• 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!

How do you hear headphone 'soundstage'

How do you perceive headphone stereo image (without any trickery/Binaural)

  • In my head (Left, Right and inbetween)

  • In the back of my head (Left, Right and inbetween)

  • Slightly in front of my head (Left, Right and inbetween)

  • a full 3D image (all around me)

  • a 2D image clearly in front of me

  • I don't care about this aspect

  • It depends on the headphone (from between to in front of me)


Results are only viewable after voting.

JanesJr1

Addicted to Fun and Learning
Joined
Jun 9, 2021
Messages
504
Likes
449
Location
MA
I reckon I agree with your first paragraph there, but for the bit I bolded in your post I don't think has to be a total total construct, as in I think slightly angled drivers or pads could help provide a few more cues as a portion of your HRTF is injected into the stream that you receive at the eardrum. I mean that goes along with the idea that some designs of headphone have better soundstage than others once you've tried to account for measured frequency response on a dummy head (although I'm aware of the limitations of unit to unit variation throwing a spanner in the works if you're trying to match frequency responses, not to mention the more unpredictable shenanigans above and around 10kHz that you don't really have reliable visibility on from measurements). So I do agree with your first paragraph there, but regarding the bit I bolded in your post - I think it is a construct like you say, but not a total one, there are still cues that are happening & coming from the headphone's characteristic physical design (the soundstage potential of that headphone model).
Agreed.
 

pseudoid

Master Contributor
Forum Donor
Joined
Mar 23, 2021
Messages
5,105
Likes
3,411
Location
33.58 -117.88
I am diggin' @solderdude sharing all this knowledge.
10q!
He had the HD800S and Edition X that he found to have good spatial qualities compared to some other headphones he had at that moment to test.
How does a headphone-designer impart such 'spatial qualities' to 2 independently (L/R) excited electro-mechanical transducers?

This 'spatial-gene' seems to reside in the algorithms of our brain and not just within the primary auditory cortex (temporal lobe), nor just within one side (L/R) of our brain.
I am aware of some MRI studies, of how human brains process music, have shown that there is no one magic brain location to analyze music.
My understanding is that individual electrical signals (L and R) from the auditory nerve is sent directly to the thalamus, which is just above the brainstem.

Based on MRI results (brainfacts.org), it is said that the brainstem/thalamus are responsible for computation of a "sound's" direction and location.
I seems as though this is where the algorithm for signal's spatial processing initially occurs; prior to being sent to the other brain's processing/analysis (e.g. rhythm, beat, tone, pitch, loudness, dynamics, etc.) centers.
MRI studies also found that the gooey-part of our brain that process our emotions also fires up in response to music.

[I am being told that emotions are just a construct! Discussing it would be OT;)]
 
Last edited:

okok

Senior Member
Joined
Oct 31, 2020
Messages
377
Likes
161
speakers got much better day and night soundstage
 
OP
solderdude

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,842
Likes
35,808
Location
The Neitherlands
How does a headphone-designer impart such 'spatial qualities' to 2 independently (L/R) excited electro-mechanical transducers?

That must be a trade secret, I have found no info on how it is 'done' except for the often used 'angled drivers'.
Obviously Dan has made some remarks about it. Willing to bet it won't work for the majority of people.
There must be certain info available otherwise the DSP modules for 3D processing would not work.

For us mortals that do not seem to poses 3D capabilities in our brain (only hear music in the head from left to right) also do not seem to be able to do this with any headphone they tried.
As I have not explored all 3D software/smyth/whatnot I have no idea how I would react nor is this my goal.
I can fully enjoy music on headphones as is so there's that.

The poll showed that the majority of people hears 'spatial qualities' as I do (inside the head) maybe depending on headphone/recording a bit more forward to the head.
Only very few (about 5%) state they get 3D sound (and assume they mean by listening with their normal headphones to regular, not processed, 2CH recordings.
 

JohnnyNG

Active Member
Joined
Dec 24, 2019
Messages
176
Likes
183
I've never had imaging all around me, even with binaural recordings. With binaural, I get some imaging behind my head but nothing in front. And with "normal" stereo it's all in my head. As such, I don't really enjoy headphone listening as the scale is reduced to fit in my head.
 

JanesJr1

Addicted to Fun and Learning
Joined
Jun 9, 2021
Messages
504
Likes
449
Location
MA
Yet... the Stealth and Expanse are not U-shaped at all and are supposed to have better stereo imaging
Dan addresses this in his interview with Head-Fi on the Expanse (search youtube head-fi expanse). He states that the Expanse is less U-shaped than the Stealth, which he says conforms to Harman more closely. He says that with the closed-back Stealth, the lower FR in the upper bass was intended to help the soundstaging, but the same constraint was unnecessary on the Expanse. He wished to achieve an "experientially flat" upper bass for the Expanse, because conforming to the Harman curve sounds too recessed.

Below are the frequency response curves to compare. The Expanse is tuned flatter in the upper bass than either the closed-back Stealth or closed-back Noire. It appears on the Head Fi FR curve to be about 2 db +/- of warmth from about 90-400 Hz.

EXPANSE VS STEALTH: Stealth is more U-shaped from <100Hz to >400 Hz.
1665333393102.png



EXPANSE VS HARMAN:
1665332443878.png


STEALTH VS HARMAN: Stealth has a bit of the 'Dan Clark bump' from about 60-160 Hz, but the Harman-conforming U above that range. I think the Harman curve is fitted just a little higher on this chart vs the Expanse chart above, perhaps letting the Stealth "U" look a little diminished; the RMS level (vertical axis) bottoms out at just under 300 Hz for the Expanse and 550 Hz for the Stealth, as suggested in the Head Fi curves above.
1665332540202.png


NOIRE VS HARMAN: The Noire has a firm bump, followed by a U above 160 Hz. The U partly conforms to Harman, and dips a bit below in the ~200 Hz and 700-1100 Hz bands.

1665333116407.png
 

Attachments

  • 1665330990874.png
    1665330990874.png
    182.1 KB · Views: 34
Last edited:

Emlin

Addicted to Fun and Learning
Joined
Jul 8, 2018
Messages
738
Likes
1,026
For it is like listening to speakers with big soundstage and excellent imaging, except that it does wrap around at the sides a bit with hard panned instruments. Nothing behind me and definitely nothing between my ears(!).

Headphones and iems.

Imaging does seem to improve with better FR.

I don't fit into your categories so can't vote.
 

Phorize

Major Contributor
Forum Donor
Joined
Apr 26, 2019
Messages
1,527
Likes
2,053
Location
U.K
Caveat Emptor: Everything I will write is subjective:
Depends on headphones or IEMs for me as well:

DCA Expanse: A full 3D image with less hard panning than HD800
DCA Stealth: Slightly in front of my head with less 3D depth than Expanse
DCA Aeon X Open / Aeon 2 Noires: Slightly in front of my head without the 3D depth of both Stealth and Expanse
DCA Voce: In my head without the 3D depth of both Stealth and Expanse

AKG K712/K702: A full 3D image with less hard panning than HD800 (closest to Expanse though K712 pans sideways more than Expanse while Expanse has better depth perception)

Raal SR1a: Slightly in the back of my head with ultra-wide panning
HD 800 (S): Slightly in the back of my head with ultra-wide panning
Warwick Aperio and Bravura: Slightly in front of my head with less 3D depth than Expanse
Hifiman Susvara and Focal Utopia: Slightly in front of my head with less 3D depth than Expanse

IEMs:

headstage in this context is defined as not perceiving soundstage as out of the head

CA Andromeda: Full 3D image with out of my head soundstage
AK Pathfinder: Full 3D image within headstage
Oriolus Traillii: Full 3D image that's slightly wider and taller than Andromeda (yes it's more spacious than "headstage")
Vision Ears EXT: Full 3D image that's slightly wider and taller than Andromeda (yes it's more spacious than "headstage")
AK x EE Odyssey: Slightly in front of my head (yes it's more spacious than "headstage")
64 Audio U12T: Slightly in front of my head (yes it's more spacious than "headstage")

Last but not least: Airpods 3 + Custom Spatial Audio: Full 3D image that's more spacious than "headstage"
That’s an impressive headphone collection!
 

majingotan

Major Contributor
Forum Donor
Joined
Feb 13, 2018
Messages
1,508
Likes
1,779
Location
Laguna, Philippines
That’s an impressive headphone collection!

I wish I have all of those lol. I was fortunate enough to demo those headphones at CanJam recently. My only personal units are Dan Clark Aeon X Open and Campfire Andromeda 2018 Blue Color
 

artburda

Active Member
Joined
Feb 12, 2019
Messages
286
Likes
361
Location
Switzerland
I hear the sound stage very well to each side or to the back, but not to the front. What is supposed to be in front sounds more like it's in my head. For example when I listen to this test track
 
OP
solderdude

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,842
Likes
35,808
Location
The Neitherlands
Dan addresses this in his interview with Head-Fi on the Expanse (search youtube head-fi expanse). He states that the Expanse is less U-shaped than the Stealth, which he says conforms to Harman more closely. He says that with the closed-back Stealth, the lower FR in the upper bass was intended to help the soundstaging, but the same constraint was unnecessary on the Expanse. He wished to achieve an "experientially flat" upper bass for the Expanse, because conforming to the Harman curve sounds too recessed.

Below are the frequency response curves to compare. The Expanse is tuned flatter in the upper bass than either the closed-back Stealth or closed-back Noire. It appears on the Head Fi FR curve to be about 2 db +/- of warmth from about 90-400 Hz.

EXPANSE VS STEALTH: Stealth is more U-shaped from <100Hz to >400 Hz.
View attachment 236330


EXPANSE VS HARMAN:
View attachment 236321

STEALTH VS HARMAN: Stealth has a bit of the 'Dan Clark bump' from about 60-160 Hz, but the Harman-conforming U above that range. I think the Harman curve is fitted just a little higher on this chart vs the Expanse chart above, perhaps letting the Stealth "U" look a little diminished; the RMS level (vertical axis) bottoms out at just under 300 Hz for the Expanse and 550 Hz for the Stealth, as suggested in the Head Fi curves above.
View attachment 236322

NOIRE VS HARMAN: The Noire has a firm bump, followed by a U above 160 Hz. The U partly conforms to Harman, and dips a bit below in the ~200 Hz and 700-1100 Hz bands.

View attachment 236327

This is not in anyway related to spatial properties though. Only with bass impact. I fully agree b.t.w. bass impact is in that region.
I made this plot many years ago: Notice the 'punch, slam' descriptors. Too much of it becomes negative: muddy, boomy, rubbery bass'.

descriptors2.png


For it is like listening to speakers with big soundstage and excellent imaging, except that it does wrap around at the sides a bit with hard panned instruments. Nothing behind me and definitely nothing between my ears(!).

Headphones and iems.

Imaging does seem to improve with better FR.

I don't fit into your categories so can't vote.

That would be a 3D experience in my book and out of the head.
5% voted for that.
 
Last edited:

xaviescacs

Major Contributor
Forum Donor
Joined
Mar 23, 2021
Messages
1,493
Likes
1,971
Location
La Garriga, Barcelona
Nice poll. Regardless on the headphone, over ear or IEM, I hear the music, I imagine the players, just in front of me, very close. With IEMs there is a bit of "In my head". This is however quite loose, and if while listening I imagine they are far away or inside my ears, suggestion works and perception changes.

P.S.: With my digital piano (GP-510) it sounds more in my head.
 

Emlin

Addicted to Fun and Learning
Joined
Jul 8, 2018
Messages
738
Likes
1,026
This is not in anyway related to spatial properties though. Only with bass impact. I fully agree b.t.w. bass impact is in that region.
I made this plot many years ago: Notice the 'punch, slam' descriptors. Too much of it becomes negative: muddy, boomy, rubbery bass'.

descriptors2.png




That would be a 3D experience in my book and out of the head.
5% voted for that.
But you have 3D as going all around. I don't get that except with binaural recordings.
 
OP
solderdude

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,842
Likes
35,808
Location
The Neitherlands
3D can be 'all around you' or as an image speakers would create so with depth and in front of you.
It is quite remarkable you do not hear sounds 'in' our head at all.

2D is more like a line through or just in front of the head so left to right and in between but not 'depth'
 

Emlin

Addicted to Fun and Learning
Joined
Jul 8, 2018
Messages
738
Likes
1,026
3D can be 'all around you' or as an image speakers would create so with depth and in front of you.
It is quite remarkable you do not hear sounds 'in' our head at all.

2D is more like a line through or just in front of the head so left to right and in between but not 'depth'
Thanks, I'll vote 3D then. I confirm that that is my head is empty, both ways.
 

DualTriode

Addicted to Fun and Learning
Joined
Oct 24, 2019
Messages
891
Likes
593
Hello All,

This may be of interest to some.

The gang of PhD's over a Facebook Reality Labs have recently written a AES published paper on this topic.

The Harman Target transfer function may be preferred for "music content" and a flat transfer function may be preferred for "spatial content".

Personally I feel that this concept needs more investigation. Music can be pretty spatial.


Thanks DT
 
Last edited:

JanesJr1

Addicted to Fun and Learning
Joined
Jun 9, 2021
Messages
504
Likes
449
Location
MA
This is not in anyway related to spatial properties though. Only with bass impact. I fully agree b.t.w. bass impact is in that region.
I made this plot many years ago: Notice the 'punch, slam' descriptors. Too much of it becomes negative: muddy, boomy, rubbery bass'.

descriptors2.png




That would be a 3D experience in my book and out of the head.
5% voted for that.
I always wondered who made that wonderful graphic, the best one that I've seen. Thank you; I have kept it close by on my computer for some time.

In my other posts, I've made the point that a realistic, U-shaped FR response may enhance the fidelity of playback in a number of ways related to soundstage. In conjunction with other spatial cues, it does help the placement of soundstage outside of my head. it improves on a mids-focus FR in general or on other spatial cues by themselves in generating a realistic soundstage. My point has been simply that it is at least plausible that not only manipulations of arriving sound signals at the eardrum with directionality or other interaural cues, but also other spatial cues in the recording and it's EQ, contribute to a perception of soundstage.

Simple, but it works for me, repetitively.

Why should it be logical to rule out the idea that fidelity to the spatial nature of a recording, in addition to artificial sound wave manipulations, could enhance soundstage? I have lots of evidence in my own careful listening; that's subjective, even if carefully done, and thus only a datapoint. But why rule out the idea a priori?

PS Dan Clark does repeatedly seem to state his belief that a U-shaped upper bass FR contributes to soundstage in Youtube commentary. This is in addition to the punch/slam thing in the upper bass. I wonder if the two ideas simply sound like overlapping assertions in the context of live interviews as they evolve from subject to subject. Meanwhile I don't know why to agree or disagree on the soundstage side of upper bass emphasis; my own A/B testing included a more broad FR adjustment (alternately above and slightly below Harman) well up into the mids (~250 Hz to >2 kHz). I also A/B'd hedaphones with weak out-of-the-box soundstage to ones with a better soundstage and a more U-shaped out-of-the-box FR ... Closed X vs Noire. I don't have the Expanse or Stealth to include.
 
Last edited:

majingotan

Major Contributor
Forum Donor
Joined
Feb 13, 2018
Messages
1,508
Likes
1,779
Location
Laguna, Philippines
Hello All,

This may be of interest to some.

The gang of PhD's over a Facebook Reality Labs have recently written a AES published paper on this topic.

The Harman Target transfer function may be preferred for "music content" and a flat transfer function may be preferred for "spatial content".

Personally I feel that this concept needs more investigation. Music can be pretty spatial.


Thanks DT

Awesome read! I admit I was skimming through it and went for the results

It seems that "individual flat" EQ targets the best spatial perception out of all targets. Hence why I choose IEMs that have those 3D spatial imaging capabilities. What's interesting between these two IEMs that give that around the head spatial imaging to my perception is the FR curve trend from 4.5-8 KHz is quite similar

graph.png
 

JanesJr1

Addicted to Fun and Learning
Joined
Jun 9, 2021
Messages
504
Likes
449
Location
MA
Awesome read! I admit I was skimming through it and went for the results

It seems that "individual flat" EQ targets the best spatial perception out of all targets. Hence why I choose IEMs that have those 3D spatial imaging capabilities. What's interesting between these two IEMs that give that around the head spatial imaging to my perception is the FR curve trend from 4.5-8 KHz is quite similar

View attachment 236395
To be clear, are you suggesting that the "leveling" of the Harman peak from 2.6 to 8 kHz by both IEM's might enhance spatial cues by being a flat response compared to Harman?

If so, a quick check of the FR's of three headphones that get spatial kudos (Senn 800s, AKG 371, DCA Expanse) doesn't really show that pattern, if in fact it should. It makes me wonder if we tallied the FR's for all the generally-recognized headphones that have spatial chops*, would we find flat FR curves compared to Harman, as the published paper suggests we might?

*The Senn and DCA headphones are mentioned on ASR frequently as spatial champs. The AKG is a Sean Olive fave for spatial presentation.
 
Last edited:

majingotan

Major Contributor
Forum Donor
Joined
Feb 13, 2018
Messages
1,508
Likes
1,779
Location
Laguna, Philippines
To be clear, are you suggesting that the "leveling" of the Harman peak from 2.6 to 8 kHz by both IEM's might enhance spatial cues by being a flat response compared to Harman?

If so, a quick check of the FR's of three headphones that get spatial kudos (Senn 800s, AKG 371, DCA Expanse) doesn't really show that pattern, if in fact it should. It makes me wonder if we tallied the FR's for all the generally-recognized headphones that have spatial chops*, would we find flat FR curves compared to Harman, as the published paper suggests we might?

*The Senn and DCA headphones are mentioned on ASR frequently as spatial champs. The AKG is a Sean Olive fave for spatial presentation.

It could be that IEM FR does not translate to over the ear headphone FR in this case. Even the Dan Clark Aeon X Open has good spatial qualities that’s more stretched out to the sides than the two IEMs mentioned, Expanse and Stealth.

If I were to put the spatial shape. The two IEMs and Expanse track a more spherical spatial presentation while Stealth is more like the Earth (slightly obtuse) while the Aeon X Open is more oblong than the rest. i.e. the sensation of panning more to the left and right rather than panning less yet having a sensation that’s it’s coming from the front or back, and up and down. The up and down, front to back perception is a lot less present on the Aeons compared to the rest of them
 
Top Bottom