Even though most headphone amplifiers have a fairly flat frequency response across the entire audio band, most headphones don't it seems. Such as these curves for the AKG K550:
The Wonderfully Competent AKG K550 Sealed Headphone Page 2 | InnerFidelity (scroll down)
I'm curious if anyone knows of any headphone equalization hardware and/or software that can equalize the frequency response curves of specific headphone models. By this I mean being able to select a pre-determined set of equalization values from a table to correct a given headphone model's specific frequency response glitches, not just fiddling with the sliders on a multi-band parametric equalizer.
I'm aware of two so far. This hardware "headphone correction filter PCB" by "solderdude"
Headphone | DIY-Audio-Heaven
which is essentially a high and low shelving filter followed by a 2 band parametric equalizer. The "low" shelving filter is the same circuit I use in my desktop headphone amp for bass boost. He has a table of part values to equalize the curves on 35 different models of headphones.
Then there is the headphone equalization software from Audyssey built into their Media Player app for the iphone:
Improve Your Headphones | Audyssey Media Player
Audyssey apparently has well over 100 headphone models in their equalization database. The software is a great solution if you use an iphone, but I don't and Audyssey says they have no plans for an Android or PC version.
Any others out there?
The Wonderfully Competent AKG K550 Sealed Headphone Page 2 | InnerFidelity (scroll down)
I'm curious if anyone knows of any headphone equalization hardware and/or software that can equalize the frequency response curves of specific headphone models. By this I mean being able to select a pre-determined set of equalization values from a table to correct a given headphone model's specific frequency response glitches, not just fiddling with the sliders on a multi-band parametric equalizer.
I'm aware of two so far. This hardware "headphone correction filter PCB" by "solderdude"
Headphone | DIY-Audio-Heaven
which is essentially a high and low shelving filter followed by a 2 band parametric equalizer. The "low" shelving filter is the same circuit I use in my desktop headphone amp for bass boost. He has a table of part values to equalize the curves on 35 different models of headphones.
Then there is the headphone equalization software from Audyssey built into their Media Player app for the iphone:
Improve Your Headphones | Audyssey Media Player
Audyssey apparently has well over 100 headphone models in their equalization database. The software is a great solution if you use an iphone, but I don't and Audyssey says they have no plans for an Android or PC version.
Any others out there?
Last edited:
I have a Superlux HD681 headphone.
I have made a similar filter to the one with the 470 mH and 3.3 mH coils in DIY-Audio-Heaven.
I used LTspice and modelled the output impedance of my HiFimeDIY UA23 DAC and input impedance of my phones to filter out the peak in the upper frequencies. I used the graphs from the Headphone Learning website.
I used polyester capacitors and wound my own air core coils. I used variable resistors for tweaking, which I still need to do. Cannot find the time in the near future. I mounted the parts on a piece of wood, intending a prettier box later.
My coil and capacitor sizes were slightly different. I wound my own coils, so not limited to manufactured values.
This system is a significant improvement over unmodified Sennheiser HD580's, in that the bass is more natural.
The results are fantastic. I cannot stop listening to music now. True audiophile quality for about $100 USA, including the DAC, filter, and headphones. It works with any computer.
I have made a similar filter to the one with the 470 mH and 3.3 mH coils in DIY-Audio-Heaven.
I used LTspice and modelled the output impedance of my HiFimeDIY UA23 DAC and input impedance of my phones to filter out the peak in the upper frequencies. I used the graphs from the Headphone Learning website.
I used polyester capacitors and wound my own air core coils. I used variable resistors for tweaking, which I still need to do. Cannot find the time in the near future. I mounted the parts on a piece of wood, intending a prettier box later.
My coil and capacitor sizes were slightly different. I wound my own coils, so not limited to manufactured values.
This system is a significant improvement over unmodified Sennheiser HD580's, in that the bass is more natural.
The results are fantastic. I cannot stop listening to music now. True audiophile quality for about $100 USA, including the DAC, filter, and headphones. It works with any computer.
How-To: Auto-EQ is published -
Funny you should ask --- I published an article in Linear Audio magazine last year that does just that --> auto-EQ for headphones using the Audysses system but any type of auto-EQ will work, of course. It is part of my Hi-End headphone amplifier article. Its all there on how to do it. Not a generic EQ curve but an EQ for your head/ear/phones using a probe microphone which you can easily/cheaply make.
Its all right there for you.
Thx-Richard Marsh
Funny you should ask --- I published an article in Linear Audio magazine last year that does just that --> auto-EQ for headphones using the Audysses system but any type of auto-EQ will work, of course. It is part of my Hi-End headphone amplifier article. Its all there on how to do it. Not a generic EQ curve but an EQ for your head/ear/phones using a probe microphone which you can easily/cheaply make.
Its all right there for you.
Thx-Richard Marsh
Thanks for the feedback! I have actually been thinking about getting that HiFime DIY async DAC. I just noticed last week that he had those available now. I have his $30 ODAC clone. Good to know that the eq circuit for the 681 does the job. Yeah I've been kind of guessing that some accurate equalization would go a long way to closing the gap between midrange phones and higher end cans. Sounds like it does!
Richard - thanks for the heads-up on the article! Looks like that is volume 3 from last year. I'll get it on order from Jan. I've been wanting to read an issue of LA. Good point that ears are part of the system and should ideally be included in the equalization too.
Last edited:
As soon as I hear that the async DAC is better than the original UA23, I will order one.
If there is a USB DAC that gives a better bang for the buck, I would like to know.
If there is a USB DAC that gives a better bang for the buck, I would like to know.
Auto EQ:
Would equalizing for a flat response direct on the ear canal not be the wrong thing ?
If you take into account the influence of the head shape and the outer ear, would not
a diffuse-field equalization be more appropriate ?
Auditory environment
Would equalizing for a flat response direct on the ear canal not be the wrong thing ?
If you take into account the influence of the head shape and the outer ear, would not
a diffuse-field equalization be more appropriate ?
Auditory environment
In my limited experience, a narrow peak in a frequency response in more annoying than a wide peak.
A narrow dip in a response is less annoying than a narrow peak of the same magnitude.
The overall flatness of the response in not as important and distortions.
If distortions are produced in an effort to flatten a frequency response, I would choose the less flat response. Is digital frequency modification distortion free? Clarity of sound in paramount.
A narrow dip in a response is less annoying than a narrow peak of the same magnitude.
The overall flatness of the response in not as important and distortions.
If distortions are produced in an effort to flatten a frequency response, I would choose the less flat response. Is digital frequency modification distortion free? Clarity of sound in paramount.
MLS water-fall plot of active LP filter -
Yes to all of the above.
Note that a flat frequency response with filters (crossovers) can be low distortion as well BUT it still is not percieved as Flat. This is due to the Q of the filters adding energy to the response at the cross-over freq. The total energy at the filters corner has its audible affects.
Thx-RNMarsh
🙂
Yes to all of the above.
Note that a flat frequency response with filters (crossovers) can be low distortion as well BUT it still is not percieved as Flat. This is due to the Q of the filters adding energy to the response at the cross-over freq. The total energy at the filters corner has its audible affects.
Thx-RNMarsh
Sorry for missing some typo errors in the previous posting.
Thank you Richard for your interesting point about filters.
I had a most interesting visit with Joseph Manger in Germany at his home some years ago. I was impressed with his Manger speakers, in that they gave the orchestra some sense of depth in addition to the expected stereo spread. Also there was a huge sweet spot, completely different from electrostatic "head in a vice" experience.
The main concern he expressed about speakers is the importance of phase coherence. His theory was that the human ear was developed to hear the sound of the tiger creeping up behind your back, and the cracking of a stick would locate its position, based on phase difference of the sound reaching your ears. Listening fatigue is caused by crossovers in multi way systems, because the ear is genetically programmed to sort out location by phase, which is messed up by multi way speakers.
This has implications for the arguments used in feedback discussions, which sometimes leave me quite confused.
The Superlux headphones, have a very simple peak near the high end. It can be tamed by a simple filter, well about the most important listening frequencies, and hopefully does not ruin the phase problems through the main auditory frequencies. This makes it an outstanding headphone for its low price. The filter, in my opinion though is mandatory, not an option.
Thank you Richard for your interesting point about filters.
I had a most interesting visit with Joseph Manger in Germany at his home some years ago. I was impressed with his Manger speakers, in that they gave the orchestra some sense of depth in addition to the expected stereo spread. Also there was a huge sweet spot, completely different from electrostatic "head in a vice" experience.
The main concern he expressed about speakers is the importance of phase coherence. His theory was that the human ear was developed to hear the sound of the tiger creeping up behind your back, and the cracking of a stick would locate its position, based on phase difference of the sound reaching your ears. Listening fatigue is caused by crossovers in multi way systems, because the ear is genetically programmed to sort out location by phase, which is messed up by multi way speakers.
This has implications for the arguments used in feedback discussions, which sometimes leave me quite confused.
The Superlux headphones, have a very simple peak near the high end. It can be tamed by a simple filter, well about the most important listening frequencies, and hopefully does not ruin the phase problems through the main auditory frequencies. This makes it an outstanding headphone for its low price. The filter, in my opinion though is mandatory, not an option.
Last edited:
I just visited the headphone learning site again:
Learning Center - Build a Headphone Graph | HeadRoom Audio
The graph for the HD681 headphone is a little different for the one I used 9 months ago. I took the peak to be about 7.5 kHz. This graph now shows 8.5 kHz.
Maybe the best thing to do is a measure the high frequency peak with a audio oscillator of the phones you are actually using.
Learning Center - Build a Headphone Graph | HeadRoom Audio
The graph for the HD681 headphone is a little different for the one I used 9 months ago. I took the peak to be about 7.5 kHz. This graph now shows 8.5 kHz.
Maybe the best thing to do is a measure the high frequency peak with a audio oscillator of the phones you are actually using.
Here is a free LMS software package to add the missing dimension to the use of filters --
www.audiostandart.com/software/55-freelms.html
Thx-RNMarsh
www.audiostandart.com/software/55-freelms.html
Thx-RNMarsh
Here is the Google translation to English from the Cyrillic for that: 🙂
Google Translate
Richard - I just ordered that Volume 3 of Linear Audio with your headphone eq article. Should be some good reading!
Google Translate
Richard - I just ordered that Volume 3 of Linear Audio with your headphone eq article. Should be some good reading!
I am curious to know if you try the EQ the way I did, if you find it was an improvment for you. It was for me and so added it to the headphone article.
-Richard
-Richard
Richard, thank you so much for your suggestion about the MLS software. I would love to make the comparisons.
I am using Linux Ubuntu. I was not able to get it into my computer last week. The MLS seems to be for MS Windows. I could try to run it under Wine, but do not have the time to fool around with it.
I am in Canada visiting family for a couple of weeks. When I get back to the UK, I will be very busy renovating a new house we recently moved into. I will not have much time for audio till I get the renovations done.
I am using Linux Ubuntu. I was not able to get it into my computer last week. The MLS seems to be for MS Windows. I could try to run it under Wine, but do not have the time to fool around with it.
I am in Canada visiting family for a couple of weeks. When I get back to the UK, I will be very busy renovating a new house we recently moved into. I will not have much time for audio till I get the renovations done.
Last edited:
Another option is a hardware solution http://www.smyth-research.com/products.html
Probably the best option on this market. Relatively expesive though.
Probably the best option on this market. Relatively expesive though.
Start out flat freq and then adjust from that base -
Only if you wanted to recreate the sound As If you were listening to speakers in a room with your two ears. But with headphones, you are not doing that, of course. Recordings made just for headphone playback or optimized for headphone listening are done more and more often now and that requires a flat response. So, the answer is -- it all depends. But in most cases reducing unwanted peaks and valley's in the response still ends up as better sounding on average - over all. Then, if you want, adjust/compensate for affect as you feel is best or more accurate.
Thx-RNMarsh
Only if you wanted to recreate the sound As If you were listening to speakers in a room with your two ears. But with headphones, you are not doing that, of course. Recordings made just for headphone playback or optimized for headphone listening are done more and more often now and that requires a flat response. So, the answer is -- it all depends. But in most cases reducing unwanted peaks and valley's in the response still ends up as better sounding on average - over all. Then, if you want, adjust/compensate for affect as you feel is best or more accurate.
Thx-RNMarsh
Last edited:
I read a little about that -- thx for the web location... Seems interesting to me and would hope Ver 2.0 and beyond get developed... eventually in a smaler package .. maybe even portable or into an IC. takes time but should be encouraged, IMO.
Thx-RNMarsh
- Status
- Not open for further replies.