How do you hear headphone 'soundstage'

[dasdoing]

Waves Abbey Road studio 3........

I hated it.
only could it make kind of enyoible turning "circumference" and "ear to ear" down to the lowest values:

Shannon Pierce's BRIRs are the best there is.
tested many others

 

[wiggum]

Soundstage is quite an unknown property in headphones, how it's created, and which if any measurements relate to it. So it is the dark art side of headphones!

From what I have read, the notch near 10kHz when listening to a pair of speakers at ear level and interaural time difference(ITD) helps our brain locate the source in a three-dimensional space. If the source is above the ear plane, the notch moves to the right like 11kHz or above.

Many over-ear/on-ear headphones do create the 10kHz notch, but they are unable to create ITD, so soundstage is difficult to reproduce, though I have felt some soundstage(not always) with open headphones like Philips SHP9500. I guess the sound leaks through the grills and arrives at the opposite ear.

 

[Robbo99999]

From what I have read, the notch near 10kHz when listening to a pair of speakers at ear level and interaural time difference(ITD) helps our brain locate the source in a three-dimensional space. If the source is above the ear plane, the notch moves to the right like 11kHz or above.

Many over-ear/on-ear headphones do create the 10kHz notch, but they are unable to create ITD, so soundstage is difficult to reproduce, though I have felt some soundstage(not always) with open headphones like Philips SHP9500. I guess the sound leaks through the grills and arrives at the opposite ear.

I don't think it's about sound leakage from one cup to the other, I don't think that's gonna be loud enough to be noticeable. Large earcups and angled pads/drivers whilst not touching the ear if possible seems to me to be one of the recipes, and measured frequency response will affect it too, it's a combination of things I think, but we just don't know why certain headphone models seem to have common trends in how people describe the soundstage properties. From my own point of view I can't offer anymore insight than that.

 

[iGude]

I don't think it's about sound leakage from one cup to the other, I don't think that's gonna be loud enough to be noticeable.

Actually, the leakage from one one cup to the opposite cup in open back headphones is weak but above the threshold of perception, i.e. it does influences the perceived soundstage to some extent. However, it is significantly weaker than the crosstalk of speakers in a stereo triangle.

 

[solderdude]

Actually, the leakage from one one cup to the opposite cup in open back headphones is weak but above the threshold of perception, i.e. it does influences the perceived soundstage to some extent. However, it is significantly weaker than the crosstalk of speakers in a stereo triangle.

Can you show me the research on that ?
I have tried some crude measurements and tests myself and the leakage is insignificant, even with fully open planars and using test tones.
About 30dB L-R differences might be audible with beatles type of recordings.

 

[Ilkless]

Many over-ear/on-ear headphones do create the 10kHz notch, but they are unable to create ITD, so soundstage is difficult to reproduce, though I have felt some soundstage(not always) with open headphones like Philips SHP9500. I guess the sound leaks through the grills and arrives at the opposite ear.

Crosstalk is negligible -- the more likely explanation is that ITD (and ILD) does exist embedded in the recording as a function of the microphone, recording and mixing technique. That's why one school of thinking in loudspeaker playback aims to remove crosstalk! A good eardrum FR that approximates the frequency aspect of the HRTF may well lead to plausible externalised spatial cues being perceived.

Per Prof Edgar Choueiri of Princeton's 3D Audio and Applied Acoustics lab:

Any properly recorded stereo album has ITD and ILD cues embedded in the recording, and it is these cues that present the normal stereo image. If you record using Spaced Omnis, for example, these will capture a strong ITD signal, and will capture, for example, the reverb of the recording environment.

ORTF recordings use cardioid mikes, which tend to emphasize the ILD cues, since they are typically too close together for a strong ITD signal. So most acoustically recorded recordings will produce an extraordinarily impressive and satisfying 3D spatial image. So you will hear a very strong 3D image, and not the normal stereo image locked to the speakers. It just won’t necessarily be spatially accurate.

 

[wiggum]

Can you show me the research on that ?
I have tried some crude measurements and tests myself and the leakage is insignificant, even with fully open planars and using test tones.
About 30dB L-R differences might be audible with beatles type of recordings.

Rtings actually measure these parameters.

Look at their explanation here.

https://www.rtings.com/headphones/learn/open-vs-closed-back. (Read the soundstage section).

 

[wiggum]

Crosstalk is negligible -- the more likely explanation is that ITD (and ILD) does exist embedded in the recording as a function of the microphone, recording and mixing technique. That's why one school of thinking in loudspeaker playback aims to remove crosstalk! A good eardrum FR that approximates the frequency aspect of the HRTF may well lead to plausible externalised spatial cues being perceived.

Per Prof Edgar Choueiri of Princeton's 3D Audio and Applied Acoustics lab:

Look at the above rtings link I posted above. I don't think crosstalk is negligible. If the crosstalk is above the threshold of hearing, then it is audible. While psyco-acoustics might come into play and we may not be cognizant of it, the ITD and ILD contributions of those crosstalk affect the perception of the soundstage.

 

[solderdude]

Rtings actually measure these parameters.

Look at their explanation here.

https://www.rtings.com/headphones/learn/open-vs-closed-back. (Read the soundstage section).

What they measure is leakage measured just outside of the driver.
What isn't measured is how much of that leakage makes it to the other ear and is heard.
It is high frequency mainly and this means the phase and amplitude will differ from what is shown in the measurements.
Also the high frequencies that make it outside will be attenuated on the other ear (so leakage + isolation should be considered) including attenuation by the 'path' the leakage has to take.

The below is not shown in measurements but rather an assumption and theory.
A driver measured from 2cm distance and attenuated rear sound 'bouncing on walls' meters away will be immensely attenuated before it reaches the other ear.
The theory below is fun and needed to 'create' a number but IME not well researched and just theorized.

Open and semi-open headphones leak a lot, which allows some of the sounds that escape to bounce off the walls and objects in your environment, providing the subtle ambiance of your surroundings. This gives the sound quality a bit more of a speaker-in-a-room feel than just regular headphones on your head.

 

[iGude]

Can you show me the research on that ?

I’m not brave enough to write a statement like this on ASR without backup from research

From the Loudspeaker and Headphone Handbook by John Borwick:
The attenuation [of the interaural leakage] at 2 kHz is 30 dB, with an interaural delay of 1 ms improving out-of-head localization for a single channel when the other is silent.

Hence, it can be considered negligible for stereo signals, but not for research e.g. on audibility thresholds etc.

 

[solderdude]

But what does bone conduction do and sound conduction via the headband ?
How relevant is all of this with actual music and how to prove that it actually would influence stereo imaging.
One would have to do blind tests with the most 'open' of (semi)-open headphones and find ways to ensure that 'leakage around the head' could be removed.
I wonder if anyone would really do such testing or already did and documented it.
I think research is more focussed on simulating room effects than trying to determine if sound leakage is really problematic with open headphones.

 

[wiggum]

But what does bone conduction do and sound conduction via the headband ?
How relevant is all of this with actual music and how to prove that it actually would influence stereo imaging.
One would have to do blind tests with the most 'open' of (semi)-open headphones and find ways to ensure that 'leakage around the head' could be removed.
I wonder if anyone would really do such testing or already did and documented it.
I think research is more focussed on simulating room effects than trying to determine if sound leakage is really problematic with open headphones.

You can do a simple test. Select some music where you have heard soundstage and pickup a open headphone. Listen to it at a corner of a room and middle of a room at the same volume level.

Do you feel a difference? That would give you an idea.

 

[Robbo99999]

What they measure is leakage measured just outside of the driver.
What isn't measured is how much of that leakage makes it to the other ear and is heard.
It is high frequency mainly and this means the phase and amplitude will differ from what is shown in the measurements.
Also the high frequencies that make it outside will be attenuated on the other ear (so leakage + isolation should be considered) including attenuation by the 'path' the leakage has to take.

The below is not shown in measurements but rather an assumption and theory.
A driver measured from 2cm distance and attenuated rear sound 'bouncing on walls' meters away will be immensely attenuated before it reaches the other ear.
The theory below is fun and needed to 'create' a number but IME not well researched and just theorized.

Yeah, I read that bit too over on RTings re the sound "bouncing off walls", seemed like utter rubbish to me, no way I can imagine that happening in the scope of headphone listening, the levels just aren't loud enough, I don't know how they could even think about writing that. (You explained the problems well).

 

[solderdude]

You can do a simple test. Select some music where you have heard soundstage and pickup a open headphone. Listen to it at a corner of a room and middle of a room at the same volume level.

Do you feel a difference? That would give you an idea.

That is not controlled and highly subjective. It will tell you nothing really.

I have done some tests and measurements and nothing pointed to any significant crosstalk and me hearing sound bouncing of walls.

 

[Ilkless]

The more likely answer is that ITD and ILD cues lie in wait, embedded in a source recording (as highlighted by Prof Choueiri), to be played back on a headphone (with or without EQ) that produces a favourable HRTF; when combined, you get plausible spatial + tonal cues. Crosstalk (if even audible; auditory masking was not accounted for in above analyses) would actually add a confounding layer of ITD and ILD on top of what is embedded in the recording as a consequence of microphone and mixing technique.

 

[Deleted member 58722]

I hated it.
only could it make kind of enyoible turning "circumference" and "ear to ear" down to the lowest values:

Shannon Pierce's BRIRs are the best there is.
tested many others

Each to their own....

 

[iGude]

The more likely answer is that ITD and ILD cues lie in wait, embedded in a source recording

Agree, still there seems to be an audible difference between open and closed headphones in addition:
“Thus a closed headphone can be corrected to have the same response as a given open headphone. They should therefore sound the same, and yet listening tests with normal stereophonic recordings confirm that IHL [in-head-localization] is worse with closed headphones.”
From the same source cited above.

 

[solderdude]

Rear reflections against the cup may be much more audible in closed headphones than when a majority of the rear fired sound is reflected back.
There is only very little attenuation and a very short time delay between direct and reflected sound.
This, however, has nothing to do with reflections in the room which is what is sometimes claimed to be the reason.
Also note this aspect can create a 'cuppy' sound signature that cannot be or DSP'ed nor EQ'ed out.

 

[Sgt. Ear Ache]

I drew a diagram of what soundstage feels like for me with headphones...yay me! lol

something like that. My head is the little brown circle and the big black circles are 3D balloons of space rather than flat horizontal platters. The way soundstage seems "better" or "worse" for me with different headphones manifests as bigger or smaller black orbs. With really good headphones it feels like the space around my ears is big and there's lots of room between the instruments and sounds. It's never a sense that there's a physical stage in front of me with musicians standing there playing. It's sounds floating in positions around my head in 3 dimensions (although never really far behind) with an impression of spaciousness or compactness of space.

 

[Robbo99999]

I drew a diagram of what soundstage feels like for me with headphones...yay me! lol

View attachment 285476

something like that. My head is the little brown circle and the big black circles are 3D balloons of space rather than flat horizontal platters. The way soundstage seems "better" or "worse" for me with different headphones manifests as bigger or smaller black orbs. With really good headphones it feels like the space around my ears is big and there's lots of room between the instruments and sounds. It's never a sense that there's a physical stage in front of me with musicians standing there playing. It's sounds floating in positions around my head in 3 dimensions (although never really far behind) with an impression of spaciousness or compactness of space.

For me if I was gonna use your diagram I'd position my head right at the centre of the those two circles, mainly because I don't get masses of forward projection but I certainly hear some rearward projection in some tracks (that have that baked in effect) as well as always left & right out of head (with correct headphone) and forwards a little bit in most tracks general presentation, but not massively forwards.