Proximity effect and the psychoacoustically wrong target FR curves

[Geert]

(I'm also aware that near-field monitoring requires less bass; not sure how it adds to the discussion)

You just brought it up, I didn't mention it.

Now that the brain recognizes that the sound source is nearer than a speaker-in-a-room, yet has the FR of speaker-in-a-room, it is "confused".

People have spend 2 pages arguing why this confusing probably doesn't exist but it seems like these arguments didn't stick. One more attempt:

For large music events we usually use very directional low frequency speakers (horns, arrays, cardioid systems). They produce a wave front alike sound field, which means the level of lows doesn't drop of with distance as fast as what happens with hifi speakers. Now do you believe when you walk from a far distance to the stage your brain is getting confused because it expects the sound to change while it doesn't? It doesn't, instead the whole idea of such a setup is to get a uniform sound field across the whole area.

I am trying to give a theory as to why headphone preference targets are warmer than the DF target, even if the DF target is bass-corrected. I am not rejecting preference targets, but trying to derive them from combining DF target and proximity effect.

You should first start with validating if the problem you envisioned is real, meaning a statistically relevant amount of people share your idea. Until then we have the actual preference curves, which represent the current scientific consensus.

 

[ChrisG]

I think @IAtaman hit the nail on the head with "acoustic loading" as a reason for the difference. Take a small transducer (such as a 4" full-range) that is "tuned" for the box into which it is inserted, and take it out. Lay it on the table. The same difference in sound will occur as a pair of headphones, and yet it's not proximity effect. It's loading and cancellation .... nothing to do with the brain.

Jim

You've introduced a confounding variable: removing a speaker from an enclosure creates a dipole system, which are notoriously lossy.

Long time HF user here. Decided to start this discussion on ASR since afaik you are more welcoming to science-related discussions.

The proximity effect in audio is an increase in bass or low frequency response when a sound source is close to a microphone. When you speak closely into a microphone, your voice sounds more chesty. When you speak far into a microphone, your voice sounds more "normal". So far so good. Our eardrums also have proximity effects, because they function much like a microphone.

There's a subtlety that you've missed.

Almost all directional microphones exhibit proximity effect. The select few that don't have to pull off a lot of trickery to avoid it. See: EV RE20.
Directional microphones are all somewhere along the line of pressure (omnidirectional pattern) and pressure gradient (aka velocity, fig-8 pattern) transducers. Cardioid is the half-way point, and likely the most common directional pattern among microphones. Sub/wide-cardioids are closer to omnidirectional/pressure, while super/hypercardioids are closer to fig-8/velocity.
Omnidirectional mics do not exhibit proximity effect, but fig-8 mics do. As expected, all the mics along that scale exhibit proximity effect to greater or lesser degree.


The very important thing you've missed in your reasoning is this: our ears act like pressure-based microphones. ie, omnidirectional, ie, do NOT exhibit proximity effect.


Anyone caught up in the microphone-related-proximity-effect discussion, then, is automatically wrong. It's not happening here.



What IS happening here is this: baffle step.

"Baffle step" (and its associated treatment, Baffle Step Compensation) is a step-down in the frequency response that occurs when you take a loudspeaker driver with a flat response on a large baffle, and put it on a small baffle. Over a frequency range related to the dimensions of the baffle, the low-frequency output drops by up to 6dB. This is because the radiation has essentially gone from hemispherical to omnidirectional. Half the pressure going forwards means 6dB is lost on-axis.

This is why close-up loudspeaker measurements require additional processing before they're useful: when you put a microphone in the acoustic nearfield, you capture the signal before 6dB is lost by wrapping around the cabinet and going backwards. Pull the mic further away, and the change in frequency response is obvious.


So, when somebody whispers in your ear and it sounds all bassy, it's because their voice is no longer suffering from baffle-step losses.



Finally, we can go back to headphones.

Headphones have a few different regions of operation. At the mid-high range, you've got a relatively large source pointing at your outer ear, and that's going to cause all sorts of interesting effects. The desirable frequency response is non-linear, for sure.

We're concerned about bass here, though. Thankfully, a helpful forum member has pretty much solved this for us: https://www.audiosciencereview.com/...or-people-wearing-glasses.24574/#post-1759664

When glasses are worn, the headphone driver is essentially still in the same place relative to the ear. The big difference is that the acoustic chamber containing the headphone driver, air space and our ear becomes leaky.

Below the modal region of a room, it acts as a pressure vessel, which will exhibit a 12dB/octave rising response towards the LF. Real-world rooms are leaky, so you won't get all twelve of those decibels, but you'll probably get some.
Headphones work in the same way, except obviously the pressure zone starts at a much higher frequency. @solderdude 's post shows this quite clearly.

Now, there's all sorts of caveats about how well-sealed our headphones will be, and even things like hair between the earcup and skin will show some loss in the bass. Chances are that some averaging has to be done for manufacturers to work out what's going to work for most people, most of the time.


Hopefully this clears up a few things.

Chris

 

[fpitas]

I guess I always assumed that the discrepancy was down to lack of baffle step power effect with a headphone, and the fact that there is no tactile bass with a headphone. When you adjust a speaker's baffle step for good bass, you've usually boosted the bass power going into the room considerably.

 

[thanrl]

There's a subtlety that you've missed.

Almost all directional microphones exhibit proximity effect. The select few that don't have to pull off a lot of trickery to avoid it. See: EV RE20.
Directional microphones are all somewhere along the line of pressure (omnidirectional pattern) and pressure gradient (aka velocity, fig-8 pattern) transducers. Cardioid is the half-way point, and likely the most common directional pattern among microphones. Sub/wide-cardioids are closer to omnidirectional/pressure, while super/hypercardioids are closer to fig-8/velocity.
Omnidirectional mics do not exhibit proximity effect, but fig-8 mics do. As expected, all the mics along that scale exhibit proximity effect to greater or lesser degree.


The very important thing you've missed in your reasoning is this: our ears act like pressure-based microphones. ie, omnidirectional, ie, do NOT exhibit proximity effect.


Anyone caught up in the microphone-related-proximity-effect discussion, then, is automatically wrong. It's not happening here.



What IS happening here is this: baffle step.

"Baffle step" (and its associated treatment, Baffle Step Compensation) is a step-down in the frequency response that occurs when you take a loudspeaker driver with a flat response on a large baffle, and put it on a small baffle. Over a frequency range related to the dimensions of the baffle, the low-frequency output drops by up to 6dB. This is because the radiation has essentially gone from hemispherical to omnidirectional. Half the pressure going forwards means 6dB is lost on-axis.

This is why close-up loudspeaker measurements require additional processing before they're useful: when you put a microphone in the acoustic nearfield, you capture the signal before 6dB is lost by wrapping around the cabinet and going backwards. Pull the mic further away, and the change in frequency response is obvious.


So, when somebody whispers in your ear and it sounds all bassy, it's because their voice is no longer suffering from baffle-step losses.



Finally, we can go back to headphones.

Headphones have a few different regions of operation. At the mid-high range, you've got a relatively large source pointing at your outer ear, and that's going to cause all sorts of interesting effects. The desirable frequency response is non-linear, for sure.

We're concerned about bass here, though. Thankfully, a helpful forum member has pretty much solved this for us: https://www.audiosciencereview.com/...or-people-wearing-glasses.24574/#post-1759664

When glasses are worn, the headphone driver is essentially still in the same place relative to the ear. The big difference is that the acoustic chamber containing the headphone driver, air space and our ear becomes leaky.

Below the modal region of a room, it acts as a pressure vessel, which will exhibit a 12dB/octave rising response towards the LF. Real-world rooms are leaky, so you won't get all twelve of those decibels, but you'll probably get some.
Headphones work in the same way, except obviously the pressure zone starts at a much higher frequency. @solderdude 's post shows this quite clearly.

Now, there's all sorts of caveats about how well-sealed our headphones will be, and even things like hair between the earcup and skin will show some loss in the bass. Chances are that some averaging has to be done for manufacturers to work out what's going to work for most people, most of the time.


Hopefully this clears up a few things.

Chris

Thanks for the perspective. Are you postulating that tapping my fingers 1 inch away sounds more bassy than 2 inches away is due to the closing in of the fingers providing a "seal" instead of due to a proximity effect?

Regarding omnidirectional mics: are there researches regarding the change in FR when an omnidirectional mic is placed at the end of a tube? For example, end correction can contribute to that. I suspect the trend of the effect is similar to the directional microphone proximity effect.

Regardless if the increase in bass is due to a proximity(-like) effect or a better seal, I think my conclusion stands: they are cues that alert our brains to anticipate a FR that is warmer than a FR that simulates speaker-in-a-room. Do you think it is fair to draw such a conclusion given the limited understanding about the behavior of an omnidirectional microphone in a tube?

 

[thanrl]

You should first start with validating if the problem you envisioned is real, meaning a statistically relevant amount of people share your idea. Until then we have the actual preference curves, which represent the current scientific consensus.

I'm so confused. The problem I envision is to explain the discrepancy between the DF target and the preference targets. If there is a discrepancy, then the problem is real. Are you trying to say that you (or statistically many people) enjoy the DF target?

 

[Geert]

Are you trying to say that you (or statistically many people) enjoy the DF target?

As far as I know most people don't listen to DF (diffuse field) targets. Instead they use the Harman curve as a basis, which has elevated bass.

The problem I envision is to explain the discrepancy between the DF target and the preference targets

There is a difference between these targets because of they way they are obtained. The Harman curve has elevated lows especially to make it sound like how we here sound in a room. It's done on purpose as a solution, so how is this discrepancy a problem?

Put a hifi loudspeaker in open air and you'll notice it also lacks bass. Just like with headphones, the reason is it's missing the bass build up like we have in a room. We don't need to invent a proximity effect to explain this.

If you believe the proximity effect is real you need to test and proof it on it's own, not via analogies which introduce other variables (like your example of the sound of a headphone on a table where we know it no longer functions as a pressure chamber, or bringing an omnidirectional source closer to the ears).

 

[thanrl]

There is a difference between these targets because of they way they are obtained. The Harman curve has elevated lows especially to make it sound like how we here sound in a room. It's done on purpose as a solution, so how is this discrepancy a problem?

The DF target (ISO 11904-2) is a deduced target, obtained by dummy heads and microphones. The Harman target is an induced target, obtained by humans and equalizers. Both claim to recreate speakers in an echoic chamber. There is a discrepancy between the deductive and the inductive science. I respect you for not being interested in resolving this discrepancy.

 

[Geert]

The DF target (ISO 11904-2) is a deduced target, obtained by dummy heads and microphones. The Harman target is an induced target, obtained by humans and equalizers. Both claim to recreate speakers in an echoic chamber.

Looks like we finally arrived at the crux of the matter. The Harman curve is based on the frequency response of a dummy head captured in a "critical listening room" while the Diffuse Field curve is measured in a room with no direct sound or direct reflections, leaving only reverberation (sound arrives from all angles at equal level) . So these rooms are not identical rooms. The critical listening room has a bass build up. Research shows the Harman curve is what people prefer https://www.aes.org/e-lib/browse.cfm?elib=16768.

Dr. Sean Olive: "The goal from the beginning was to make a headphone sound like a pair of good loudspeakers in a reference listening room. We believe that because of this, we're as close as we can get to making a headphone sound like a pair of good loudspeakers in a critical listening room".

Ooh, and both the DF response and Harman response were measured on dummy heads, so how could the difference between both curves ever be the result of a proximity effect?

I respect you for not being interested in resolving this discrepancy.

You must be kidding. By now I've probably spend more than an hour discussing it with you (and a lifetime studying it).

 

[ChrisG]

Thanks for the perspective. Are you postulating that tapping my fingers 1 inch away sounds more bassy than 2 inches away is due to the closing in of the fingers providing a "seal" instead of due to a proximity effect?

Regarding omnidirectional mics: are there researches regarding the change in FR when an omnidirectional mic is placed at the end of a tube? For example, end correction can contribute to that. I suspect the trend of the effect is similar to the directional microphone proximity effect.

Regardless if the increase in bass is due to a proximity(-like) effect or a better seal, I think my conclusion stands: they are cues that alert our brains to anticipate a FR that is warmer than a FR that simulates speaker-in-a-room. Do you think it is fair to draw such a conclusion given the limited understanding about the behavior of an omnidirectional microphone in a tube?

1 - No. It's almost certainly due to the construction of your head. For me, tapping right behind my hear gives a very "tappy" sound - not much bass at all. My fingertip tells me there's a thin layer of skin before hitting bone, and I expect that's the cause.

2 - The tube is very short/small, only making much difference in the kHz range. Below that, the tube is of basically zero consequence.

3 - No. I don't think our brain's expectations factor into this. An example of speaker vs distance was given previously. Having worked with a wide range of speakers at a wide range of distances, they tend to sound the same (disregarding room acoustics - I've done this outdoors) until either air absorption kicks in (losing treble at large distances), or you get really close and nearfield effects (baffle-step, as noted previously) kick in.


Chris

 

[thanrl]

1 - No. It's almost certainly due to the construction of your head. For me, tapping right behind my hear gives a very "tappy" sound - not much bass at all. My fingertip tells me there's a thin layer of skin before hitting bone, and I expect that's the cause.

Well that's surprising. Try tapping closer to your concha? I can hear more bass tapping near the concha than from an inch away, if I "normalize" the difference in volume in my mind.

2 - The tube is very short/small, only making much difference in the kHz range. Below that, the tube is of basically zero consequence.

I think the consequence may be bigger than you think. Try listening to the same IEM (preferably bullet-shaped ones) at different insertion depths. Or even better, use the same size eartips but vary the depth in which they are fitted onto the IEM nozzle by adding o-rings to the back of the nozzle. The difference in warmth by going from a 1mm o-ring to no o-ring is night and day, especially when the insertion is deep.

 

[ChrisG]

Well that's surprising. Try tapping closer to your concha? I can hear more bass tapping near the concha than from an inch away, if I "normalize" the difference in volume in my mind.

I think the consequence may be bigger than you think. Try listening to the same IEM (preferably bullet-shaped ones) at different insertion depths. Or even better, use the same size eartips but vary the depth in which they are fitted onto the IEM nozzle by adding o-rings to the back of the nozzle. The difference in warmth by going from a 1mm o-ring to no o-ring is night and day, especially when the insertion is deep.

As expected from my earlier post, changing the volume of the acoustic chamber will alter the LF response. This is not the same as altering the directivity pattern.

Chris

 

[thanrl]

As expected from my earlier post, changing the volume of the acoustic chamber will alter the LF response. This is not the same as altering the directivity pattern.

Chris

If changing the volume *by changing the distance to the sound source* (e.g. by 1mm) alters the LF response at the ears, how is that practically different from a proximity effect? Are you trying to explain the cause of the proximity effect at the ears by the baffle step?

 

[ChrisG]

If changing the volume *by changing the distance to the sound source* (e.g. by 1mm) alters the LF response at the ears, how is that practically different from a proximity effect? Are you trying to explain the cause of the proximity effect at the ears by the baffle step?

No, I'm talking about changing the volume of the acoustic chamber containing the headphone driver and the eardrum:

Below the modal region of a room, it acts as a pressure vessel, which will exhibit a 12dB/octave rising response towards the LF. Real-world rooms are leaky, so you won't get all twelve of those decibels, but you'll probably get some.
Headphones work in the same way, except obviously the pressure zone starts at a much higher frequency. @solderdude 's post shows this quite clearly.

Now, there's all sorts of caveats about how well-sealed our headphones will be, and even things like hair between the earcup and skin will show some loss in the bass. Chances are that some averaging has to be done for manufacturers to work out what's going to work for most people, most of the time.

Smaller chamber = higher frequency at which the modal region stops and the pressure zone starts. Given that the average ear canal is 25mm long, it doesn't surprise me that 1mm movements (4% of the volume) are audible.


Chris

 

[thanrl]

No, I'm talking about changing the volume of the acoustic chamber containing the headphone driver and the eardrum:


Smaller chamber = higher frequency at which the modal region stops and the pressure zone starts. Given that the average ear canal is 25mm long, it doesn't surprise me that 1mm movements (4% of the volume) are audible.


Chris

I think there are a few issues, some of which we agree upon, some we don't:
1. Changing the distance of the sound source from the outer ear canal to the inner ear canal, when the ear canal is sealed (think IEM), is there a proximity effect? I think we agree that there is, and you gave an explanation above. My explanation is end correction, but in the end they may very well be compatible.
2. Changing the distance of the sound source from the outer ear to the outer ear canal, when the ear canal is semi-sealed (think open back headphones), is there a proximity effect? I'd argue yes, because even open back headphones can create a semi-sealed pressure chamber. Proximity effect becomes audible when the chamber size decreases.
3. Changing the distance of the sound source from an open space to the outer ear, when the ear canal is open (think near-field monitoring), is there a proximity effect? You say no, and now I tend to agree with you.

And then I think we disagree on the brain's reaction to these proximity effects. Whether the cause of the proximity effect is due to the decrease in distance or the decrease in the size of the chamber, I postulate that when the brain recognizes this change (possibly through the change in temperature, change in airflow or the physical contact of the seal), it anticipates more bass and warmth.

 

[markanini]

The proximity effect is related to microphone technology. It's of concern to recording engineers and mic designers. It's a known phenomenon for specific devices under specific conditions.

It's a big leap to assume that things are carried over when changing the fundamental conditions. That's like assuming alcohol is a healthy part of your diet because it kills germs.

 

[GaryH]

Even if a headphone has a theoretically perfect FR simulation of a speaker-in-a-room (i.e. the DF target)

DF is absolutely not a simulation of a speaker in a normal room. It's a simulation of a speaker in a reverberation chamber, which is obviously not how music is mixed and mastered, so is clearly the wrong target.

Now that the brain recognizes that the sound source is nearer than a speaker-in-a-room, yet has the FR of speaker-in-a-room, it is "confused".

Sounds like you're actually referring to the known 'SLD' (source-to-listener distance) effect: apparent acoustic source distance influences perceived loudness / frequency response.

 

[sq225917]

Having the mic close to the listener and using iems aren't the same thing. When I use iems i want close micd stuff to sound close and far micd stuff to sound far micd.

 

[ahofer]

This strikes me as similar to the discussion of loudness - what if an orchestra somehow played quietly, we would hear the FR differently, so there’s no need to make Fletcher-Munson corrections….or do we want to hear a scaled-down version of what the real thing sounds like. Hard to know the answer.

 

[Cbdb2]

Most of what I listen to is multi-tracked rock produced in the studio so the panning/soundstage is completely artificial anyway...

Not quite. A stereo mic can give real panning cues. Mic distance is also audible without the proximity effect. You hear more room and less upper mid freq. detail. Less breath and lips on vocals, less string squeek on stringed instruments, keys on trumpest, etc.
One of the most obvious examples, the vocal at 30sec. The reverb is a big part of it but try to ignore that and listen to the difference in tone.

I recorded ADR (actor comes in the studio to add/replace dialog shot on set) for tv/movies for a few years and often you would record a line in the middle of a speech so the aim was to match the tone (often to get rid of noise, like an airplane in a period piece). Not as easy as it sounds. Yould think EQ and all the other processing available would make it easy but there were other things to consider, one of the big ones was distance from the mic. There's no processing that can make someone recorded 6 inches from the mic sound like there 6 feet from the mic.

 

[Cbdb2]

Thanks for the perspective. Are you postulating that tapping my fingers 1 inch away sounds more bassy than 2 inches away is due to the closing in of the fingers providing a "seal" instead of due to a proximity effect?

Or is it Fletcher-Munson?