Has the meta on IEMs moved away from Harman's "pinna gain"?

[Limopard]

While my preferences for over-ears roughly match with the Harman curve I found its in-ear equivalent mostly too bright and biting. I tried to adjust my UE900 (which has a quite "wild west" response) to Harman - it was a different kind of bad.

 

[muza_1]

 

[muza_1]

Well now that I have tried with the new Crin tool with an accurate and customizable to preference target (JM-1: Bass +10, Tilt -0.8 for me) all I can say is that first, I have been doing it wrong and the with wrong targets in the past but with this target and measurements from Paul Wasabi (PW) all I can say is that I like it and it took just a little EQ in comparison with past attempts.
The IEM is the CCA Trio/KZ Vader Balanced with all switches up.

Lowering the area from 3500Hz to 6000Hz balanced the whole presentation much better than just lowering the pinna (3000Hz) region, which at first I thought it was a bit
elevated.

I'm going to try with different IEMs and EQ just to be sure but for the time being it sounds good.

 

[InfiniteJester]

Great video. JM1 DF still has way too much ear-gain to sound neutral to me. It is funny that its creator think exactly the same.

 

[muza_1]

Now I know why I like it especially compared to other (bad) attempts I made with the "meta" tuning, I have always liked the Harman target with -2dB or so less of highs, as it turns out the JM-1 target properly displayed and adjusted (tilt and bass) to my preference is not that far off (just a few dB less than I'm used to).

Somehow I always missed it when comparing targets in other squigs and tools, it might have been that for most it's a fixed target (or a variation) and trying to add a bass shelf, tilt and then center and overlay it over the Harman target was not as intuitive as with this one (there is also the small difference in tilt in my preference).

And there is the so called "meta" tuned IEMs like the Aful Explorer that are not tuned to the JM-1 they just have a general lower pinna gain or Mega5EST that are DF tuned and some other variations and that makes "meta" much more difficult to judge.
I tried and returned the Explorer thinking "meta" is not for me but are they "meta" because they are vastly different from other "meta" IEMs. Is the JM-1 target "meta"?

 

[anaveragebear]

I've found the JM1 with a -.8dB/Oct works well for me from 3-10kHz, but I like more highs cuz I like the sense of "sparkle" and "air" and I also prefer no tilt down to 200Hz. I don't know if I fully agree with the idea of tilting the response to match speakers in a room because, well, IEMs aren't really in a room, but I am pretty confident that research will show in time that this is "the right idea."

 

[markanini]

I've found the JM1 with a -.8dB/Oct works well for me from 3-10kHz, but I like more highs cuz I like the sense of "sparkle" and "air" and I also prefer no tilt down to 200Hz. I don't know if I fully agree with the idea of tilting the response to match speakers in a room because, well, IEMs aren't really in a room, but I am pretty confident that research will show in time that this is "the right idea."View attachment 393326

I would bet that more will prefer adjustments different for what's currently offered. More highs is something I prefer myself. I don't need a lot of bass, but when I want some bass it works better, relative to JM-1, if the mids have some V-shape to them, and highs are a bit elevated. Otherwise, the overall presentation comes off too wooly to me.

I think at this point the future of Meta depends on making an effort to studying preference. JM-1 as it is now will lead to outcomes parallel to Harman 2019. Some brands will adopt Meta on select models, but for the most part do their own thing w.r.t. frequency response. The Meta honeymoon will eventually pass and Meta tuned IEMs will start getting criticized for X, Y and Z issues to their tuning.

 

[Jimbob54]

Now I know why I like it especially compared to other (bad) attempts I made with the "meta" tuning, I have always liked the Harman target with -2dB or so less of highs, as it turns out the JM-1 target properly displayed and adjusted (tilt and bass) to my preference is not that far off (just a few dB less than I'm used to).

View attachment 393187

Somehow I always missed it when comparing targets in other squigs and tools, it might have been that for most it's a fixed target (or a variation) and trying to add a bass shelf, tilt and then center and overlay it over the Harman target was not as intuitive as with this one (there is also the small difference in tilt in my preference).

View attachment 393189

And there is the so called "meta" tuned IEMs like the Aful Explorer that are not tuned to the JM-1 they just have a general lower pinna gain or Mega5EST that are DF tuned and some other variations and that makes "meta" much more difficult to judge.
I tried and returned the Explorer thinking "meta" is not for me but are they "meta" because they are vastly different from other "meta" IEMs. Is the JM-1 target "meta"?

View attachment 393194

I might be wrong but I dont think what you've done in your first graph is "right"

You cant overlay a 711 based target on a 5128 target and simply compare- the rigs have different responses.

Variations and Nova are about the most Harman 2019v2 compliant IEM out there. Your first graph indicates a pure H2019 compliant IEM should align to your JM1 +10 -0.8 curve apart from the deviations 2-8k (below) and 8+ (above)

But that aint so on Crins 5128 measurements of those IEM. But I will admit my grasp of this, especially when using the different rigs is shaky at best!

For reference the Project Meta is pretty much bang on JM1 +10 -0.8, shown graphed on the 711 vs the more Harman Nova and VAriations. The midrange hump is way lower in the JM1 meta

 

[muza_1]

I might be wrong but I dont think what you've done in your first graph is "right"

You cant overlay a 711 based target on a 5128 target and simply compare- the rigs have different responses.

Variations and Nova are about the most Harman 2019v2 compliant IEM out there. Your first graph indicates a pure H2019 compliant IEM should align to your JM1 +10 -0.8 curve apart from the deviations 2-8k (below) and 8+ (above)

But that aint so on Crins 5128 measurements of those IEM. But I will admit my grasp of this, especially when using the different rigs is shaky at best!

For reference the Project Meta is pretty much bang on JM1 +10 -0.8, shown graphed on the 711 vs the more Harman Nova and VAriations. The midrange hump is way lower in the JM1 meta

View attachment 393334

View attachment 393336

Yes you are correct, each target is designed for an specific coupler, I should have exported the 5128 measurements and EQ based on that, I'm going to try and see how much different the curves end up being.

I was just fooling around with the tool and completely blank out on that.

 

[InfiniteJester]

I've found the JM1 with a -.8dB/Oct works well for me from 3-10kHz, but I like more highs cuz I like the sense of "sparkle" and "air" and I also prefer no tilt down to 200Hz. I don't know if I fully agree with the idea of tilting the response to match speakers in a room because, well, IEMs aren't really in a room, but I am pretty confident that research will show in time that this is "the right idea."View attachment 393326

Harman IE is physically painful to me. JM-1 has been a blessing, despite not being perfect.

 

[Jimbob54]

Yes you are correct, each target is designed for an specific coupler, I should have exported the 5128 measurements and EQ based on that, I'm going to try and see how much different the curves end up being.

I was just fooling around with the tool and completely blank out on that.

Youre right that the 4-8khz area is too high on Harman, and the 3k summit is too much too. (Caveat- for my ears and listening levels)

My suspicion is Harman might sound good to people that listen comparatively quieter but when you notch it up- things go sideways. Zero evidence to base that on of course.

 

[InfiniteJester]

Youre right that the 4-8khz area is too high on Harman, and the 3k summit is too much too. (Caveat- for my ears and listening levels)

My suspicion is Harman might sound good to people that listen comparatively quieter but when you notch it up- things go sideways. Zero evidence to base that on of course.

But if you lower it too much you stop being able to hear the mid-range.

 

[markanini]

Youre right that the 4-8khz area is too high on Harman, and the 3k summit is too much too. (Caveat- for my ears and listening levels)

My suspicion is Harman might sound good to people that listen comparatively quieter but when you notch it up- things go sideways. Zero evidence to base that on of course.

Bassheads like Harman 2019 too. Look at what Hawaii Bad Boy says about Truthear Zero Red vs Blue. He also says a bunch of good things about Truthear Nova.

 

[muza_1]

Youre right that the 4-8khz area is too high on Harman, and the 3k summit is too much too. (Caveat- for my ears and listening levels)

My suspicion is Harman might sound good to people that listen comparatively quieter but when you notch it up- things go sideways. Zero evidence to base that on of course.

There is a problem with the tilt of the JM-1 target and the various sources, which one is the correct one?

Yellow is the one used by majority of Squig pages and it should have a tilt of -1, the other two I took form AutoEq target section in GitHub.


Also there is the problem with the measurements, both the 5128 (converted?) and 711 (uploaded) for the CCA Trio are practically the same, aren't they suppose to be different?

EDIT. I added the graph with bass boost.

 

[Jimbob54]

There is a problem with the tilt of the JM-1 target and the various sources, which one is the correct one?

View attachment 393411View attachment 393415
Yellow is the one used by majority of Squig pages and it should have a tilt of -1, the other two I took form AutoEq target section in GitHub.


Also there is the problem with the measurements, both the 5128 (converted?) and 711 (uploaded) for the CCA Trio are practically the same, aren't they suppose to be different?

View attachment 393413

EDIT. I added the graph with bass boost.

Like I say, Im shaky at best with this.

There are clearly 4 very different curves there and I think they are all intended to be different- either by tilts or bass/ treble filters or both (this is part of the problem with new "standards"- they arent standardised.

Re the measurements that purport to be compensated from 711 to 5128- think youve just shown how that might not be reliable. Which is why I am sticking to Crins 5128 measurements and target for 5128 and squig generally for 711 measurements and targets.

For my money the Crin JM1 with an 8ish db bass boost and a (roughly) -0.4 tilt is where I am happy for now - but that may only be applicable to his 5128 actual measurements.

 

[muza_1]

There are clearly 4 very different curves there and I think they are all intended to be different- either by tilts or bass/ treble filters or both (this is part of the problem with new "standards"- they arent standardised.

Exactly.

Re the measurements that purport to be compensated from 711 to 5128- think youve just shown how that might not be reliable. Which is why I am sticking to Crins 5128 measurements and target for 5128 and squig generally for 711 measurements and targets.

I also think Crin ones are some of most reliable at the time, the down side is there is a limited selection of measured IEMs (considering the amount of available models).

From what I've tried (and trying now with other IEMs measured by him) I think I prefer a tilt of -0.8 with a bass boost of 10dB but I tend to listen outside and not loud.

 

[markanini]

The JM-1 curves are different because they account for coupler difference. The correct one is the one is the one used for the individual database. In other words, use the selectable JM-1 with Paul Wasabis squig link, use the selectable JM-1 with Gizaudios squig link and so on.

 

[muza_1]

The JM-1 curves are different because they account for coupler difference. The correct one is the one is the one used for the individual database. In other words, use the selectable JM-1 with Paul Wasabis squig link, use the selectable JM-1 with Gizaudios squig link and so on.

I tried but I ended with a huge mid-bass hump and dark sound I still think the tilt is not implemented correctly.

It was supposedly implemented this way for ease of use and interchangeable results, I think I'll stick with 711 and Harman (adjusted to my preference) for the time being, it's has been more reliable and I had good results EQing form different sources and different IEMs in the past. Also I've been getting good results with the new Crin Squig using his measurements with tilt and boost to my preference (with the limited experimentation at the moment).

Edit. Crin and AutoEq are one of the most popular sources when it comes to 5128 and their targets are closer to each other, AutoEq uses 3rd party measurements and is supposed to be a correct and reliable way to do it, at least as far as I know, why implement a system that only works with the tool using it when the already implemented one is standardized (even with some low cost clones).

 

[InfiniteJester]

A tilt of -1.5 seems to sound reasonably natural for me. My HRTF must be insane. Everything sounds always too shouty and bright. I have a vory big head and hears, I wonder if that could be the cause.

 

[markanini]

I tried but I ended with a huge mid-bass hump and dark sound I still think the tilt is not implemented correctly.

That means you don't like JM-1 with 10dB tilt. Nothing strange, JM-1 is just a starting point to add your preference adjustments on top of.