R.S. measurements or whatever, I'm never happy with the result. I've found that the measurements can be good in cluing you in to where there's a peak or a dip and maybe it's relative magnitude, but if you go whole hog on canceling that peak or dip you'll likely end up with somewhat unnatural results.

Also, there's acoustic stuff going on that I don't properly understand, but some dips can actually be caused by a resonance and EQ'ing those dips down rather than up can sometimes seem to me like it makes things sound better.

 

[solderdude]

I've found that the measurements can be good in cluing you in to where there's a peak or a dip and maybe it's relative magnitude, but if you go whole hog on canceling that peak or dip you'll likely end up with somewhat unnatural results.

This..