The IEM Harman Target 2019 sounds "off" to me. Is it just me?

[dasdoing]

From what it seems the internet/community kind of forced the target into a standard. manufactures don't have much choice once this happens

the problem is that as a subjective target it will always be changing lol

 

[Chromatischism]

I push this topic because Sean Olive just confirmed on Twitter (X) that there is a new Harman In-Ear Target on the way.
His statement implies that it might be closer to Crinacles "Red" IEM (I don't own this IEM).

Oratory1990: "Friends of mine are using the @TruthearLab Zero/Zero:Red for on-stage in-ear monitoring. This gave me a chance of measuring them. They broadly follow the Harman Target curve."
Sean Olive: "We are working on a new in-ear HARMAN target curve so it may not "broadly" follow it soon."

https://twitter.com/i/web/status/1710432330098319804

Interesting

 

[Doltonius]

The Harman Target is just an average of a population sound preference, gives you a quite important information and is a great achievement. But mathematically speaking most of us will fall around it, not necessarily in the death center, your preference will fall barely off to greatly off from it., it is expected. Nothing wrong with that.

However, you will struggle to find iems tuned to have even more upper mid range than the harman target. I have the feeling the iem target is actually at the upper bound of population preference range in the upper mids, rather than in the middle of the range. Many people feel that the upper mid range is too much and call it shouty.

 

[CedarX]

However, you will struggle to find iems tuned to have even more upper mid range than the harman target. I have the feeling the iem target is actually at the upper bound of population preference range in the upper mids, rather than in the middle of the range. Many people feel that the upper mid range is too much and call it shouty.

Any data or study to support your statements?

 

[Chromatischism]

Any data or study to support your statements?

Check out the other preference targets at Squiglink. It's worth asking why none of the other IEM reviewers' targets have an upper mid-range/treble as high as Harman.

 

[CedarX]

Check out the other preference targets at Squiglink. It's worth asking why none of the other IEM reviewers' targets have an upper mid-range/treble as high as Harman.

Are other reviewer’s targets backed by any science? I don’t dispute the “authority” of these reviewers, but how does it make the Harman IEM target wrong as a preference basis?

 

[MacClintock]

Are other reviewer’s targets backed by any science? I don’t dispute the “authority” of these reviewers, but how does it make the Harman IEM target wrong as a preference basis?

Well, true, the other reviewer‘s targets are not exactly science, but somehow contribute to the listener’s preference average. And the new Harman 2024 beta has even more elevated ear gain region. It is not clear to me why. But then, I think Sean Olive mentioned that just 13 persons participated in the study, weighing this against the many complaints asks for caution in the judgements. I must say that I tried also the SoundGuys target, which faired equally good as both Harman targets and liked it quite a bit.

 

[Chromatischism]

Are other reviewer’s targets backed by any science? I don’t dispute the “authority” of these reviewers, but how does it make the Harman IEM target wrong as a preference basis?

We're not making claims like "backed by science". But consider that what you're referring to is a target derived, in theory, by averaging a statistically significant sample of people. If you then look at a lot of other targets in use by others and find all of them have less upper mid-range, leaving Harman as an outlier, something is wrong.

 

[Doltonius]

Are other reviewer’s targets backed by any science? I don’t dispute the “authority” of these reviewers, but how does it make the Harman IEM target wrong as a preference basis?

A good basis should be in the middle of the preference range. But Harman just doesn't seem like it is in the middle, in terms of its upper mids. I have never seen a single reviewer reference curve or a single person online whose favorite iem having more elevated upper mids than the Harman target; wanting the Harman target as is, is already at the extreme, e.g., Amir and Sharur. Also, the Harman results have been out for years. Yet, manufacturers just so rarely release iems with more upper mids than the Harman target. Even iems marketed as "Harman-tuned' usually still fall a little bit short somewhere in the 2k-8khz range, compared to the Harman, e.g., Variations, Chopin, Zero Blue, etc. Probably only the Nova with foam tips truly get there. But considering that most people use the silicone tips, Truthear probably tuned the iem according to its response on the silicone tips, which means that it by default (with silicone tips, that is) also falls short of the Harman target between 3khz and 7khz. If the Harman is really in the middle of the preference range, we wouldn't be in this situation.

 

[CedarX]

So I take that the opinion of multiple reviewers trumps whatever science Harman puts to develop their IEM target.

Another point that puzzles me: if you align the base Harman target to the others in the highs, then it’s not so much “shooty”, but more “bass-light”… unless you are in this Harman group who like more bass and you bump the bass… then it becomes essentially V-shaped as compared to the others…
So what’s off with the Harman IEM target?

 

[Doltonius]

So I take that the opinion of multiple reviewers trumps whatever science Harman puts to develop their IEM target.

Another point that puzzles me: if you align the base Harman target to the others in the highs, then it’s not so much “shooty”, but more “bass-light”… unless you are in this Harman group who like more bass and you bump the bass… then it becomes essentially V-shaped as compared to the others…
So what’s off with the Harman IEM target?

If what others have said is true, i.e., that the Harman IE target was developed with only 13ish trained listeners with upper-mid and treble levels that cannot be adjusted, and its competitors in the testing seem to be targets that are very different, I really don't think there is much science to back Harman's target up. And it is not just about the reviewers. The audience, and all the manufacturers, no one goes above Harman in the upper mids. Additionally, I am sure that Apple have done their own research, just didn't release the results. Their best-selling TWS's don't have Harman-level upper mids. Apple would sure want to know that is going to be the popular preference and they have the resource to conduct more comprehensive research than Harman. Also interesting is that fact that Samsung, who owns Harman, also doesn't tune its TWS to the Harman-level upper mids, despite good compliance in the bass and up to 3khz.

What I am talking about is the region between 3khz and 8khz (I guess many will no longer call this upper mids, but low treble, but this is what I mean). You can try to align the iems and the Harman target at 3khz. Then observe the shape of the curve after 3khz and before 8khz. Harman is more convex (elevated, bulging upwards) compared to most iems, including those claiming to be Harman-tuned. Many iems and TWS can conform to the Harman very well, before 3khz; at around 3khz, they might be even slightly more elevated (normalized with volume). But afterwards, they will usually fall short of Harman.

Also shoutiness is exactly determined the balance between bass and upper mids.

 

[CedarX]

If what others have said is true, i.e., that the Harman IE target was developed with only 13ish trained listeners with upper-mid and treble levels that cannot be adjusted, and its competitors in the testing seem to be targets that are very different, I really don't think there is much science to back Harman's target up.

@RayDunzl, post #19, mentions 283 individuals in 4 countries…. If this is true, I think there is more science to back it up.

Also shoutiness is exactly determined the balance between bass and upper mids

Again, based upon what “reviewers” are saying, shoutiness is about female upper vocals for example… not a balance thing at all.

I guess I’ll have to admit that I just can’t see the science that unequivocally demonstrates the Harman IEM target is off… and move on!

 

[MacClintock]

@RayDunzl, post #19, mentions 283 individuals in 4 countries…. If this is true, I think there is more science to back it up.

That was for the over-ear target, the IEM target was done with much fewer people, less than 20, all Harman employees, as told by Sean Olive in a recent video. And the options were also much more restricted than for the over-ears.

Again, based upon what “reviewers” are saying, shoutiness is about female upper vocals for example… not a balance thing at all.

I guess I’ll have to admit that I just can’t see the science that unequivocally demonstrates the Harman IEM target is off… and move on!

I don't think the Harman2019IEv2 target is bad, but when listening for a longer time, or, more importantly, cranking up the listening level, I have the urge to tone down the 2-6 kHz region a bit, maybe by 2-3 dB. Then it becomes much more pleasant.

 

[RayDunzl]

@RayDunzl, post #19, mentions 283 individuals in 4 countries…. If this is true, I think there is more science to back it up.

So I put what I found way back when into google to see if I could find it again.

Search for
Dr. Olive and his researchers conducted blind studies on 283 individuals in four countries and 11 test locations. Although most of the participants were Harman employees, the science team was careful to achieve diversity in their gender, age, and listening experience.

Here's what it came back with:

 

[Doltonius]

@RayDunzl, post #19, mentions 283 individuals in 4 countries…. If this is true, I think there is more science to back it up.


Again, based upon what “reviewers” are saying, shoutiness is about female upper vocals for example… not a balance thing at all.

I guess I’ll have to admit that I just can’t see the science that unequivocally demonstrates the Harman IEM target is off… and move on!

As @MacClintock says, that person mixed up the number of participants in the OE and IE studies. The in-ear studies are done on a much smaller scale and with restricted options.

The Harman target is supposed to be an "average preference curve." For a given region in the frequency response, i.e., 3khz-8khz, almost all observable preference in the actual audience in the real world have preferences below the Harman curve. I have not seen a single person who says that Harman upper mids are still too tame. How could an "average preference curve" behave like that? If it is right at the average, then there will be people (in fact, half the people!) complaining this point about the Harman target, because they would prefer more. But that is so far from the reality. You only ever see people who say it is too shouty or just right.

Indeed, shoutiness is perceived most clearly with female vocals. However, for a given pair of iems, if you find it shouty, just add more bass (and/or lower midrange). You will find it less shouty immediately. Therefore it still has to do with balance. Frequency response is all about relative amplitude between the frequencies, i.e., balance.

 

[Ze Frog]

To be honest the Harman Curve likely isn't really aimed at audiophiles perse, more a setting that is designed to appeal to most people instead. That's not to say there's anything wrong with the Harman Curve, but different strokes for different folks really. What seemingly works for the massive majority always tends to be a little bit safe in many areas of life.

 

[CedarX]

I have not seen a single person who says that Harman upper mids are still too tam

I don’t want to put @Doodski on the spot, but take a look at his “typical EQ”:
https://audiosciencereview.com/foru...rent-are-that-many-confused.9245/post-1928760

I don’t know if he uses IEMs… But if he does, he may very well be that single person you’ve never seen before !!!
This is anectodal evidence… with no data to support it, but how is it different, or of less value from your demonstration? Don’t we need a little more evidences to support all of this?

[End of my rant - Was just trying to defend Science, not Harman ]

 

[markanini]

To be honest the Harman Curve likely isn't really aimed at audiophiles perse, more a setting that is designed to appeal to most people instead. That's not to say there's anything wrong with the Harman Curve, but different strokes for different folks really. What seemingly works for the massive majority always tends to be a little bit safe in many areas of life.

It's goal is defeating the circle of confusion: https://seanolive.blogspot.com/2009/10/audios-circle-of-confusion.html?m=1

 

[Doltonius]

I don’t want to put @Doodski on the spot, but take a look at his “typical EQ”:
https://audiosciencereview.com/foru...rent-are-that-many-confused.9245/post-1928760

I don’t know if he uses IEMs… But if he does, he may very well be that single person you’ve never seen before !!!
This is anectodal evidence… with no data to support it, but how is it different, or of less value from your demonstration? Don’t we need a little more evidences to support all of this?

[End of my rant - Was just trying to defend Science, not Harman ]

This is pointless if we don't know what gear they listen on. As I said, if Harman IE is supposed to be the average preference curve, you should see half the people complaining that its upper mids are too tame. Now you just give me this one person where we don't even know what type of equipment they listen to??? And I am afraid the PEQ is for their speakers, if you look closely at the image they posted.

Harman's IE target was not based on solid science, to begin with. There is barely any science to defend. If the 13ish trained testers can't even adjust the upper-mids and treble, you can't legitimately call the IE Harman target science-based average preference curve of the general population. And as it doesn't actually predict the average preference of the audience in the real world, it means that its insufficient methodological rigor does have bad empirical repercussions. Bad science, at best, if at all considered science.

 

[Chromatischism]

Indeed, shoutiness is perceived most clearly with female vocals. However, for a given pair of iems, if you find it shouty, just add more bass (and/or lower midrange). You will find it less shouty immediately. Therefore it still has to do with balance. Frequency response is all about relative amplitude between the frequencies, i.e., balance.

Agree with everything you're saying except this. Increasing bass is a crutch and doesn't completely solve the problem, but creates another--now the mids are pushed to the background.