Harman preference curve for headphones - am I the only one that doesn't like this curve?

[dasdoing]

how does anybody kow how flat sounds?
unless you were created with the perfect speaker in an anchoic room you wont know.
flat is not equal percieved loudness of test tones

 

[dasdoing]

flat in Harman terms means "sounds like an average room"

 

[aldarrin]

It does sound 'flat'. That's the point.

It can sound flat to you if that's what you are used to. But flat + bass boost isn't flat, it's a colored sound that many people find pleasing.

I think "flat" or neutral for headphones looks something like this green line. The dotted line is Harmon's 2018 OE target. The added bass shelf is fun, but not what I'd call flat.

 

[HereYaGo]

It can sound flat to you if that's what you are used to. But flat + bass boost isn't flat, it's a colored sound that many people find pleasing.

View attachment 110778
I think "flat" or neutral for headphones looks something like this green line. The dotted line is Harmon's 2018 OE target. The added bass shelf is fun, but not what I'd call flat.

I found the ear sensitivity chart going up in the bass interesting. i dont know much about it tho really. just i notice it resembles a headphone with a treble boost, wide soundstage, and bass boost.

 

[bluefuzz]

it's a colored sound

Try it for yourself. Listen to a series of 1/3 octave sine tones or band limited pink noise and hear what sounds even to you ...

 

[bluefuzz]

The added bass shelf is fun, but not what I'd call flat.

It's not an 'added bass shelf'. If you actually listen to a sine series on a headphone EQ'ed to the Harman target the bass will still be appreciably less loud than the midrange.

 

[aldarrin]

Try it for yourself. Listen to a series of 1/3 octave sine tones or band limited pink noise and hear what sounds even to you ...

I did using my HD800S EQ'd and measured to Harman 2018 OE. Harman sounds V-shaped to me.

See slide 10:
http://www.juloaudio.sk/Umiestnenie_reprosustav/History of Harman Target Curve.pdf

Green dotted line is "flat sounding like speakers in the room." Black what people thought sounded good when they could play with the bass & treble. Which one looks like the 2018 OE target?

 

[dasdoing]

Try it for yourself. Listen to a series of 1/3 octave sine tones or band limited pink noise and hear what sounds even to you ...

again, what sounds even in this exercise is not flat: https://en.wikipedia.org/wiki/Equal-loudness_contour

 

[wolf_walker]

I'm usually pretty happy with Oratory1990's EQ's, at most they only require a little tweaking for my half-worn out ears.
That's usually a little bump at 4Kish and some more bass, but it's a bit track/mixing specific too.
Nothing is universal, or a blanket for everyone. For whatever reason, my ears don't seem to like crinacle presets,
they just sound funky, but I'm sure they are quite competent.
Sometimes, I want more of one thing than another, mood, or music in question, but that's why
Peace has easily changeable presets I suppose.
It's a many splendored audio world we live in.

 

[bluefuzz]

Green dotted line is "flat sounding like speakers in the room."

No it isn't. Read that slide again (bold emphasis mine):

The green dotted curve is response of the loudspeaker equalized to a flat in-room curve. Listeners did not like this baseline curve and adjusted the bass 6.6 dB higher and the treble -2.4 dB lower. More evidence that the in-room loudspeaker target should have a 9-10 dB downward slope from 20-20 kHz

In other words not 'flat sounding like speakers in a room', but artificially measuring flat. Listeners adjusted the bass to sound more 'flat sounding', i.e. with a downward slope.

 

[andreasmaaan]

No it isn't. Read that slide again (bold emphasis mine):

In other words not 'flat sounding like speakers in a room', but artificially measuring flat. Listeners adjusted the bass to sound more 'flat sounding', i.e. with a downward slope.

You're absolutely correct that the green curve is the (smoothed) response of a pair of speakers EQ'd to have a flat steady-state in-room response (although some comments from Dr Olive earlier in the thread regarding windowing seem to cast just a shade of doubt on this, it's probably worth mentioning).

The thing is though, the Harman target is not by design an attempt at a "flat" or "flat-sounding" curve. If your read the methodology by which the curve was derived, you'll see that it's an average preference curve; nothing more nor less.

Also note that the in-room steady-state response at the listening position of a neutral speaker varies wildly with speaker dispersion characteristics, room size/acoustics, and listening distance, so it's just not possible to say that a particular downward-sloping response is the response of a neutral speaker in a room.

That the Harman curve happens to quite closely resemble the in-room steady-state response of a neutral, direct-radiating box monopole in Harman's listening room is an interesting coincidence, but it's not possible to read too much into it.

 

[bluefuzz]

the Harman target is not by design an attempt at a "flat" or "flat-sounding" curve. [...] you'll see that it's an average preference curve;.

My point (if I have one) is that it effectively amounts to the same thing. Most people prefer a slightly downward sloping response from loudspeakers in a room and the Harman headphone curve approximates that response - which unsurprisingly sounds flat-ish ...

 

[andreasmaaan]

My point (if I have one) is that it effectively amounts to the same thing. Most people prefer a slightly downward sloping response from loudspeakers in a room and the Harman headphone curve approximates that response - which unsurprisingly sounds flat-ish ...

Ok, but my point is that:

the in-room steady-state response at the listening position of a neutral speaker varies wildly with speaker dispersion characteristics, room size/acoustics, and listening distance, so it's just not possible to say that a particular downward-sloping response is the response of a neutral speaker in a room.

Hence my disagreement with your point...

 

[bluefuzz]

Hence my disagreement

Obviously, there is some handwaving involved, but in my (admittedly limited) experience the preferred sound of any speaker in any room does not vary wildly from that ~10dB slope. But lets agree to disagree on how we quantify 'wildly' ... ;-)

 

[andreasmaaan]

Obviously, there is some handwaving involved, but in my (admittedly limited) experience the preferred sound of any speaker in any room does not vary wildly from that ~10dB slope. But lets agree to disagree on how we quantify 'wildly' ... ;-)

Ok, FWIW, having measured systems from studios to living rooms with varying levels of furnishing/treatment, and with various kinds of speakers, the slope you get from a neutral speaker (one that's flat on-axis, with nothing too strange about the dispersion characteristic) may vary from ~3dB (nearfield in treated studio) to ~12dB (large living room).

Happy to disagree about whether those variations are "wild" or not

 

[bluefuzz]

the slope you get from a neutral speaker (one that's flat on-axis, with nothing too strange about the dispersion characteristic) may vary from ~3dB (nearfield in treated studio) to ~12dB (large living room).

My take away from that (with suitable hand waving) is that bass needs to be as loud – and preferably a bit to a lot louder – than any other part of the frequency spectrum to sound 'right'. The vast majority of headphones sadly do not sound like that unless EQ'ed to (or designed to) approximate the Harman curve ...

 

[dasdoing]

in my (admittedly limited) experience the preferred sound of any speaker in any room does not vary wildly from that ~10dB slope.

two dirac impulses, then 2 diferent noises after about 100ms (one is pink noise, the other grey noise):

if this were 2 rooms they would sound equaly flat, because that 100ms reflection we would clearly percieve as one

yet this is how theses "rooms" meassure in a 500ms window:

if you don't have excatly the same reflection (in response and time) you can't compare two rooms in the 500ms window.


this thread sheds some light on the dificulty to determine a measurement window that actualy shows what we are hearing: https://www.hometheatershack.com/threads/feature-request-frequency-dependent-windowing.99673/

 

[andreasmaaan]

My take away from that (with suitable hand waving) is that bass needs to be as loud – and preferably a bit to a lot louder – than any other part of the frequency spectrum to sound 'right'.

I essentially agree with you here

What I disagree with is that the Harman target is the correct degree of tilt for a neutral (i.e. what you've been calling "flat") response. FYI, the most recent Harman target (2017) has a downward tilt of approximately 11dB with respect to a flat diffuse field response.

Moreover, equally importantly in terms of the discussion of perceived neutrality, the Harman "tilt" is not linear (i.e. it is not a straight downward-sloping line). Rather, it includes a little dip in the lower-midrange and a little peak in the mid-bass (relative to a straight downward-sloping line). FWIW, this dip/peak is a result of the Q of the filters used in the Harman study (if it's not clear what I mean by this, please let me know and I'll explain).

But anyway, I do agree with you that some degree of downward slope in a headphone target response is likely necessary for subjectively neutral sound. But I would disagree with the idea that the Harman target is the most neutral. In particular, given the Harman research investigated listener preference and not perceived neutrality, I'm not sure you could argue that the Harman target should even sound (on average) neutral to most people; but, as per the goals and methodology of the research, it is very likely to be (on average) among the most preferred.

 

[Feelas]

I essentially agree with you here

What I disagree with is that the Harman target is the correct degree of tilt for a neutral (i.e. what you've been calling "flat") response. FYI, the most recent Harman target (2017) has a downward tilt of approximately 11dB with respect to a flat diffuse field response.

Moreover, equally importantly in terms of the discussion of perceived neutrality, the Harman "tilt" is not linear (i.e. it is not a straight downward-sloping line). Rather, it includes a little dip in the lower-midrange and a little peak in the mid-bass (relative to a straight downward-sloping line). FWIW, this dip/peak is a result of the Q of the filters used in the Harman study (if it's not clear what I mean by this, please let me know and I'll explain).

But anyway, I do agree with you that some degree of downward slope in a headphone target response is likely necessary for subjectively neutral sound. But I would disagree with the idea that the Harman target is the most neutral. In particular, given the Harman research investigated listener preference and not perceived neutrality, I'm not sure you could argue that the Harman target should even sound (on average) neutral to most people; but, as per the goals and methodology of the research, it is very likely to be (on average) among the most preferred.

That's all good and understandable, and it would be informative if you provided what type of tilt you believe is neutral, so we can get one more opinion on this. Thanks!

 

[aldarrin]

That's all good and understandable, and it would be informative if you provided what type of tilt you believe is neutral, so we can get one more opinion on this. Thanks!

https://www.reddit.com/r/oratory1990/comments/eik71c