I know everyone says that headphones sound better than speakers and are more accurate, and I feel like I agree with stuff under $100. But, I just don't see that with mid level headphones or slightly higher end stuff? Btw, just putting it out there, I haven't listened to any truly high end headphones. Maybe my ears are broken or something, idk, but for example, I have a pair of definitive technology computer speakers that I got for $199, and they sound better to me than my friend's Audio Technica m50x's. I can hear more detail in my speakers, and it sounds cleaner I guess? Now I know sound quality is subjective, but my speakers just sound flatter than her m50x's except for the bass on these speakers which is slightly boosted. I looked up the frequency response of the m50s here: Headphones - 1.4 - Graph - RTINGS.com
I'm not sure how accurate those measurements are, but I'm not surprised, that's what it sounded like to me. Those measurements don't look flat compared to my Definitives which looked a lot more flat to me when I measured them. And I'm sure there are flatter bookshelf speakers than my definitives given the price point. So what gives? Why do people say headphones are much more accurate than equivalent speakers? Like I said, maybe there's something wrong with me XD but can someone clear up my confusion?
I'm not sure how accurate those measurements are, but I'm not surprised, that's what it sounded like to me. Those measurements don't look flat compared to my Definitives which looked a lot more flat to me when I measured them. And I'm sure there are flatter bookshelf speakers than my definitives given the price point. So what gives? Why do people say headphones are much more accurate than equivalent speakers? Like I said, maybe there's something wrong with me XD but can someone clear up my confusion?
Headphones are not supposed to be flat at your ears. They aren't like speakers measured in free field. The closed chamber between headphone and ear is a very different space.
That said, they should still be smooth. I measure my headphones with in ear mics and base my EQ on that. I try for a farily flat response up to 1kHz and descending after that. Works for me.
That said, they should still be smooth. I measure my headphones with in ear mics and base my EQ on that. I try for a farily flat response up to 1kHz and descending after that. Works for me.
Headphones need a slight boost in the bass to compensate for the fact that they don't give you the physical 'punch' that you get from a room speaker.
Headphones also need to be rolled-off at high frequencies to compensate for their proximity to the ear.
Dummy ear measurements of headphones usually show lots of ups & downs in the high frequency range. This is a consequence of the reflection of sound waves by the convoluted structure of the pinna, which results in peaks and troughs due to constructive and destructive interference.
Headphones also need to be rolled-off at high frequencies to compensate for their proximity to the ear.
Dummy ear measurements of headphones usually show lots of ups & downs in the high frequency range. This is a consequence of the reflection of sound waves by the convoluted structure of the pinna, which results in peaks and troughs due to constructive and destructive interference.
That response isn't the actual output of the headphone. The graph has been "compensated" to remove the headphone's response characteristics designed to mimic your ear/head's effect on free field sound. The exact nature of the correct target response for headphones is still a subject of debate. Google HRTF or Sean Olive's work for details.
Is there a reason similarly priced speakers sound better to me than headphones though? Is it that I haven't listened to any decent headphones? Or are my ears broken? 🙂
The graphs are used as comparisons to others, in conjunction with having experience listening to the others.
Member
Joined 2009
Paid Member
I still remember my first hi end headphones, a pair of Grado 60e imported from the US (they sold for touch-your-toes prices in Canada at the time) and they were so not flat that I was very disappointed. Most of the time I would hope for improvement with ‘burn in’ but it never came and each time I had to rip them off my head before blood came pouring out of my ears. I’ve never been brave enough to buy headphones since although I have some HD650’s in my Amazon wish list I just haven’t got over the horror of those Grado’s.
fyi - bose noise cancelling headphones used during flights sound fine but lose out to the airyness of speakers.
fyi - bose noise cancelling headphones used during flights sound fine but lose out to the airyness of speakers.
Last edited:
At your ears or in your ears? It depends where you measure. Yes, there are reasons. You can find much better detailed explanations than I can provide if you just do a little searching on the Interwebs.
This should get you started:
Headphone Measurements Explained | InnerFidelity
That might be a bit strong but it's certainly the wild west. Agreement appears to end with the belief that headphones with a perfectly flat response won't sound like flat speakers. The exact headphone response necessary to sound like a flat speaker appears to vary widely between testers. I believe RTings uses their own response weighting subject to occasional revision. To add even more ambiguity a strong case can be made that extra-aural and deep insertion in-ear headphones each require different responses to sound 'natural'.
My uneducated guess is that an over-the-ear headphone, measured through an in-ear microphone (tiny microphone in the ear canal) and a "flat" loudspeaker measured through the same in-ear microphone, should sound the same if the headphone is equalized for the same frequency response as the loudspeaker.
Having recently done that, I can confirm, yes - mostly. At least that is how I base my headphone EQ, plus a downward slope like the classic B&K room curve. I do find it works well.
There is debate about whether the ear canal should be blocked or not.
There is debate about whether the ear canal should be blocked or not.
I have had a Symth Realizer (A8) demo...
the Smyth Realizer shows that your individual HRTF needs to be measured, then with the headphone' FR you can get amazing emulation of external sound fields - but both your head/ears measured personal response and the headphone's measured FR on your head, at your ear canal opening are needed
CanJam SoCal 2018 - Smyth Research Realiser A16 Headphone Surround System | InnerFidelity
always check any Smyth Realizer review for having a proper personalized calibration - the generic setting or other people's HRTF personalization files won't be colse enough to fool your brain
the Smyth Realizer shows that your individual HRTF needs to be measured, then with the headphone' FR you can get amazing emulation of external sound fields - but both your head/ears measured personal response and the headphone's measured FR on your head, at your ear canal opening are needed
CanJam SoCal 2018 - Smyth Research Realiser A16 Headphone Surround System | InnerFidelity
always check any Smyth Realizer review for having a proper personalized calibration - the generic setting or other people's HRTF personalization files won't be colse enough to fool your brain
So I have been told. One person's setting will sound very weird to someone else.
I have used in-ear measurements to gently EQ my headphones with very good results. My target is the classic downward slope of 1.8 dB per octave above 1kHz. Just did a new EQ this morning and am happy with it.
I have used in-ear measurements to gently EQ my headphones with very good results. My target is the classic downward slope of 1.8 dB per octave above 1kHz. Just did a new EQ this morning and am happy with it.
Speaking as a chap who suffers from hyperacusis, where listening to any speakers/headphones that do stupid things with the critical 1-3kHz region, either though frequency response or directivity, is a complete nono. It's actually physically painful when they get things wrong. The HD650s are simply the easiest thing to listen to that I've ever experienced. They are the exact opposite of ears bleeding.
Pano,
How do you know, what is flat? (I mean what is the reference) Or did you eq to your "known flat" loudspeaker room response?
Unfortunately I don't know. So I take a leap of faith and EQ to what I believe is correct. Flat from 20Hz to 1000Hz, then descending 1.8dB/octave.
The present EQ for my Fostex is to remove a large wide hump at 140 Hz, add a little at 500Hz, then a high shelf filter at 3kHz to boost the top end. There is also a big peak at 9K that I'm not sure belongs there or not. Usually with EQ I don't change each filter as much as measurement would indicate, but often use about half that amount. For example instead of a 5.5dB reduction at 140, I used 3.5dB. Too much EQ usually sounds wrong.
What I end up with is a much more natural sound, a tonal balance that I would expect to hear from good speakers or from live, acoustic music.
The present EQ for my Fostex is to remove a large wide hump at 140 Hz, add a little at 500Hz, then a high shelf filter at 3kHz to boost the top end. There is also a big peak at 9K that I'm not sure belongs there or not. Usually with EQ I don't change each filter as much as measurement would indicate, but often use about half that amount. For example instead of a 5.5dB reduction at 140, I used 3.5dB. Too much EQ usually sounds wrong.
What I end up with is a much more natural sound, a tonal balance that I would expect to hear from good speakers or from live, acoustic music.
If you EQ flat relative to your HRTF you should end up with a flat response at the opening to your ear canal, same as you should have when listening to speakers. Very oversimplified but there's a lot of good discussion here:
Measuring HRTF for headphone use | Audio Science Review (ASR) Forum
Last edited:
Member
Joined 2009
Paid Member