Ive looked at Stax, hd800, hd 650, akf k701, denon, ECT. the behavior in the FR of most headphones seems quite consistent and definitely is far from being flat.
Doesnt matter if its flagship headphones or cheap headphones like earpods. The headphones Frequency Response are quite stunning
why and how do they still sound good?
I mean, they are so far from flat that it ould be totally unnacceptable if speakers would measure that way!
Ive compared most flagship headphones and they all measure TERRIBLE!!
overall, headphones seems consistent in their terrible unflatness:
In general, and thats a rough approximation:
from 60hz to about 7-800 hz, they are relatively flat so lets say its at 0= flat
but after that, its a mess...
from 1khz it stars to show rolls off up to 2khz normally showing a good - 3 to - 8 db from flat and that with most headphones
from 2khz to 5khz, we are often -10 db, with stax showing -15db from 2 khz to 5khz????
5khz to 10khz normally comes back with a approx -5db.
10khz will depend but its never flat, its at least again -5db from flat and sometimes -15 even -20 db from flat!!!
I have to be honest, I find that most headphones sound quite pleasing.
Following what I know about the importance of a relatively flat speaker, please help me understand this weird phenomenon.
Should we EQ speakers, not in that spectacular manner, but with that as reference?
Would that be more pleasing to EQ our speakers to mimic a bit the heapdhones FR?
something like:
bass and midrange up to 1khz flat
1khz to 2khz -3db
2khz to 5khz -5db
5khz to 10khz -3db
10khz and up -5db
how can we still enjoy headphones when they are so not flat and ultimately not transparent at all?!
please let me understand a bit more this very contreversial subject.
Google Image Result for http://cdn.head-fi.org/a/a7/a72366c7_HD800AnaxMod.gif
Google Image Result for http://cdn.innerfidelity.com/images/AKG_Q701_graph_q701k701frcompare.jpg
Google Image Result for http://i374.photobucket.com/albums/oo181/acharpen/CSD/CSD_InnerFidelityData_110924R1_StaxSR007_vs_SR009_FRF.jpg
Google Image Result for http://cdn.innerfidelity.com/images/Denon_AHDx000_graph_NewOldRight.jpg
Google Image Result for http://en.goldenears.net/en/files/attach/images/254/149/015/c9a56b352f59374decb2822c6267e4e0.png
Google Image Result for http://i837.photobucket.com/albums/zz296/solderdude/allM50graphs.jpg
Google Image Result for http://cdn.innerfidelity.com/images/Audeze_LCD2LCD3_graph_lcd3OldAndNew.jpg
Google Image Result for http://graphs.headphone.com/graphCompare.php%3FgraphType%3D0%26graphID%255B%255D%3D853%26graphID%255B%255D%3D573%26graphID%255B%255D%3D863%26graphID%255B%255D%3D2881
Doesnt matter if its flagship headphones or cheap headphones like earpods. The headphones Frequency Response are quite stunning
why and how do they still sound good?
I mean, they are so far from flat that it ould be totally unnacceptable if speakers would measure that way!
Ive compared most flagship headphones and they all measure TERRIBLE!!
overall, headphones seems consistent in their terrible unflatness:
In general, and thats a rough approximation:
from 60hz to about 7-800 hz, they are relatively flat so lets say its at 0= flat
but after that, its a mess...
from 1khz it stars to show rolls off up to 2khz normally showing a good - 3 to - 8 db from flat and that with most headphones
from 2khz to 5khz, we are often -10 db, with stax showing -15db from 2 khz to 5khz????
5khz to 10khz normally comes back with a approx -5db.
10khz will depend but its never flat, its at least again -5db from flat and sometimes -15 even -20 db from flat!!!
I have to be honest, I find that most headphones sound quite pleasing.
Following what I know about the importance of a relatively flat speaker, please help me understand this weird phenomenon.
Should we EQ speakers, not in that spectacular manner, but with that as reference?
Would that be more pleasing to EQ our speakers to mimic a bit the heapdhones FR?
something like:
bass and midrange up to 1khz flat
1khz to 2khz -3db
2khz to 5khz -5db
5khz to 10khz -3db
10khz and up -5db
how can we still enjoy headphones when they are so not flat and ultimately not transparent at all?!
please let me understand a bit more this very contreversial subject.
Google Image Result for http://cdn.head-fi.org/a/a7/a72366c7_HD800AnaxMod.gif
Google Image Result for http://cdn.innerfidelity.com/images/AKG_Q701_graph_q701k701frcompare.jpg
Google Image Result for http://i374.photobucket.com/albums/oo181/acharpen/CSD/CSD_InnerFidelityData_110924R1_StaxSR007_vs_SR009_FRF.jpg
Google Image Result for http://cdn.innerfidelity.com/images/Denon_AHDx000_graph_NewOldRight.jpg
Google Image Result for http://en.goldenears.net/en/files/attach/images/254/149/015/c9a56b352f59374decb2822c6267e4e0.png
Google Image Result for http://i837.photobucket.com/albums/zz296/solderdude/allM50graphs.jpg
Google Image Result for http://cdn.innerfidelity.com/images/Audeze_LCD2LCD3_graph_lcd3OldAndNew.jpg
Google Image Result for http://graphs.headphone.com/graphCompare.php%3FgraphType%3D0%26graphID%255B%255D%3D853%26graphID%255B%255D%3D573%26graphID%255B%255D%3D863%26graphID%255B%255D%3D2881
Last edited:
Think Ive got my answer!
''
First, ALL measurements are of questionable value. Most often there is non, at the very best very limited correlation between "measures better" and "sounds better" for any given measurement.
Second, measuring headphones at high frequencies is non-trivial and much of what you see in HF ripples/non-flatness are a combination of attempting to provide a certain degree of diffuse field EQ and of the reflections within the cavity formed by the headphones ear-cup, the head and the ear-pads.
Third, no two sets of headphone measurements are comparable with each other or any given speaker result. While there is a standardised test "head simulator" it is very expensive and so not commonly used. Further, some measured responses will include a calibration for a presumed "standard diffuse field response" with everyone who uses it using their version and everyone having their own idea what the correct parameters are.''
''
First, ALL measurements are of questionable value. Most often there is non, at the very best very limited correlation between "measures better" and "sounds better" for any given measurement.
Second, measuring headphones at high frequencies is non-trivial and much of what you see in HF ripples/non-flatness are a combination of attempting to provide a certain degree of diffuse field EQ and of the reflections within the cavity formed by the headphones ear-cup, the head and the ear-pads.
Third, no two sets of headphone measurements are comparable with each other or any given speaker result. While there is a standardised test "head simulator" it is very expensive and so not commonly used. Further, some measured responses will include a calibration for a presumed "standard diffuse field response" with everyone who uses it using their version and everyone having their own idea what the correct parameters are.''
You need an instrumented head/ear canal model with mics mounted where the eardrums are. The ear canal makes a difference. That is why it still sounds good.
For IEM's you can use piece of tubing of same diameter as ear canal and make it same depth.
Linkwitz talks about measuring IEM's and a EQ circuit to reduce the 7kHz peak generated by the eardrum bounce and length of the ear canal.
For IEM's you can use piece of tubing of same diameter as ear canal and make it same depth.
Linkwitz talks about measuring IEM's and a EQ circuit to reduce the 7kHz peak generated by the eardrum bounce and length of the ear canal.
Sorry, but I'm going to disagree here.
I sat down with some components for a small PA system - 12" two-way cabinets. I measured each component individually. The woofers had a 2.5kHz spike that needed notching out, even with a 4th order crossover. The compression driver/horn combination had a lot of output around 5kHz, with a couple of high-Q peaks in the passband, and tailed off towards 20kHz.
The resultant frequency response was much flatter, and the sound many many times better than before those bits were EQ'd flat. Even at low levels, the 5kHz output on the HF made me wince when some harmonic (particularly on voices) got to that region.
It's simple enough to try for yourself, too. Fire up some software that has EQ available, and make the frequency response measurement look awful. Then listen.
I'm not saying that frequency response is the be-all and end-all, just that I've always found an improvement with the sound when the response measures flat, or close. I suspect others have found the same.
Chris
The headphone speaker is close to the ear in the near field not in the far field like a loudspeaker or a musical instrument. The diffraction of sound around the upper body, head and ear which our brain expects to be present is missing. The distortion of the headphone frequency response is an attempt to compensate for this as well as one or two other things.
also, ive been reading a bit more in headphone forums.
I know hi frequency travel faster then lower frequenciess.
I know that with distance, hi frequency lose velocity.
which could explain also the need to have a not flat response from headphones.
I begin to agree with flat measuring speakers. I use one18 amphion right now, they are the best speaker I have ever had or heard and they are, by far, the one that measure the best in my room.
I have measured recently the wilson audio cub recently, it was awful in terms of measurments
I know hi frequency travel faster then lower frequenciess.
I know that with distance, hi frequency lose velocity.
which could explain also the need to have a not flat response from headphones.
I begin to agree with flat measuring speakers. I use one18 amphion right now, they are the best speaker I have ever had or heard and they are, by far, the one that measure the best in my room.
I have measured recently the wilson audio cub recently, it was awful in terms of measurments
This is how your body - torso, head, ear external and internal - modifies sound before you hear it. Headphones, depending on whether they're over ear or deep insert, can bypass some or most of these modifications. The theory is that building the mods into the headphones results in a more natural listening experience.
Attachments
yup, headphones are great.
Time alignment, no acoustics.
Too bad headphones localize to inside your head.
And I flop on the couch while doing other things, I need jukebox sound instead of headphones or monitors.
Norman
Time alignment, no acoustics.
Too bad headphones localize to inside your head.
And I flop on the couch while doing other things, I need jukebox sound instead of headphones or monitors.
Norman
My 2c:
-Non linear distortion is often amazingly low in headphones compared to speakers of a similar budget. Non-flat frequency response is often far less objectionable when there are no accompanying non-linear distortion problems occurring at the same frequencies.
-Frequency response is highly dependant on how the headphones are positioned on the ear and people will subconsciously adjust the positioning to achieve the frequency response that they like the most
-Measuring frequency response of headphones is problematic due to the wide variation in heads shape/size, ear shape/size and personal preference for positioning of the headphones.
Measuring headphones still has some merit. Particularly when comparing the effects of modifications or the general tone of difference headphones - e.g. Headphone X has more bass and less midrange than Headphone Y. Then if you have listened to headphones X or Y you may have some idea of what to expect before forking out the $ to buy the other one without first hearing it.
-Non linear distortion is often amazingly low in headphones compared to speakers of a similar budget. Non-flat frequency response is often far less objectionable when there are no accompanying non-linear distortion problems occurring at the same frequencies.
-Frequency response is highly dependant on how the headphones are positioned on the ear and people will subconsciously adjust the positioning to achieve the frequency response that they like the most
-Measuring frequency response of headphones is problematic due to the wide variation in heads shape/size, ear shape/size and personal preference for positioning of the headphones.
Measuring headphones still has some merit. Particularly when comparing the effects of modifications or the general tone of difference headphones - e.g. Headphone X has more bass and less midrange than Headphone Y. Then if you have listened to headphones X or Y you may have some idea of what to expect before forking out the $ to buy the other one without first hearing it.
Last edited:
- Status
- Not open for further replies.