[money4me247]

Just wondering what would be the best way to get an accurate volume match for a direct AB comparison of two different headphones.
 
Thank you
 
edited: lol noted that there was some confusion from the "x." yeaa, really just trying to do a subjective A-B comparison but as objectively as possible. i know that volume can influence listener preference, so just trying to remove that confounding variable. thanks for everyone's feedback!

 

[Strangelove424]

You need a microphone or SPL meter, but you can probably get away with doing it on the cheap. If you needed an exact reading of the db level hitting your ear, that would require precise, calibrated equipment. But since you only need a relative comparison between two levels you could use something as basic as a smartphone w/ SPL meter app, or even better a microphone you may have handy fed to a program like audacity that can display the input level. Just make sure your microphone mimics ear position as best as possible and mounting position is consistent between tests. Avoid laying the headphones and microphone both on a surface, you need to support/suspend them both away from reflecting surfaces. The mounting of the microphone and headphones will be the creative part. Start by matching something static like noise or test tones, then proceed to music listening. 

 

[RonaldDumsfeld]

The easiest way is to install foobar with the ABX Comparator component.
 
Use the options to get the plug in to level match for you automatically.

 

[castleofargh]

it's a nightmare pure and simple. even if you solve the equipment problem, what do you volume match for? 1khz tone? white noise? depending on the headphone's signature you will have a very stupid result.
 
I gave up on direct headphone comparisons unless they have a really close signature.

 

[esldude]

I am with castleofargh here.  ABX of headphones is not going to tell you much.  For one thing switching delays dramatically reduce the ability to perceive differences and switching headphones is really going to impact that.  Measuring the level is going to be a crapshoot.  And headphone response varies enough using a single tone isn't going to work. 
 
Now if you just insist, I would suggest matching by ear using pink noise that was band limited to 200-2000 hz.  I don't think abx this way will be useful, but if you wanted to compare headphones blind it would somewhat work.  Have a tester level match them ahead of time.  Then rather than ABX them use a MUSHRA protocol.  MUSHRA isn't as sensitive as ABX, but is intended for intermediate to large differences which is a good description of headphone differences.
 
http://en.wikipedia.org/wiki/MUSHRA
 
A short paper describing MUSHRA
 
http://downloads.bbc.co.uk/rd/pubs/whp/whp-pdf-files/WHP038.pdf
 
Now it wasn't made for headphone comparisons, but seems to be more applicable as a blind testing method of headphones than ABX.
 
Also, I have not done this, it just seems like a better approach than ABX with something like headphones.
 
Essentially you are having people listen to sound and rate it as Excellent, Good, Fair, Poor or Bad.  You have a hidden reference which is a degraded signal to ensure people are reliably hearing the differences.  I have of this being adapted so that listeners rate bass, midrange and treble separately for each test signal. 

 

[castleofargh]

  I am with castleofargh here. 

/me changes status to "in a relationship".

 

[money4me247]

lol..... so basically no way to accurately compare headphones & remove volume bias?
 
*whole head-fi website made irrelevant* hahaha

 

[Strangelove424]

  what do you volume match for? 1khz tone? white noise? depending on the headphone's signature you will have a very stupid result.
 

 
I'd look at the FR graph and find a frequency range they are both equally sensitive to, or as esldude mentioned use band-limited pink noise. If you go about trying to match a 60hz tone between Denons and Grados, just as an extreme example, you will have issues. 

 

[esldude]

  lol..... so basically no way to accurately compare headphones & remove volume bias?
 
*whole head-fi website made irrelevant* hahaha

Yes, in many ways oh yes. 
 
Actually it is two things.  Frequency response is too variable between headphones.  So what frequencies do you match?  Matching almost anywhere will mean very audible FR differences somewhere.  Which is one reason FR is the main reason headphones sound different from each other.  Even if you could match FR, you have issues measuring and matching loudness due to how it sits on individual ears. 
 
So doesn't mean nothing can be told listening, but when people start crowing about tiny issues from headphone cabling well yes irrelevant. 
 
Which is also why MUSHRA-like methods are enough for headphone comparisons.  Now if one headphone has nothing below 200 hz vs another that works nicely to 20 hz you can tell that with pretty sloppy methodology even without fine level matching.  If one is a couple db hot in that range you may need more care in matching to discern it reliably.    The reason I picked 200hz- 2khz pink noise is that region is usually the flattest in most headphones.  That region also encompasses fundamentals of voice and most instruments.  So the overtone differences will stand out if the fundamentals are close to matched. 
 
Listening impressions are fine if not taken too far.

 

[esldude]

  /me changes status to "in a relationship".

Oh my is that a field we need added to head-fi profiles????? 

 

[Claritas]

Even if you could match FR, you have issues measuring and matching loudness due to how it sits on individual ears. 

So doesn't mean nothing can be told listening, but when people start crowing about tiny issues from headphone cabling well yes irrelevant.

If the headphone isn't very position-sensitive, the problems with matching might not effect tests on DACs, amps, and cables using the same pair of headphones.

 

[esldude]

If the headphone isn't very position-sensitive, the problems with matching might not effect tests on DACs, amps, and cables using the same pair of headphones.

 
Yes, but that is answering an easier question leaving unanswered the OP's initial question.  He wanted to ABX two headphones. 

 

[Dark_wizzie]

  I am with castleofargh here.  ABX of headphones is not going to tell you much.  For one thing switching delays dramatically reduce the ability to perceive differences and switching headphones is really going to impact that.  Measuring the level is going to be a crapshoot.  And headphone response varies enough using a single tone isn't going to work. 

For me if I switch headphones with 5 second delay and cannot hear a difference, then the difference is too small to ever matter. Five seconds? People say they recall how headphones sound like well enough to compare against them from a month ago! I don't even remember the lyrics of my favorite tracks that well.

 

[money4me247]

Okay, so I have two headphones and I would like to compare their audio performance in the most objective and unbiased way possible.
 
They blatantly feel different on the head, so technically I can't really call my comparison a true ABx test as I can tell X just by feel. Impossible to blind test these headphones.
 
I am just trying to be as objective about this process as possible. Audio memory is definitely unreliable, so having both pairs of headphones at the moment, I would like to do a direct comparison.
 
These headphones have different impedance and efficiency, so they playback at different volumes when software volumes are held constant. I know that volume is an important variable to control. It is funny because many comparison reviews around here never try to control for volume, so I was curious what an easy way to do that would be. Seems like all possibilities for volume-matching are quite convoluted and may not be reliable without specialized equipment or preparation, so it is probably impossible for a hobbyist to really do such an experiment...
 
It is extremely amusing to me that a hobbyist forum dedicated to headphone performance does not have some sort of universal testing system established with volume-matching for comparing the performance differences between headphones.
 
My conclusion is that probably all impressions/comparisons of headphones made ever around here are probably significantly influenced by lack of volume-matching & probably is not reliable... The fact that most comparisons are not done directly means that our unreliable audio memory also plays a big part in our subjective impressions...
 
Thank you @Strangelove424, your suggestion of using a SPL meter app is the the suggestion that sounds most feasible and I will attempt this experiment over the weekend. Thank everyone else for the general information about which frequency response to match.
 
I guess I can always just write a review based on how I feel with lots of flowery audiophile terms heh

 

[castleofargh]

 

  I am with castleofargh here.  ABX of headphones is not going to tell you much.  For one thing switching delays dramatically reduce the ability to perceive differences and switching headphones is really going to impact that.  Measuring the level is going to be a crapshoot.  And headphone response varies enough using a single tone isn't going to work. 

For me if I switch headphones with 5 second delay and cannot hear a difference, then the difference is too small to ever matter. Five seconds? People say they recall how headphones sound like well enough to compare against them from a month ago! I don't even remember the lyrics of my favorite tracks that well.

because those are different things. instant memory has the highest fidelity, and tests have long proved that for audio, instant switch is by far the most accurate way to discriminate 2 sounds. ABx is based on this and the protocol itself asks for near instantaneous switches.
you can remember sound just fine for about 3 seconds, but even in that timeframe, you will have less accuracy than with instantaneous switches.
and after than you can remember key points for a long time, usually linked to something you already knew(more bass than X headphone, less sibilant than Y headphone...). and depending on your training and number of key points, you can indeed get pretty good audio accuracy over longer periods of time. but that's just not ABX and just not as accurate.
 
 

  Okay, so I have two headphones and I would like to compare their audio performance in the most objective and unbiased way possible.
 
They blatantly feel different on the head, so technically I can't really call my comparison a true ABx test as I can tell X just by feel. Impossible to blind test these headphones.
 
I am just trying to be as objective about this process as possible. Audio memory is definitely unreliable, so having both pairs of headphones at the moment, I would like to do a direct comparison.
 
These headphones have different impedance and efficiency, so they playback at different volumes when software volumes are held constant. I know that volume is an important variable to control. It is funny because many comparison reviews around here never try to control for volume, so I was curious what an easy way to do that would be. Seems like all possibilities for volume-matching are quite convoluted and may not be reliable without specialized equipment or preparation, so it is probably impossible for a hobbyist to really do such an experiment...
 
It is extremely amusing to me that a hobbyist forum dedicated to headphone performance does not have some sort of universal testing system established with volume-matching for comparing the performance differences between headphones.
 
My conclusion is that probably all impressions/comparisons of headphones made ever around here are probably significantly influenced by lack of volume-matching & probably is not reliable... The fact that most comparisons are not done directly means that our unreliable audio memory also plays a big part in our subjective impressions...
 
Thank you @Strangelove424, your suggestion of using a SPL meter app is the the suggestion that sounds most feasible and I will attempt this experiment over the weekend. Thank everyone else for the general information about which frequency response to match.
 
I guess I can always just write a review based on how I feel with lots of flowery audiophile terms heh

 there is nothing wrong with subjectivity as long as it's not mislabelled

.