• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

"Bias" of some members towards headphone measurements?

I stuck a mic near a driver of my HD650.

Got this:

1737041293811.png


So, I suppose come kind of rig is needed to more accurately measure what it outputs vs what you hear.
 
Which method do you mean ?
All standard measurement rigs produce different results and are not accurate above several kHz. They are pretty accurate between 100Hz and 6kHz which arguably is the most important part of the music spectrum.
This is clearly verifiable and requires a sizable investment.

I was responding to @Jaxx1138 who said "current method", which I assumed meant judging broad compliance with Harman OE 2018 target. I agree with your points, they are some of the reasons I said "albeit not perfect".
 
What do you mean "the current method"?
All current HATS or just head and pinna systems are not consistent with one another(they arrive at slightly different results) ie; there is no real standard. Just deviations on a theme.
I feel you're striving for a perfection which doesn't exist
Isn't that the goal of audiophiles, audio engineers, and designers in general? In some sense anyway?

I think it's a good argument and goes back to my point about perfection. If you are correct then it should be possible to make a pair of headphones which, by applying a generic impulse response for speakers in a room (generic in the sense it's not for a specific HRTF), sound the same for everyone (the same in the sense that it sounds like sitting in that room). The fact Smyth exists, Creative had a go with S-Xfi, to name just 2, demonstrates it's not this simple because of the way headphone fundamentally interact with each person's anatomy differently.
I agree with your general conclusion here. The only part I feel is missing the point. Is that Smyth is emulating a defined sound system in a defined space. It is not designing a playback system but using an existing one and then overlaying a series of filters for each ear. This holographic effect is basically a really advanced binaural recording. It's not all that mysterious. There is a lot more utilized in their processing but there is no reason to go down that rabbit hole. But let's be clear, the Smyth system has nothing to do with understanding good measurements for headphones. Conflating HRTF's with objective analysis of some independent playback system is erroneous. Just because loudspeakers in a room is a different way of listening to music does not mean that the way we analyze objective metrics of headphones needs to be anti-scientific and confusing. Reading a headphone metric should be as easy as reading a Klippel NFS set of data for a loudspeaker. Any reasoning that it shouldn't be becomes some religious cult belief that deviates from the physics of acoustics.
 
I stuck a mic near a driver of my HD650.

Got this:

View attachment 421562

So, I suppose come kind of rig is needed to more accurately measure what it outputs vs what you hear.
It's not wrong per se. But it does not represent what is heard by the listener in the low frequencies. The acoustic loading on the driver is incorrect as compared to when the headphone is on your head, even with an open back design.
 
I am not certain about what the is the very best way to analyze headphones acoustically. I only understand (after using many of the current HATS systems) that the results we obtain currently are unreliable and not very repeatable (hence why so many rigs tell you to run an average which in turn smooths the curve). The old argument about subjectivity has been used for a long time by audiophiles. Research shows us that this is untrue and that among large segments of the population prefer flat response with smooth directivity to a statical significance in double blind studies. But here we go again with headphones, and now we are back too "its all subjective". While we ignore looking into accurately quantifying the actual acoustic performance of a given headphone. HRTF's are not subjective but are unique to each individual. Audio reproduction is science whether you like it or not. Steely Dan is subjective (whether you like it or not).
 
But here we go again with headphones, and now we are back too "its all subjective". While we ignore looking into accurately quantifying the actual acoustic performance of a given headphone. HRTF's are not subjective but are unique to each individual.

I don't think anyone is saying "it's all subjective" when it comes to headphones or giving up on a scientific approach.

Despite the broad and well-established preference for flat response / smooth directivity, there is room for individual preference in both speakers and headphones, but let's leave that aside.

The fact that everyone has a unique HRTF means that the same headphone actually does sound different to each listener, unlike speakers. A flat speaker sounds flat to everyone, but everyone hears a slightly different headphone.

As I understand it, this drives the variety and uncertainty in headphone measurement. They try to come up with measurement apparatus that will better approximate what average listeners hear / prefer.

However, the industry converging on a single measurement standard or target curve wouldn't solve the fundamental challenge of individuals having different HRTFs.

For that you either need a wide variety of well-characterized headphones (and consumers who have somehow figured out how their HRTF interacts with them) or for everyone to get HRTFs captured to compensate their headphones with. I feel like the latter is actually easier.
 
I don't think anyone is saying "it's all subjective" when it comes to headphones or giving up on a scientific approach.

Despite the broad and well-established preference for flat response / smooth directivity, there is room for individual preference in both speakers and headphones, but let's leave that aside.

The fact that everyone has a unique HRTF means that the same headphone actually does sound different to each listener, unlike speakers. A flat speaker sounds flat to everyone, but everyone hears a slightly different headphone.

As I understand it, this drives the variety and uncertainty in headphone measurement. They try to come up with measurement apparatus that will better approximate what average listeners hear / prefer.

However, the industry converging on a single measurement standard or target curve wouldn't solve the fundamental challenge of individuals having different HRTFs.

For that you either need a wide variety of well-characterized headphones (and consumers who have somehow figured out how their HRTF interacts with them) or for everyone to get HRTFs captured to compensate their headphones with. I feel like the latter is actually easier.
Again, the reference does not change for anyone. Miles Davis in a concert hall will play the same and have the same acoustic output for you and for me as long as we are both in the relatively same location. The speaker is not re-tuned for each listener. Although your pinna and my pinna are different and our hearing may have different responses, when listening to a KEF Blade loudspeaker, that reference does not change. Headphones are NO DIFFERENT. I fail to understand why no one understands this. The subjectivity of my hearing vs your hearing does not change the mix and mastering of the source in any way. The reference stays the same regardless of the subject. HRTF's have NOTHING to do with making a good headphone. So long as the correct target function is met, then it's all the same. A speaker with free field flat response with good directivity does not need to be modified in order to be accurate for people with different hearing. Accurate is always accurate. There is no scientific research that indicates otherwise. Everyone listens to the same Led Zeppelin recordings. If you own a B&W 800D and someone else owns one then B&W does not change the X-over for each listener. Room modes are a separate thing. You cannot conflate HRTF's with any electronic sound reproduction system. HRTF's are just human transfer functions. They have nothing to do with IEM's, headphones or loudspeakers. the physics of acoustics do not change with how humans hear.
 
I don't think anyone is saying "it's all subjective" when it comes to headphones or giving up on a scientific approach.

Despite the broad and well-established preference for flat response / smooth directivity, there is room for individual preference in both speakers and headphones, but let's leave that aside.

The fact that everyone has a unique HRTF means that the same headphone actually does sound different to each listener, unlike speakers. A flat speaker sounds flat to everyone, but everyone hears a slightly different headphone.

As I understand it, this drives the variety and uncertainty in headphone measurement. They try to come up with measurement apparatus that will better approximate what average listeners hear / prefer.

However, the industry converging on a single measurement standard or target curve wouldn't solve the fundamental challenge of individuals having different HRTFs.

For that you either need a wide variety of well-characterized headphones (and consumers who have somehow figured out how their HRTF interacts with them) or for everyone to get HRTFs captured to compensate their headphones with. I feel like the latter is actually easier.
You don't compensate for HRTF's. There is no need. The target function for a headphone is different than a free field response target for a loudspeaker. That is due to, proximity of the source, angle of source to the ear, room gain(Schroeder Frequency) , the current metric is not free field (for headphones), and head positioning is not reliable (in many cases). Once that transfer function is identified (scientifically) , which Harman has started the process with it's subjective preference tests, then it is a simple process to identify what is accurate.
 
Again, the reference does not change for anyone. Miles Davis in a concert hall will play the same and have the same acoustic output for you and for me as long as we are both in the relatively same location. The speaker is not re-tuned for each listener. Although your pinna and my pinna are different and our hearing may have different responses, when listening to a KEF Blade loudspeaker, that reference does not change. Headphones are NO DIFFERENT. I fail to understand why no one understands this. The subjectivity of my hearing vs your hearing does not change the mix and mastering of the source in any way. The reference stays the same regardless of the subject. HRTF's have NOTHING to do with making a good headphone. So long as the correct target function is met, then it's all the same. A speaker with free field flat response with good directivity does not need to be modified in order to be accurate for people with different hearing. Accurate is always accurate. There is no scientific research that indicates otherwise. Everyone listens to the same Led Zeppelin recordings. If you own a B&W 800D and someone else owns one then B&W does not change the X-over for each listener. Room modes are a separate thing. You cannot conflate HRTF's with any electronic sound reproduction system. HRTF's are just human transfer functions. They have nothing to do with IEM's, headphones or loudspeakers. the physics of acoustics do not change with how humans hear.
Yep the music does not change and we all hear live recordings as live and real (no matter what coloration the room throws at it).

It becomes funny when you are in the audience and record the sound where one stands and listens to it.
Now... do that at 2 different locations and compare those recordings... they will sound very different and very different from how the live music was perceived.

Our brain calibrates itself constantly for sounds around us and as long as one isn't autistic filters out a lot of irrelevant crap. This is particularly true when we can use the eyes (and body) as input as well.

It becomes a different thing when listening to headphones. Simply because our brain is missing cues (no crossfeed and artificial hall/echo) as the sounds come from the sides.
Our ears are not really designed for this.
Also sounds coming from 'in front of us' (speakers/live sound) is reaches us in a certain way and sight helps (unless one is visually impaired then things change).
We don't see sound sources from the side. Brain only 'knows' it doesn't come from in front of us.
Yet this is what we do with headphones. Sure we can use crossfeed or other trickery and modify the FR of the recording to suit our HRTF but still the coupling between 2 small speakers a few cm away in a small (closed) 'chamber' differs from actual point sources far away and in front of us.

HRTF.png


Add to that the rather 'wonky' FR response (but usually good phase response) differs from what speakers on a distance do (other type of FR modifications, echo's, hall, phase and nulls). Not to mention seal and positioning on the head, lack of tactile feel which not everyone will 'hear' the same way.

And agreed, with speakers you don't compensate for HRTF, the brain does that for you and is remarkably good at that.
Speakers and headphones, however, with sounds coming from very different directions have different HRTF so it makes sense to compensate the difference between hearing sounds from the front and side of us if the goal is to get 'realistic' sound.
Recordings are means to be heard from 2 speakers in front of us ... unless they have been recorded binaural.

It is well known that ears and ear canals differ individually (and differ from fixtures as well but is another matter). For sounds coming from in front of us this does not really matter as the brain takes care of it.
But sounds coming from close proximity from the sides where for instance the pinna is also deformed (when touching the baffle, bypassed partly or completely) the ears are not 'calibrated' by the brain and hear it differently for coupling reasons. Especially above a few kHz the differences can be dramatically different from sounds coming from the front of us.
 
Yep the music does not change and we all hear live recordings as live and real (no matter what coloration the room throws at it).

It becomes funny when you are in the audience and record the sound where one stands and listens to it.
Now... do that at 2 different locations and compare those recordings... they will sound very different and very different from how the live music was perceived.

Our brain calibrates itself constantly for sounds around us and as long as one isn't autistic filters out a lot of irrelevant crap. This is particularly true when we can use the eyes (and body) as input as well.

It becomes a different thing when listening to headphones. Simply because our brain is missing cues (no crossfeed and artificial hall/echo) as the sounds come from the sides.
Our ears are not really designed for this.
Also sounds coming from 'in front of us' (speakers/live sound) is reaches us in a certain way and sight helps (unless one is visually impaired then things change).
We don't see sound sources from the side. Brain only 'knows' it doesn't come from in front of us.
Yet this is what we do with headphones. Sure we can use crossfeed or other trickery and modify the FR of the recording to suit our HRTF but still the coupling between 2 small speakers a few cm away in a small (closed) 'chamber' differs from actual point sources far away and in front of us.

View attachment 421727

Add to that the rather 'wonky' FR response (but usually good phase response) differs from what speakers on a distance do (other type of FR modifications, echo's, hall, phase and nulls). Not to mention seal and positioning on the head, lack of tactile feel which not everyone will 'hear' the same way.

And agreed, with speakers you don't compensate for HRTF, the brain does that for you and is remarkably good at that.
Speakers and headphones, however, with sounds coming from very different directions have different HRTF so it makes sense to compensate the difference between hearing sounds from the front and side of us if the goal is to get 'realistic' sound.
Recordings are means to be heard from 2 speakers in front of us ... unless they have been recorded binaural.

It is well known that ears and ear canals differ individually (and differ from fixtures as well but is another matter). For sounds coming from in front of us this does not really matter as the brain takes care of it.
But sounds coming from close proximity from the sides where for instance the pinna is also deformed (when touching the baffle, bypassed partly or completely) the ears are not 'calibrated' by the brain and hear it differently for coupling reasons. Especially above a few kHz the differences can be dramatically different from sounds coming from the front of us.
ut sounds coming from close proximity from the sides where for instance the pinna is also deformed (when touching the baffle, bypassed partly or completely) the ears are not 'calibrated' by the brain and hear it differently for coupling reasons. Especially above a few kHz the differences can be dramatically different from sounds coming from the front of us.
The confusion is because we are listening to the sound power of loudspeakers in a room. So the reference changes from what a loudspeaker is doing(Frequency Response). But we currently measure these so differently that they have little to do with one another. Having said that the timbre or target function can be simulated as it already is to some degree in headphones with the built in pinna gain as seen above. The part that gets confusing is the total convolution due to the fact that the room is both diffuse and non-reflective according to the reflective and absorptive coefficients of the room combined with the loudspeakers directivity or overall sound power. Then we have the spatialized part which is the nature of stereo listening . Otherwise known as the HRTF's. There is a timbre function associated with it that adds to the overall sound power or perceived timbre. The typical shelf function around 2K from the right ear listening to the left loudspeaker and the shading from the head (part of HRTF's) But that can be added back into the target function(overall timbre). The only thing I do not address is the cross correlation between the two ears. This can all be measured and accounted for. None of it will change what is essentially an accurate headphone is, regardless of humans all having slightly different HRTF coefficients. IEM's admittedly will always have issues because they bypass our pinna filters, and since that is a known variable then without taking DRP metric on each person it will be close but not entirely accurate for everyone. The assumption being made by the industry is that since HRTF's are all different, then the target functions are all different. This is not correct. Perception of each individuals experience will not and does not alter the reference. Somehow there seems to be a consensus that our perception determines the target function of a headphone. No one in science thinks that color accuracy on a televisions is derived by opinions. They use objective measurements. You can surely alter the picture if you like, but no one in the industry would use that altered version as a reference. In the science of the reproduction of sound, fidelity is not a moving target. If you are not seeking accuracy that's fine, but engineers seek to quantify the physics of sound in order to more reliably and repeatedly get closer too true fidelity of the signal. The HRTF's as a function of timbre are very relevant , but in a general or average of the population way. Not tuning to each persons ears. You want the frequency balance that the loudspeakers create. If you can add the cross correlation , even better, but you would have to create standards for that as well. Someday soon...
 
Yep the music does not change and we all hear live recordings as live and real (no matter what coloration the room throws at it).

It becomes funny when you are in the audience and record the sound where one stands and listens to it.
Now... do that at 2 different locations and compare those recordings... they will sound very different and very different from how the live music was perceived.

Our brain calibrates itself constantly for sounds around us and as long as one isn't autistic filters out a lot of irrelevant crap. This is particularly true when we can use the eyes (and body) as input as well.

It becomes a different thing when listening to headphones. Simply because our brain is missing cues (no crossfeed and artificial hall/echo) as the sounds come from the sides.
Our ears are not really designed for this.
Also sounds coming from 'in front of us' (speakers/live sound) is reaches us in a certain way and sight helps (unless one is visually impaired then things change).
We don't see sound sources from the side. Brain only 'knows' it doesn't come from in front of us.
Yet this is what we do with headphones. Sure we can use crossfeed or other trickery and modify the FR of the recording to suit our HRTF but still the coupling between 2 small speakers a few cm away in a small (closed) 'chamber' differs from actual point sources far away and in front of us.

View attachment 421727

Add to that the rather 'wonky' FR response (but usually good phase response) differs from what speakers on a distance do (other type of FR modifications, echo's, hall, phase and nulls). Not to mention seal and positioning on the head, lack of tactile feel which not everyone will 'hear' the same way.

And agreed, with speakers you don't compensate for HRTF, the brain does that for you and is remarkably good at that.
Speakers and headphones, however, with sounds coming from very different directions have different HRTF so it makes sense to compensate the difference between hearing sounds from the front and side of us if the goal is to get 'realistic' sound.
Recordings are means to be heard from 2 speakers in front of us ... unless they have been recorded binaural.

It is well known that ears and ear canals differ individually (and differ from fixtures as well but is another matter). For sounds coming from in front of us this does not really matter as the brain takes care of it.
But sounds coming from close proximity from the sides where for instance the pinna is also deformed (when touching the baffle, bypassed partly or completely) the ears are not 'calibrated' by the brain and hear it differently for coupling reasons. Especially above a few kHz the differences can be dramatically different from sounds coming from the front of us.
Interesting. So this drawing is the FR perception our brain has to deal with. Speakers headphones or more or less both?
 
@Jaxx1138
yes, w/ headphones the proximity of the transducer results in reflections that are not present when the sound source is positioned further away. that applies to 'in ears' too as the timpan is a reflective surface.

also, the 'Harman preference curve' involved both trained and untrained listeners. while on many aspects subjects agreed, discrepancies indicate the subjective/preference gap is unbridgeable in some areas for numerous reasons, one being, our aging hearing needs different things for max intelligibility as time passes...

this is Dr. Sean Olive, Senior Research Fellow for Harman International in his article titled "The Perception and Measurement of Headphone Sound Quality - What Do Listeners Prefer?", a summary of the Harman headphone research (19+) papers found on Audio Engineering Society:

(page 5): "There was already prior evidence that younger males and less experienced listeners preferred higher levels of bass and treble in their headphones compared with females, experienced, and older listeners (Olive e tal., 2013a; Olive and Welti, 2015)."

(page 9): "The Harman target curve is one example that is preferred by a majority (64%) of listeners from a broad range of ages, listening experiences, and genders. Slight adjustments in the bass and treble levels may be necessary to compensate for variance in the quality of recordings and to satisfy individual tastes. The Less Bass Is Better class (21% of listeners) includes a disproportionate percentage of females and older listeners and none of the trained listeners. The More Bass Is Better class is skewed toward males versus females by a factor of 4 to 1. There is no evidence that sound quality preferences are geographically influenced. Recognition of good sound reproduction seems to be universal. Objective measurements of the headphones using standard ear simulators can predict how good they sound. The further the frequency response a headphone deviates from the Harman target response, the lower its perceived sound quality will be. A simple linear model based on these deviations can predict how listeners would rate it in controlled listening tests."
 
Somehow there seems to be a consensus that our perception determines the target function of a headphone. No one in science thinks that color accuracy on a televisions is derived by opinions.
If the television were somehow connected directly to your optic nerve, bypassing the eye and retina, I could easily imagine needing color correction for it to look right. This is like the situation with headphones and IEMs.

You don't compensate for HRTF's. There is no need.
I'll ask again. If this is the case, why do spatial audio processors need to compensate for HRTFs to work properly? It's not only the phase / timing that matters here.
 
No, no and no! As you make something for humans the human ability in that regard is crucial other than ability of the device it self or if you want their limitations. This doesn't mean you can't make preference better by studying all relevant factors influencing it. You won't buy bad seats for the movie or concert because you know your experience will be less than optimal there. You need to understand how our brain sees such intentions as hostile and overwhelming and tries to filter it out and focus interpret able speech area. It goes so far that we have difference between genders which area is more in the line of focus. In other words your brain will fight back against you putting it under unnecessary overwhelming processing state that's exhausting for it. It's even worse when it's put under in synthetic not natural conditions where he doesn't know what to do and holds in threaten state for prolonged amount of time. Headaphones/earphones are exactly that because there is no consequent time domain.
This is perfectly natural, attacking it on biased bases will only make things worse.
No one in science doubt people perceive colour's differently from many reasons; being trained differently, colour blind, various vision deficits, ambient lightning and so on. There are to what's considered absolute colourimetry and from what we so far reached 80% in best HDR format Dolby Vision there so far are device profiles darker/brighter and to ambient light and probably will be more in the future.
Anyway in every other even sensory area scientific experiment's advanced much more than in audio. Only answer is; work, work and more work. It's been done much better in other areas experiencing similar or even worse subjective impact but it didn't fall out from the sky. When ever someone mentions the "Harman" research experiment's as exampary and something that should be taken as relevant hair on my head rises up. Only thing that such can be example of is how not to do it and show how people are delusional and as a prime example of statistically not relevant or significant model. Future more you are all discussing things that you claim aren't measured accurately but in reality are simulated bad trying to change and influence things in the way it simply doesn't work and base your opinions on why such failed to meet desired demand in the first place.
 
Interesting. So this drawing is the FR perception our brain has to deal with. Speakers headphones or more or less both?
For speakers it is somewhere between the (upper in the drawing) 30 to 0 degrees mixed with the -30 degrees.
For headphones it is the 90 degrees response. Some drivers are at an angle but this is rarely more than a few degrees (often around 6 degrees).

The idea is that the tuning of headphones (when the goal is speakers in a room) is done in such a way that it would mimic flat speakers in a room (so around 30 degrees plot)
Of course it isn't that simple as we also hear reflections, there are room resonances etc.
 
No one in science thinks that color accuracy on a televisions is derived by opinions. They use objective measurements.
Indeed no one thinks that also because it is very easy to measure reliably and with accuracy. No human perception is needed... especially not color blind people.
Meaning the comparison between sight and hearing goes lame because of the nature of the measurements.


The HRTF's as a function of timbre are very relevant , but in a general or average of the population way.
As shown by Dr. Olive in the Harman research. This yielded an average... the Harman curve ... which was updated over the years as it wasn't as cut and dry as initially thought.

engineers seek to quantify the physics of sound in order to more reliably and repeatedly get closer too true fidelity of the signal.
Very easy in the electronics realm. more difficult in the acoustic and transducer realm ... and then there was preference. After all people buy based on what the prefer.
Were back to Harman research which was done using a specific test fixture (that deviated a little from the current available ones).
 
All current HATS or just head and pinna systems are not consistent with one another(they arrive at slightly different results) ie; there is no real standard. Just deviations on a theme.
Although there's truth to what you say, in fairness the work hasn't been done to confirm or deny this. At first glance I err on agreement to the extent that, if you apply the same filters used in the Harman research to the 5128 DF response, it is easy to see headphone measurements compensated to this target diverge in unpredictable ways compared to Harman target on GRAS. I don't know enough to be able to say whether this is expected and I'm being naive or something else.

Where I disagree more is we currently have a standard which is the Harman OE 2018 target on the GRAS system. The "deviations on a theme" come from elsewhere, there are people using DF tilts, preference bounds, all sorts. If you whittle it down to the basics, which it seems to me what Amir does, and just see whether a headphone broadly follows this OE2018 target when measured on the appropriate GRAS system, we seem to get most of the way to finding out if the headphone will sound "correct". I'd be more suspicious if people were measuring headphones which followed this target and thought that in reality they sounded trash, but this doesn't seem to be the case.
Isn't that the goal of audiophiles, audio engineers, and designers in general? In some sense anyway?
Maybe striving for some kind of perfection, but I said "a perfection which doesn't exist". Whatever their designing still has to be bound by physical limits. This is what me, @solderdude and others are trying to explain, it feels like you're massively simplifying the fundamental difference between headphones and speakers.
I agree with your general conclusion here. The only part I feel is missing the point. Is that Smyth is emulating a defined sound system in a defined space. It is not designing a playback system but using an existing one and then overlaying a series of filters for each ear. This holographic effect is basically a really advanced binaural recording. It's not all that mysterious. There is a lot more utilized in their processing but there is no reason to go down that rabbit hole. But let's be clear, the Smyth system has nothing to do with understanding good measurements for headphones. Conflating HRTF's with objective analysis of some independent playback system is erroneous. Just because loudspeakers in a room is a different way of listening to music does not mean that the way we analyze objective metrics of headphones needs to be anti-scientific and confusing. Reading a headphone metric should be as easy as reading a Klippel NFS set of data for a loudspeaker. Any reasoning that it shouldn't be becomes some religious cult belief that deviates from the physics of acoustics.
Again I think this shows you are oversimplifying the difference between headphones and speakers. You'd be correct if they weren't so different. However, from what I know about it, the default setup of the Smyth is using HD800S headphones. These are acoustically very open but still bypass the person's head and torso and still clearly interact with people's ears in a way that is problematic enough to require the Smyth setup to take individual HRTF measurements to compensate this difference.

We don't analyze objective metrics of headphones differently just to be "anti-scientific and confusing", we do so because they are fundamentally different to the point they can't be measured in the same way. Think of other media that can be consumed in different ways, for example images. You can take the same image and either print it on paper or project it, or view it on a screen, and perhaps more. You are trying to reproduce the same image but in completely different ways. Your argument seems to me similar to someone arguing there should be no need to consider the intensity of light needed for projecting an image because we don't need to do so for printing it on paper.

This is more extreme than the difference in headphones and speakers in some ways, but not others. Headphones are coupled to your head and ears, in doing so interacting differently with physical aspects as large as your head down to as small as your ear canal. In some cases bypassing these or even physically modifiying them, thus impacting the frequency response heard differently across different people. Even if none of this were true, you're still playing L and R channels directly into each respective ear, using small transducers held in cups which seal against your head to varying degrees, and hearing no reflections due to there literally being no room! There's no reflection or directivity.

This doesn't happen with speakers, as others have tried to explain. I'd be agreeing with you if this wasn't the case, but it is!

So when you say "Reading a headphone metric should be as easy as reading a Klippel NFS set of data for a loudspeaker." this seems to me unscientific because it is an assumption nobody else is making, and which seems completely counterintutive. Due to how different headphones and loudspeakers are, I'd say having made that statement the burden of proof is on you to explain why the differences are not as consequential as previously thought.

You've given some credentials regarding your background, which I don't come close to having, but this make me more confused that you can't appreciate this difference between headphones and speakers which makes it necessary to be measuring them differently in the manner we do. I think I and others have explained as well as we can.
 
I am not certain about what the is the very best way to analyze headphones acoustically. I only understand (after using many of the current HATS systems) that the results we obtain currently are unreliable and not very repeatable (hence why so many rigs tell you to run an average which in turn smooths the curve). The old argument about subjectivity has been used for a long time by audiophiles. Research shows us that this is untrue and that among large segments of the population prefer flat response with smooth directivity to a statical significance in double blind studies. But here we go again with headphones, and now we are back too "its all subjective". While we ignore looking into accurately quantifying the actual acoustic performance of a given headphone. HRTF's are not subjective but are unique to each individual. Audio reproduction is science whether you like it or not. Steely Dan is subjective (whether you like it or not).
This seems like a straw man argument, I don't know where you've got this view that we've thrown our hands up in despair and said "its all too subjective"? People have quantified the acoustic performance of headphones, you are just unhappy with the devices used and perhaps some of the methodology if I understand you correctly, in part because as outlined in my previous posts I think you oversimplify how different headphones and speakers are to necessitate this difference in methodology.
 
The confusion is because we are listening to the sound power of loudspeakers in a room. So the reference changes from what a loudspeaker is doing(Frequency Response). But we currently measure these so differently that they have little to do with one another. Having said that the timbre or target function can be simulated as it already is to some degree in headphones with the built in pinna gain as seen above. The part that gets confusing is the total convolution due to the fact that the room is both diffuse and non-reflective according to the reflective and absorptive coefficients of the room combined with the loudspeakers directivity or overall sound power. Then we have the spatialized part which is the nature of stereo listening . Otherwise known as the HRTF's. There is a timbre function associated with it that adds to the overall sound power or perceived timbre. The typical shelf function around 2K from the right ear listening to the left loudspeaker and the shading from the head (part of HRTF's) But that can be added back into the target function(overall timbre). The only thing I do not address is the cross correlation between the two ears. This can all be measured and accounted for. None of it will change what is essentially an accurate headphone is, regardless of humans all having slightly different HRTF coefficients. IEM's admittedly will always have issues because they bypass our pinna filters, and since that is a known variable then without taking DRP metric on each person it will be close but not entirely accurate for everyone. The assumption being made by the industry is that since HRTF's are all different, then the target functions are all different. This is not correct. Perception of each individuals experience will not and does not alter the reference. Somehow there seems to be a consensus that our perception determines the target function of a headphone. No one in science thinks that color accuracy on a televisions is derived by opinions. They use objective measurements. You can surely alter the picture if you like, but no one in the industry would use that altered version as a reference. In the science of the reproduction of sound, fidelity is not a moving target. If you are not seeking accuracy that's fine, but engineers seek to quantify the physics of sound in order to more reliably and repeatedly get closer too true fidelity of the signal. The HRTF's as a function of timbre are very relevant , but in a general or average of the population way. Not tuning to each persons ears. You want the frequency balance that the loudspeakers create. If you can add the cross correlation , even better, but you would have to create standards for that as well. Someday soon...

I still don't understand how you would define "what is essentially an accurate headphone is" in a manner other than what has been attempted already? Say I sit you in a room with speaker setup you are happy with, all the parameters you've defined, according to the research you selected etc. Just to clear any confusion so this is the room with the speakers you are agreeing is how you think listening to music over speakers should sound like. Then I give you a load of different headphones. What methodology are you proposing to be able to objectively describe how they sound in comparison to your speakers in a room? Or even better to then correct any deviations to the point where you end up with a target which you can then say, an ideal headphone should measure like this according to my methodology?
 
Back
Top Bottom