• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Other EQ target curves

Jose Hidalgo

Addicted to Fun and Learning
Joined
Aug 15, 2020
Messages
961
Likes
1,032
Location
France
Hi everybody,

For headphones, the most well-known EQ target curve may be the Harman one (and its variants for each headphone type : AE-OE or IE).
In my experience the Optimum HiFi curve is also a good alternative when you want less bass and more midrange presence.

My question would be : leaving aside Harman and Optimum HiFi, is there any other target curve that would be worth a try at least for some specific type(s) of music ? And if that's the case, what are the supposed benefits of such target curve(s) ?

Sites like Oratory and AutoEQ mention other curves, but they seem to be much less used (the Diffuse Field curve and the Free Field curve come to mind).

Thanks.
 
  • Like
Reactions: trl

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,335
Likes
6,700
Hi everybody,

For headphones, the most well-known EQ target curve may be the Harman one (and its variants for each headphone type : AE-OE or IE).
In my experience the Optimum HiFi curve is also a good alternative when you want less bass and more midrange presence.

My question would be : leaving aside Harman and Optimum HiFi, is there any other target curve that would be worth a try at least for some specific type(s) of music ? And if that's the case, what are the supposed benefits of such target curve(s) ?

Sites like Oratory and AutoEQ mention other curves, but they seem to be much less used (the Diffuse Field curve and the Free Field curve come to mind).

Thanks.

You might start with older Harman curves depending on how much less bass you want. It seems like every new curve just keeps adding more bass.
 
OP
Jose Hidalgo

Jose Hidalgo

Addicted to Fun and Learning
Joined
Aug 15, 2020
Messages
961
Likes
1,032
Location
France
Thanks Richard, but like I said, "leaving aside Harman and Optimum HiFi, is there any other target curve"...
I know already about the Various Harman curves, older and new, AE-OE and IE. This is about other curves. ;)
 

maverickronin

Major Contributor
Forum Donor
Joined
Jul 19, 2018
Messages
2,527
Likes
3,308
Location
Midwest, USA
I prefer something closer to diffuse field. That used to be the default target curve before Harman/Olive & Welti/whatever.

For me it pushes the image more forward than the Harman target, with better potential for out of head localization with good headphones.

I always use crossfeed too, which is definitely and influencing factor.
 

Bullwinkle J Moose

Active Member
Joined
Mar 31, 2021
Messages
217
Likes
90
A Pink Noise Target Curve is the bestest ever!

Open a pink noise file in iZotope RX

Select EQ Match / Learn

Now open a Horribly Recorded song with the worst EQ ever (OMG - Try Neil Young)

Now match the track to the Flat EQ Curve

Well recorded tracks are very close to flat and this will have little or no effect and sound best on Studio Monitors with a flat response

Some bad tracks may actually become listenable

This does not work on the vast majority of Original Rolling Stones tracks however as they had a different target curve for each mic / or instrument

A flat Pink Noise curve as a target for them sounds horrible

Try these >
Bob Marley - Rat Race (100% Flat)
Bob Marley - So Much Trouble in the World (72% Flat)
Neil Young - Old Man (100% Flat)


Here is one Rolling Stones Track that it DOES actually work on >
Under My Thumb (100% X 2 = Double Flat)


If Flat is not your thing, then yeah, you can try Joe's curve, Some Other Guys Curve, The Harman Curve or the BangPowZoom Curve

That last one might be Chinese - LoL
 
Last edited:

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
You might start with older Harman curves depending on how much less bass you want. It seems like every new curve just keeps adding more bass.
I've noticed this myself too, so now I wonder which curve will better reflect the real life reality, let's say for an average SPL of 85 dB?
 

Bullwinkle J Moose

Active Member
Joined
Mar 31, 2021
Messages
217
Likes
90
I've noticed this myself too, so now I wonder which curve will better reflect the real life reality, let's say for an average SPL of 85 dB?

FLAT!

If the vast majority of songs were each recorded with a different EQ curve, then Harman is the wrong curve on the vast majority of tracks

The artist should always use the best curve (whatever sounds best to them) for playback on a flat system

If you want to hear exactly what an intelligent artist intended, then flatten your playback

For the Clueless Artist, then Meh....Whatever Curve you like

Just play with them knobs all you want, but remember, the curve you prefer for one song, will not work on the next
 
Last edited:

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
Now you need to define FLAT, given human hearing curve. :)
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
Other curves.
correction-plots-2.png

Note the Harman curve is an older one which has been revised a few times.
I have my own (personal) curve which is very close to the Golden Ears target.
Below some of the curves I played with by using somewhat different listening SPL's
Ended up with the purple one.
correction1.png

Measured 'flat' but with somewhat elevated lows, just not as 'steep' and as high as Harman.
It's a few dB off between 1.5kHz and 6kHz because of the lack of a pinna in my measurements.

My preference is not 'speakers in a room at listening position a few meters away' but closer to good nearfield monitors listened to in nearfield.
This means treble is not gently sloping and bass is not really elevated.
Since I generally listen at around 70 Phon average I need some small compensation for bass by lack of SPL so I 'correct' by applying bass correction as dictated by equal loudness contours (so not a sharp bend but sloping bend).
f-m-overlayed.png


I find Harman bass 'impressive' but unnatural, like many others perhaps.
I do my own research instead of simply relying on that of others that use 'average' like-ability of sound.
Funnly enough the sound I aim for is very close to Harman but a bit 'clearer' sounding and more 'natural' instead of 'impressive' to me.
 

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
I have an '83 Electromures/Unitra amplituner (receiver) with a dedicated carbon/ribbon inside the volume potentiometer for adjusting properly the loudness in real-time. so the volume potentiometer is still 2-gangs, but with 4 pins/channel instead of 3 pins, one pin being for the inverted-logarithmic ribbon inside which is used for the loudness adjust for each channels

Waiting for the first manufacturer to revive such a loudness that changes the curve based on volume pot, not just a simple switch and a knob that can be manually adjusted.
 
Last edited:

sigbergaudio

Major Contributor
Audio Company
Forum Donor
Joined
Aug 21, 2020
Messages
2,639
Likes
5,396
Location
Norway
FLAT!

If the vast majority of songs were each recorded with a different EQ curve, then Harman is the wrong curve on the vast majority of tracks

The artist should always use the best curve (whatever sounds best to them) for playback on a flat system

If you want to hear exactly what an intelligent artist intended, then flatten your playback

For the Clueless Artist, then Meh....Whatever Curve you like

Just play with them knobs all you want, but remember, the curve you prefer for one song, will not work on the next

Depending on what you mean, this is not accurate. What typically works well is a speaker that measures flat outside or in an anechoic chamber. Such a speaker will not measure flat in a typical listening room. And as such, your headphones shouldn't have a flat curve either, it will sound too thin for most listeners.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
Waiting for the first manufacturer to revive such a loudness that changes the curve based on volume pot, not just a simple switch and a knob that can be manually adjusted.

This only works properly when gain is fixed and the transducers have a certain efficiency.
With voltage efficiency ranging 20dB for headphones, gains in amps selectable and digital volume control this won't work properly in all situations.

There were also potmeters with 2 tabs or separate 'loudness' pots alongside the regular volume control.
RME has a digital version of this b.t.w.

I'm confused. Is the topic about headphone curves or speaker/room curves? OP seemed to be asking about headphone curves, but I *think* most of the replies are about speaker/room curves.

In theory these are the same things. One tries to get the same 'tonal balance' at a specific listening level.
The only difference for headphones is the lack of tactile feel at larger SPL and subbass extension in headphones can be much better.

There are no room reflections and different speaker/listener effects either.

To me headphones and speakers are too different but in both cases you may want the perceived tonal balance to be in the same ballpark.
 
  • Like
Reactions: trl

Bullwinkle J Moose

Active Member
Joined
Mar 31, 2021
Messages
217
Likes
90
Now you need to define FLAT, given human hearing curve. :)
FLAT, as in Pink Noise Flat

Tracks are normally EQ'd by ear (with their flawed human hearing curve) to sound best on playback equipment with a Flat Response

To add another flawed curve on playback might sound good on some tracks, but those tracks were wrong to begin with

If you mix your tracks to sound best on a flat playback system, there is no need to fix them because they are not broken to begin with
 

Bullwinkle J Moose

Active Member
Joined
Mar 31, 2021
Messages
217
Likes
90
I'm confused. Is the topic about headphone curves or speaker/room curves? OP seemed to be asking about headphone curves, but I *think* most of the replies are about speaker/room curves.


Tracks are normally mixed to sound their best on studio monitors with a flat response

If your headphones truly had a flat response, those tracks would sound fine

if you "prefer" more Bass, then crank up the Bass

You do not need a specific pre-defined curve for your specific preference, or your specific ears

A harmon curve might work with "some" ears, but only with "some" music as most songs are EQ'd differently, based on the band or engineers preference

Using iZotope RX, you can "Learn" the specific EQ curve of every song in your possession and instantly see how different each one is to the next

Like I said, a pre-defined curve might only be appropriate for a few of them based on your specific equipment and ears

With a flat response as a starting point, we can see how close the artist came to a perfect response (or how bad their equipment, knowledge and ears are)

and would you even know whether a bands preferred mixing curve already matches the harmon curve or not without using iZotope RX or something similar with a learning function ???

If you were to add a harmon curve to another harmon curve without knowing that you did so, yet like the result, then the end result is just a personal preference and nothing more

You may as well just twist some knobs until you like it
 
Last edited:
OP
Jose Hidalgo

Jose Hidalgo

Addicted to Fun and Learning
Joined
Aug 15, 2020
Messages
961
Likes
1,032
Location
France
I'm confused. Is the topic about headphone curves or speaker/room curves? OP seemed to be asking about headphone curves, but I *think* most of the replies are about speaker/room curves.
You are right, I'm talking about headphones here, as stated in the OP. But I appreciate everyone's responses and try to learn from them. :)

Target curves are specific to the measurement gear. Are you making your own measurements?
I'm just talking about the usual headphone target curves as published by people like Oratory or Jaakko. For example the latest versions of the Harman curve would be (AFAIK) AE-OE 2018 and IE 2019v2. I was just wondering if I was missing on something besides Harman and Optimum HiFi. Some people here seem to root for Diffuse Field, solderdude has his own custom curve, etc..

Also, I think once we have a good Harman preset for a given headphone, it's fairly easy to derive other target curves from it : "Harman + additional filters = new target". For example we can derive an Optimum HiFi preset from any Harman preset just by a couple of additional filters, as said by Oratory himself. It could be the same for Diffuse Field, etc.. BTW I'm choosing Harman as a starting point, simply because 95% of available headphone presets are Harman ones (e.g. AutoEQ).
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
Also, I think once we have a good Harman preset for a given headphone, it's fairly easy to derive other target curves from it

One would think that but it's not the case.
You see the Harman curve does not use the same pinna as oratory/Amir/Crinacle so there is a deviation.
The correction curve is very smooth yet the measurement is anything but smooth.
Also what one measures is not exact. You can see this by overlaying measurements of the worlds most and easy to measure headphones (HD650).
Above 5kHz the wiggles are not at the same frequency and depth.
Suppose we use the HD650 as reference and take 5 different test rigs and create compensations so these 5 HATS are very closely the same in 'compensated result'.
Now take another headphone, one with larger drivers or angled and measure that on the same 5 rigs... voila 5 differing results.

Suppose we take 10 headphones and make 10 correction curves for all 5 rigs and then for all 5 rigs we create an 'average' correction so the all headphones are just slightly 'off' on average we get a nice correction curve.
That curve won't look like the Harman curve.

Have a look at a bunch of measurements from Amir, Oratory and Crinacle. Do you see the dip around 9-10kHz. If the compensation curve would be a really good fit the 'target curve' would also have this dip.
Ever noticed that almost all headphones tested after the Harman correction have huge (sometimes +10dB) peaking around 20kHz on average ?
This is because the rig cannot measure correctly (and why in plots you see 10kHz + greyed out) yet the upwards tilt is shown in the plots yet no headphones do that in reality. It's because the average is 'incorrect' and the Harman target thus is wrong.

I have no idea why this has not been 'corrected' yet.

So even when you have measurements made acc. to a standard that doesn't mean they are accurate. The raw plots are actual measurement voltages but the target curve is average.

This is why I agree with Amir. Headphone measurements are NOT an exact science and should not be treated as gospel.
They can be a good starting point from where one can 'eyeball' problem areas and then apply some EQ that roughly corrects what's terribly wrong.

Using a trace from some headphone on some rig and considering the measurement as golden standard (regardless who measured it and on what) and simply calculating the opposite so it measures 'perfect' on that rig alone and claiming only then it sounds 'perfect' is closing one's eye to reality.

It is also why there is a huge discrepancy between EQ's from Jaakko and Oratory and Amir.
 
OP
Jose Hidalgo

Jose Hidalgo

Addicted to Fun and Learning
Joined
Aug 15, 2020
Messages
961
Likes
1,032
Location
France
Actually I agree with most of what you said, so there was no need for that, lol. :)

When I wrote what I wrote, I was mostly thinking about the Optimum HiFi curve. That curve can be easily derived from Harman, with just a couple of parametric filters that essentially lower bass < 200 Hz from about -5.5 dB (plus it's more of a user-preference, really, so there's no need to be precise). From 200 Hz up, both curves are virtually identical. It's literally what Oratory said, so I won't be debating this:
Oratory1990 said:
You can always reduce the gain of the 105 Hz low-shelf filter by 5.5 dB to go from Harman OE to the Optimum HiFi Target.
In any case, the gain of the low-shelf filter at 105 Hz should always be set according to personal preference. There's no single "correct" answer here.

Probably other curves can't be derived so easily, because they are not "variants of the same curve" and they also happen to involve higher frequencies. I won't question that.

With that said, maybe an imperfect derivation is better than no derivation at all. I don't know. Most listeners aren't experts like you, solderdude. And of course 99% of them don't have the means or the will of getting an expensive professional measurement rig and performing their own measurements. Most listeners just want "good enough" ready-to-use EQ presets, that they'll be able to adjust themselves by ear if required. Which is exactly why EQ presets are so popular. And which is also why Oratory provides his newer pdf presets with included instructions on how to modify the filters.

Yes, I know about higher frequencies and rig measurement errors that should definitely NOT be corrected when EQing, but thanks for pointing it out for other people. ;)

One last thing : I didn't create this topic to debate endlessly about such things, but merely to answer the OP question (about other target curves), which thankfully has already received nice answers from some people. Thank you.
 
Last edited:
Top Bottom