Is EQ'ing headphones worth it?

[JohnYang1997]

While I generally agree, even measurements from someone like oratory1990 on Reddit have some extreme features that are not worth correcting to. For example his EQ profile for the DT1990 has a significant correction for a dip around 4 or 5KHz. While I think there's definitely something going on at that frequency, if I apply the full correction for that dip as measured on his industry standard HATS, it sounds just plain wrong.

One also has to keep in mind unit to unit variation. Audeze is notorious for this, but it's generally true that a measurement of the same model headphones as yours doesn't necessarily match a measurement of your exact headphones.

IMO the measurement results on his HATS have some issues. It's not the measurements that I'm familiar with. One example is er4's measurements are no where near the one etymotic provides(also what we measure).

Using the EARS from Minidsp is like much worse than headphones' product variation. The acoustic impedance and the physical dimension is completely fucked up. There's a standard to it, IEC-60318. 60318-4 is for the coupler, 60318-7 is for the head IIRC. ITU p57 and 58 also specify some of the stuff.

 

[BillG]

But anyway, what do you think about that EQ'ing site I listed above and what do you think about EQ'ing headphones....and does the site I listed above strive to EQ out the anomalies to try to get as close to neutral "as the artist intended" as possible?

Equalizing any transducer is fine, despite some purist's objections - it's your equipment, do what you want with it. However, some transducers are more flexible in regards as to how much EQ they can handle before distorting badly. Someone else can explain the physics behind that, if they like. I just know that it can be an issue.

I'd not concern myself with the notion of "as the artist intended"; you wont have the same configuration of equipment as they had in the vast majority of cases, nor the same listening environment or perceptual biases.

There are now very easy ways to accomplish satisfying equalization; I just happened to be previewing the following on the my smartphone a few hours ago with some pleasing results - it uses a combination of machine learning and user preferences to generate its EQ curves. Interestingly enough, it generated a response curve very similar to Harman's based upon my own preferences...

https://www.sonarworks.com/truefi/early-access

 

[JeffS7444]

Well. This is not remotely accurate. It will do more harm than good definitely.

A lot of us are using measurement tools which are less than reference-grade (think $60 measurement mics) to achieve meaningful improvements in sound quality. I see that someone's got a used Bruel & Kjaer 4128-D-001 listed on eBay at 27,500 USD: That's beyond the scope of most home hobbyists, wouldn't you say? The MiniDSP product fills this product niche just fine.

 

[A800]

No.
EQ (being filters) will introduce group delay and phase issues.

 

[JohnYang1997]

A lot of us are using measurement tools which are less than reference-grade (think $60 measurement mics) to achieve meaningful improvements in sound quality. I see that someone's got a used Bruel & Kjaer 4128-D-001 listed on eBay at 27,500 USD: That's beyond the scope of most home hobbyists, wouldn't you say? The MiniDSP product fills this product niche just fine.

That's why I suggest using someone else's work. It will be much better than EARS. You may do some calibration to your own target to make EARS somewhat useful. There are communities that are contributing to this. But if you have watched Oluv's Gadget youtube channel he was working extremely hard for each pair of headphones to make them sound like what he heard. There are just too many things wrong with that thing.
For room correction, you can get a omni mic very cheap and get reasonable results. But headphones is different, in ears would be even more difficult. I even suggest using flat plate than EARS for the sake of consistency and high frequency accuracy. Tho planars and dynamic headphones still behave very differently.

 

[JohnYang1997]

No.
EQ (being filters) will introduce group delay and therefore phase issues.

Do you know that headphones are also filters? And every minimum phase filter has a inverse filter theoretically? Eq will correct the frequency response issue and phase response at the same time if the headphone itself doesn't have excess phase. And yes clean square wave response can be reproduced via eq.

 

[JeffS7444]

Do you think most "Reference Headphones" that are sold on the basis of being 'neutral' are targeted at the Harman curve you showed me earlier?

Not necessarily. Sean Olive's research found little correlation between price and sound quality.
https://seanolive.blogspot.com/2017/02/twirt-337-predicting-headphone-sound_17.html

 

[A800]

Do you know that headphones are also filters? And every minimum phase filter has a inverse filter theoretically? Eq will correct the frequency response issue and phase response at the same time if the headphone itself doesn't have excess phase. And yes clean square wave response can be reproduced via eq.

The ear canal and the ear drum itself are also filters.

 

[JohnYang1997]

The ear canal and the ear drum itself are also filters.

Only matters is DRP response.

 

[A800]

No idea.

 

[solderdude]

The ear canal and the ear drum itself are also filters.

Those are filters that are removed by the brain as our brain calibrates itself continuously over the years by listening to real sounds all day long.
On top of that those filters differ from person to person which again does not matter at all because when we listen to real instruments our brain calibrates the 'incoming' info as real.
So while the FR, phase etc differ individually at the ear drum we all hear real sounds as real.
On top of that people get used to a certain presentation. When such is not 'too wrong' our brain accepts this.
Such becomes painfully obvious when comparing headphones directly.
It is the main reason for 'brain-in' as well but people not realizing this think the headphone needs to 'burn-in'
Also pads softening up and giving better seal or changing ear-driver distance is an important factor in sound 'changing'.
When the latter happens slowly it isn't that obvious until one compares directly.

As been said here... there is EQ and there is EQ, there is a target curve, there is taste.
No matter how one twists and turns this ALL headphones have substantial FR dips and peaks. More so than speakers. It makes perfectly sense to compensate for this.
Do it incorrectly and you may change it but not improve it (EARS) or are compensating for measurement rig anomalies that are not caused by the headphone and while improving certain points you may also be making aspects worse as they are measurement rig related and not or poorly compensated.

Yes, EQ can be difficult to get right but done 'properly' always brings about an improvement.
Headphones with specific resonances, very sharp dips or peaks can't be improved that well.
A cheap generic headphone can never be made to sound like a high-end headphone. You could very well improve tonality though (up to a certain point)

 

[Fluffy]

Here are my two cents – trying to reach a perfect response using measurements never really worked for me. I use EQ to slightly improve my already pretty nice headphones to my taste. It's all done manually and subjectively, tweaking in small increments over time until it sounds more pleasing with it than without it. I do use measurements as a reference to understand what the headphones are actually doing and to give a clue to what changes might be beneficial, but the determining factor is my ear. Some equalizations I've made had no basis in the headphones measurements, and were just done by ear.

This is a very individualistic approach and I'm sure other people will not like the changes I've made. That's why when someone comes to visit and want to listen to my headphones, I turn off all EQ and let them listen to them as they are.

 

[pwjazz]

Such becomes painfully obvious when comparing headphones directly.
It is the main reason for 'brain-in' as well but people not realizing this think the headphone needs to 'burn-in'

I've never had this issue with speakers. I can listen to the same music on my studio monitors, then my phone speaker, then my living room system, then my car system and while I can certainly hear differences in bad response and detail, none of them sounds "wrong". Switching back and forth between a DT1990 and an LCD2C, they both end up sounding flat or wrong!

I don't like that, so lately I've been aiming for headphones/EQ that eliminate this problem.

 

[Kouioui]

Etymotic has done something similar to Harman but instead of finding what's subjectively preferred, etymotic found the more accurate speakers in room response. And that's 30 years ago.

The issue I see with Ety's diffuse field target is how you define accurate speakers in a room. Most modern studio monitors are EQ'd to a room/house curve that is closer to the Harman headphone target. If that house curve is your reference, it's only logical your headphones follow. I EQ my speakers using the target from this video.

 

[flipflop]

it uses a combination of machine learning and user preferences to generate its EQ curves. Interestingly enough, it generated a response curve very similar to Harman's based upon my own preferences...

https://www.sonarworks.com/truefi/early-access

Last time I checked, SW based their EQs on actual measurements and had zero transparency with regards to their target curve.

 

[A800]

Those are filters that are removed by the brain as our brain calibrates itself continuously over the years by listening to real sounds all day long.
On top of that those filters differ from person to person which again does not matter at all because when we listen to real instruments our brain calibrates the 'incoming' info as real.
So while the FR, phase etc differ individually at the ear drum we all hear real sounds as real.
On top of that people get used to a certain presentation. When such is not 'too wrong' our brain accepts this.
Such becomes painfully obvious when comparing headphones directly.
It is the main reason for 'brain-in' as well but people not realizing this think the headphone needs to 'burn-in'
Also pads softening up and giving better seal or changing ear-driver distance is an important factor in sound 'changing'.
When the latter happens slowly it isn't that obvious until one compares directly.

As been said here... there is EQ and there is EQ, there is a target curve, there is taste.
No matter how one twists and turns this ALL headphones have substantial FR dips and peaks. More so than speakers. It makes perfectly sense to compensate for this.
Do it incorrectly and you may change it but not improve it (EARS) or are compensating for measurement rig anomalies that are not caused by the headphone and while improving certain points you may also be making aspects worse as they are measurement rig related and not or poorly compensated.

Yes, EQ can be difficult to get right but done 'properly' always brings about an improvement.
Headphones with specific resonances, very sharp dips or peaks can't be improved that well.
A cheap generic headphone can never be made to sound like a high-end headphone. You could very well improve tonality though (up to a certain point)

While I agree basically I also have the experience that EQ in general will suck some life out of speaker drivers.
In the end I always went back to no EQ at all.

 

[majingotan]

Here are my two cents – trying to reach a perfect response using measurements never really worked for me. I use EQ to slightly improve my already pretty nice headphones to my taste. It's all done manually and subjectively, tweaking in small increments over time until it sounds more pleasing with it than without it. I do use measurements as a reference to understand what the headphones are actually doing and to give a clue to what changes might be beneficial, but the determining factor is my ear. Some equalizations I've made had no basis in the headphones measurements, and were just done by ear.

This is a very individualistic approach and I'm sure other people will not like the changes I've made. That's why when someone comes to visit and want to listen to my headphones, I turn off all EQ and let them listen to them as they are.

This is precisely why I never prefer listening to headphones and IEMs at home. None of them sound tonally close to my reference speakers, even with EQ. I find that EQ on headphones and IEMs makes some genre sound more preferable to my subjective preferences but is tonally dead on other genre while this issue NEVER happens on a properly room treated speaker response.

 

[Fluffy]

This is precisely why I never prefer listening to headphones and IEMs at home. None of them sound tonally close to my reference speakers, even with EQ. I find that EQ on headphones and IEMs makes some genre sound more preferable to my subjective preferences but is tonally dead on other genre while this issue NEVER happens on a properly room treated speaker response.

Headphones that sounds FANTASTIC with some things > Speakers that sound good on everything

That's my philosophy, anyway. Plus, it's way cheaper and more practical to get an awesome sounding headphone system than a speaker system. Currently I have absolutely no option to set up a full range speaker system in a treated room.

 

[DivineCurrent]

In my opinion, EQ may be the most important tool you can use to improve your listening experience. I'd say it is 100% worth the time and effort.

I haven't had much luck equalizing headphones and IEMs to target curves in the past. They either sound off, or the frequency graphs do not translate to what I am hearing at all.
That being said, headphones like the Sennheiser HD 58X, HD 600, Focal Elex, and AKG K371 all are tuned to be relatively even and neutral sounding, without major peaks and other inconsistencies in the frequency response. Over time looking at graphs and experimenting with EQ, my sonic signature preference has evolved and keeps evolving. I used to love the stark neutral sound of the HD600, and now I prefer the slightly warmer tilt of the 58X with the small dip in the 2-3 kHz range. It is interesting to note that listening to sine sweeps, the 2-3 kHz region in the 58X sounds perfectly flat to me, even though it measures as a dip on most rigs. There is also similar behavior on the Focal Elex, another favorite of mine.

When it comes to the Etymotic curve, I think there is some really good research there. However, in the case of the ER2SE, I personally hear way too much presence from 3-6 kHz when doing sine sweeps. That area needs to be EQ'd down by about 6 dB to sound like the HD600 and 58X response. Otherwise, it sounds far brighter and more forward than the HD 600. The new Drop JVC HA-FDX1 (with the dampening filters) tones down that 3-6 kHz region to my liking, and sounds more in line with what I consider neutral and speaker-like.

 

[JohnYang1997]

The issue I see with Ety's diffuse field target is how you define accurate speakers in a room. Most modern studio monitors are EQ'd to a room/house curve that is closer to the Harman headphone target. If that house curve is your reference, it's only logical your headphones follow. I EQ my speakers using the target from this video.

View attachment 50721

Where did you get your information from? Majority of studios don't even use EQ at all, let alone choosing such room curve that you use. There's BK room curve for hifi use and there's the small room X curve which is naturally a treated room has.