Frontal reflections, depth of soundstage, and dipoles

[Bugal1998]

As @tmuikku has suggested in various posts/threads I spent this afternoon listening to musical selections and test tracks at different positions around the room. Room is 19'2"W x 24'D x 8.5'H and purpose built/professionally designed. Front baffle of the speakers are 3'6" from the front wall and center of the drivers are 4'6" from the side walls. Here is what I heard...

TL;DR--

• Was surprised to discover that volume and reverb related depth cues, while subtle experiences in general, were persistently perceptible at every location tested in the room
• From anywhere behind the speakers and at the rear wall of the room, the experience is "watching from outside"
• In/out of phase differences were audible in all locations tested EXCEPT for the front corners behind the speakers were I had zero ability to hear differences in phase.
• Left and right cues were audible behind the speakers and located between the L/R speakers, but left and right essentially disappeared, or even reversed, when listening in the corners behind the speakers.
• When standing near the front of the room depth was extremely compressed, as I moved further and further back in the room, the sense of depth stretched to fill the room, giving a maximal perception of depth at the rear wall
• There is a subtle transition from "watching from the outside" to "you are there" as the distance from the speakers increases; at the rear wall the sensation of "watching from the outside" returns (perhaps a lack of late reflections from behind?).
• Was pleased to discover that both focus and envelopment were essentially maximized at the MLP (not a surprise considering it is a purpose built and engineered room).

Boring details of each position below.

In front of the speakers...
Moving from the plane 3' in front of the speakers to the back wall of the room; position remained centered between the L/R speakers. Note: I did not adjust toe-in or the listening triangle for these tests which likely alters some perceptions of focus and width.

At all positions in front of the speakers:

• Left and right cues are extremely obvious
• Depth cues are clearly perceived
• In-phase and out-of-phase sounds are extremely obvious.

3' from the plane of the speakers:

• The experience is in between "you are there" and "watching from the outside" with reverb and sounds out to ~70 degrees left and right (probably not far from the position of the L/R speakers)
• Depth cues are compressed, as-if the entire playback happens in a space only 5' deep (even with eyes closed trying to imagine more depth)
  • Range of movement during microphone configuration sample test tracks involving front to back movement is minimal
• Stereo-mono pink noise is perceived to be 2-3' across
  • Out of phase pink noise is inside the head

6' from the plane of the speakers:

• The experience is in between "you are there" and "watching from the outside" with reverb and sounds out to ~70 degrees left and right
• Depth cues are compressed, as-if the entire playback happens in a space only 6-8' deep (even with eyes closed trying to imagine more depth)
  • Range of movement during microphone configuration sample test track involving front to back movement is still small
• Stereo-mono pink noise is perceived to be ~2' across
  • Out of phase pink noise is outside the head but only as wide as the L/R speakers and positioned in front of the listener

9' from the plane of the speakers:

• The experience is in between has just barely become "you are there" with reverb and sounds out to ~80-90 degrees left and right
• Depth cues are larger, as-if the entire playback happens in a space 15'-20' deep
  • Range of movement during microphone configuration sample test track involving front to back movement is still getting larger
• Stereo-mono pink noise is perceived to be ~2' across
  • Out of phase pink noise is outside the head and wider than the speakers but in the same plane as or just in front of the listener
  • Other stereo sounds now come from outside the speakers as well

At the MLP (~11.5' from the plane speakers):

• The experience is solidly 'you are there' with reverb and sounds out to 80-110 degrees left and right
• Depth cues are clearly perceived
• Range of movement during microphone configuration sample test track involving front to back movement is large
• Stereo-mono pink noise is perceived to be between the size of a basketball and a beachball; real world mono music is a bit smaller (perhaps due to the limited bandwidth of the selection). This is the point of maximum focus without adjusting the listening triangle or toe-in.
  • Note: If speaker toe-in is increased the sound is reduced to the size of a large softball.
  • Out of phase pink noise is the width of the room and experienced to my left and right and even slightly behind

15' from the plane of the speakers (3' Behind the MLP):

• The experience is solidly 'you are there' with reverb and sounds out to 80-110 degrees left and right
• Depth cues are clearly perceived
• Range of movement during microphone configuration sample test track involving front to back movement is very large
• Stereo-mono pink noise is perceived to be ~6' across (focus has rapidly declined);
  • Out of phase pink noise is the width of the room and experienced to my left and right and even slightly behind

20' (at the rear wall):

• The experience returns to slightly "watching from the outside" with reverb and sounds out to ~80 degrees left and right
• Depth cues are clearly perceived
• Range of movement during microphone configuration sample test track involving front to back movement is massive and perception of depth is the whole length of the room (and then some)
• Stereo-mono pink noise is perceived to be ~10' across (fills the entire space between the L/R speakers);
  • Out of phase pink noise is the width of the room but slightly in front of me

Behind the speakers...
At no time while listening behind the speakers was I ever able to localize sound to a speaker, and the experience was always "watching from the outside" (i.e. no envelopment).

Anywhere behind the plane of the speakers but between the L/R the speakers:

• The room is perceived to be completely filled with sound
• Left and right cues are audible but far less pronounced
• Depth cues based on volume and reverb are still perceived, but now the deepest sounds are at the back of the room (depth is reversed)
• Range of movement during microphone configuration sample test track involving front to back movement is somewhat ambiguous but proceeds from the back-ish area of the room up to nearly the plane of the speakers (give or take a foot)
• In-phase and out-of-phase sounds are still clearly identifiable, but less pronounced
• Stereo-mono pink noise is perceived to be ~8-10' in diameter
• Out of phase pink noise is the width of the room ('19 and mostly hollow in the middle)
• When moving left and right between the speakers the central image followed my movements with no discernable change in relative positioning or timbre
• First time I've ever experienced a sweet-spot that was wider than a single seat (or more correctly, an absolute absence of a sweet spot?)

In the corners behind the speakers:

The experience was the same in both corners so I'll share the experience from behind the right speaker.

• The perception of sound is mostly limited to the front third of the room and closer to far (left) speaker
• Left and right cues are--for any practical purposes--inaudible
• Depth cues are still perceived, but now the deepest sounds are at the left wall about a third of the way from the front of the room
• Interestingly, while standing in the corner behind the right speaker, when a sound was moving deeper and deeper behind the right speaker (when heard from the MLP), I perceived it is moving away from me and into the 1/3 point of the left wall
• Range of movement during microphone configuration sample test track involving front to back movement is still clearly audible but also somewhat ambiguous and limited. What movement can be perceived moves from the left wall up to maybe 5' in-front of me at the closest
• In-phase and out-of-phase sounds are entirely unidentifiable
  • Stereo-mono pink noise is perceived to be ~6' in diameter; real world mono music is a bit smaller (perhaps due to the limited bandwidth of the selection).
  • Out of phase pink noise is identical to in-phase; no change heard

Edit: Typos

 

[tmuikku]

Hi,
nice experimenting!

Calculated from dimensions speakers are about 10' apart so MLP at ~11.5' from base plane makes speakers about at 21deg angle from the median plane, which is narrower than equilateral triangle, right? Is toe-in angle quite small, no toe-in or perhaps speakers point toward the MLP? Your speakers were the M2? Room is purpose built, as described earlier in the thread, did you have photos somewhere?

This is interesting and nice data, thanks taking time posting it! I bet you are now more familiar with your system yourself, but also us others have a slight understanding how it sounds like in there, if we do similar tests in our own environments and compare.

While not mentioned in the report did you do walk along the centerline equidistant to speakers? eyes closed listening clarity and size of the phantom center "the focus", no sudden changes but gradual?

A quick imagination trip to your space based on your descriptions to what I perceive in mine, I could think the acoustic treatment are likely concentrated to deal early reflections for MLP, for current positioning. Assuming small toe-in and wide radiation of the M2 and "lively acoustic treatment" I think early reflections get relatively high in level if you move from MLP closer to speakers losing the focus again, just like moving farther from the MLP and speakers. This would be because the acoustic treatment is local to the MLP, you'd need to move both speakers and listener if you wanted to increase or decrease listening distance, to maintain the early reflections hitting the acoustic treatment. If these assumptions and imagination are in the ballpark, you should be able to maintain the "focus" closer to speakers also by increasing toe-in. But, this is not useful for you, just a listening test I assume should happen.

After all, main purpose for the whole exercise is to find positioning so that sound at MLP is what one wants to hear, if you have it nothign more needs to be done. Now that you scouted your space, you likely found out it's true, sound you like best is at the MLP.

Usually, on a normal living room and in my experience, the good sound doesn't extend that far, far enough into the room so that it was at practical listening position to a sofa. 3 meters / 10' you have there would be good enough I think, but doesn't happen at least in the normal living rooms I've had my speakers at, and your report suggests acoustic treatment is needed for such long distance good sound. I have good sound, the focus you describe, but it extends only maybe 5' from speaker plane, which is about 7' from speaker to ear as I have them bit wider apart than equilateral triangle, about 10' apart like you. I still have the focus close to speakers unlike you describe, but I've got toe-in and directional speakers without acoustic treatments. Reading your report makes sense to me the way I've expressed here.

 

[Bugal1998]

Hi,
nice experimenting!

Calculated from dimensions speakers are about 10' apart so MLP at ~11.5' from base plane makes speakers about at 21deg angle from the median plane, which is narrower than equilateral triangle, right? Is toe-in angle quite small, no toe-in or perhaps speakers point toward the MLP? Your speakers were the M2? Room is purpose built, as described earlier in the thread, did you have photos somewhere?

Thank you. I'm not a fountain of theoretical knowledge like many folks here; but I can do simple listening tests for my own understanding, and if anyone finds the observations to be of interest, then it's worth sharing.

Yes, the speakers are narrower than equilateral, as designed. Speakers are 10'2" apart CtC. The listening angle was designed to be 50deg combined left and right, and they are positioned as such. Due to variations in my final chair selection/seating position I'm coming up with 24.2deg each L/R or 48.4deg total when measuring from my ears to the speakers. I was skeptical of the narrow location during the design phase, so I tried wider speaker positions (55/60/65deg listening angles) and this is clearly--to my surprise--the best location to my ears. Toe-in is ~10deg per speaker; I've tried numerous toe angles and this is offers the best compromise of frequency response, focus, and envelopment to my ears. Pointed directly at the listening position was superior in some ways and inferior in others, the current toe angle is purely a subjective preference.

Yes, they are M2s, and best pics of the treatment in the space are here. We did have a similar discussion in an earlier thread, but I thought the focus on depth here was interesting.

This is interesting and nice data, thanks taking time posting it! I bet you are now more familiar with your system yourself, but also us others have a slight understanding how it sounds like in there, if we do similar tests in our own environments and compare.

While not mentioned in the report did you do walk along the centerline equidistant to speakers? eyes closed listening clarity and size of the phantom center "the focus", no sudden changes but gradual?

Yes, I (sort of) walked the space with eyes closed, and I've done that on other occasions as well; there is seating that is not easily moved, as well as a level change to the floor, so I can't truly walk the centerline uninterrupted with my eyes closed. As shared in the other thread, if I stand right behind the MLP, the focus has decreased significantly, and I think it happens about 6 inches behind my head at the MLP, but hard to tell with the chair in the way.

A quick imagination trip to your space based on your descriptions to what I perceive in mine, I could think the acoustic treatment are likely concentrated to deal early reflections for MLP, for current positioning. Assuming small toe-in and wide radiation of the M2 and "lively acoustic treatment" I think early reflections get relatively high in level if you move from MLP closer to speakers losing the focus again, just like moving farther from the MLP and speakers. This would be because the acoustic treatment is local to the MLP, you'd need to move both speakers and listener if you wanted to increase or decrease listening distance, to maintain the early reflections hitting the acoustic treatment. If these assumptions and imagination are in the ballpark, you should be able to maintain the "focus" closer to speakers also by increasing toe-in. But, this is not useful for you, just a listening test I assume should happen.

Yes, design was to manage (but not eliminate) early reflections for the MLP, and leave later reflections either untreated or diffused. And yes, I suspect part of the perception of increasing focus is due to moving into a treated zone as I approach the MLP, yielding more focus even though I'm moving further from the speakers. The speakers are on turntables so adjusting toe-in is very easy and likely would have increased the focus closer to the speakers during the test; but it would have greatly increased the time required so I skipped that step.

After all, main purpose for the whole exercise is to find positioning so that sound at MLP is what one wants to hear, if you have it nothign more needs to be done. Now that you scouted your space, you likely found out it's true, sound you like best is at the MLP.

Yes, and I had verified in the past as well. I was mostly curious, per comments by @goat76 and others about 1) the direct sound not providing envelopment (which my prior experiences and this test seems to have reinforced), and the diffuse field having no phase information (which my listening tests may or may not have contradicted, at least in a small room), and to see to what extent depth, phase, and directional cues persist, especially outside of the direct sound of the speakers. We can only simulate a diffuse field in a small room, but I was still surprised by the extent to which all cues could be heard while standing behind the speakers.

Usually, on a normal living room and in my experience, the good sound doesn't extend that far, far enough into the room so that it was at practical listening position to a sofa. 3 meters / 10' you have there would be good enough I think, but doesn't happen at least in the normal living rooms I've had my speakers at, and your report suggests acoustic treatment is needed for such long distance good sound. I have good sound, the focus you describe, but it extends only maybe 5' from speaker plane, which is about 7' from speaker to ear as I have them bit wider apart than equilateral triangle, about 10' apart like you. I still have the focus close to speakers unlike you describe, but I've got toe-in and directional speakers without acoustic treatments. Reading your report makes sense to me the way I've expressed here.

In my living room at 12' there is no focus or envelopment to speak of. Obviously the longer listening distance in the dedicated space is enabled by treatment design. Focus was clearly sharper with direct toe-in, and sharper still (pink noise the size of a baseball of softball) in a highly damped space with the same speakers. I find the subtle reduction in focus and hyper clarity makes listening more enjoyable, and the corresponding ASW and envelopment is just addictive.

 

[tmuikku]

I'm terribly sorry I did not remember the ealier discussion! I looked up your message history to find the speakers and the room, but somehow missed the listening room images and failed to make a connection. Thanks again for your report, it likely sounds really nice in there!

Depth is the topic so lets focus on that, whats the microphone test track with depth cues you were using?

 

[goat76]

Yes, the speakers are narrower than equilateral, as designed. Speakers are 10'2" apart CtC. The listening angle was designed to be 50deg left and right, and they are positioned as such. Due to variations in my final chair selection/seating position I'm coming up with 24.2deg L/R or 48.4deg total when measuring from my ears to the speakers. I was skeptical of the narrow location during the design phase, so I tried wider speaker positions (55/60/65deg listening angles) and this is clearly--to my surprise--the best location to my ears. Toe-in is ~10deg per speaker; I've tried numerous toe angles and this is offers the best compromise of frequency response, focus, and envelopment to my ears. Pointed directly at the listening position was superior in some ways and inferior in others, the current toe angle is purely a subjective preference.

Yes, they are M2s, and best pics of the treatment in the space are here. We did have a similar discussion in an earlier thread, but I thought the focus on depth here was interesting.

What a nice room you have! I'm jealous, as I have to make compromises. My listening room is just an ordinary living room with neighbors above, below, and on the other side of the speaker wall.

But still, with fairly small floor-standing loudspeakers (ATC SCM40) and two subwoofers (in stereo configuration) about 1 meter out from the wall, 2.03 meters apart, and the listening position just inside the equilateral triangle at 2.01 meters from each speaker, I have found the sweet spot for hearing "into" the recordings thanks to a rather high ratio of direct sound. The room (5.08m wide, 4.28m length, 2.6m height) has a moderate level of acoustic treatment with a few absorption panels. Still, the fairly short listening triangle is probably the main reason for the high ratio of direct sound.

The main thing I focus on when setting up the speakers and the listening positioning is to get the phantom center to sound (almost) as distinct as if a real center speaker is in place. The phantom center is like the "anchor" for the stereo image and when you get that distinct-sounding center image, everything else falls into place like finer details such as depth and width in the recording.

In other words, the two speakers must work in tandem to create a unified stereo image, and the easiest way of knowing when that is achieved is when the phantom-centered objects sound equally as distinct and equally as wide as when the same objects are hard-panned to one of the loudspeakers.
I have created a sound file containing a snare drum, a kick drum, and a female voice which are panned Left, 50% Left, Center, 50% Right, and Right. They should sound about the same in width and distinctiveness wherever they are panned in the stereo field.

Dropbox

www.dropbox.com

Yes, I (sort of) walked the space with eyes closed, and I've done that on other occasions as well; there is seating that is not easily moved, as well as a level change to the floor, so I can't truly walk the centerline uninterrupted with my eyes closed. As shared in the other thread, if I stand right behind the MLP, the focus has decreased significantly, and I think it happens about 6 inches behind my head at the MLP, but hard to tell with the chair in the way.

You can use a wheeled office chair when you are checking the sound in the centerline, that way you keep your ears at the level of your loudspeakers' acoustic center vertically, but don't fall off the cliff while doing so.

 

[Bugal1998]

I'm terribly sorry I did not remember the ealier discussion! I looked up your message history to find the speakers and the room, but somehow missed the listening room images and failed to make a connection. Thanks again for your report, it likely sounds really nice in there!

No apologies needed! I (almost) feel self-conscious engaging on this topic so much, but hearing the transformation brought about by room differences has been one of the more fascinating discoveries on my journey. How the room, binaural hearing, and cognition interacts to create such a beautiful [IMO] illusion is a fascinating topic.

Depth is the topic so lets focus on that, whats the microphone test track with depth cues you were using?

CD: Stereophile Test CD, Vol. 3.
Track: Soundstage Maps & Microphone Techniques

The first microphone configuration, and the last two are the configurations I listen to the most. The first configuration has a distinct "U" shape for depth starting deep in the front left corner of the room, moving forward to about the left speaker, across the front of the room (not a smooth pan due to microphone configuration) over to the right speaker, then receding deep into the front right corner of the room. With eyes closed you can pretty well convince yourself the person is 40-50' away in the corners, but it's a delicate illusion.

On the the last two configurations I listen to the segments starting at the back of the church, walking past the microphone, and up to the front of the church. The banging starts deep in the front of the room, well beyond the front wall, moves out into the room to where it feels like you can reach up and touch the cowbell (whatever he's banging on) as it passes the microphone, then recedes back to plane of the speakers or slightly beyond.

The ambiance/reverb throughout the track also serves as a pretty good reference for "watching from outside" verses "you are there" based on how enveloping the reverb becomes. Though even in the direct dominate sound field I find the reverb spacious and enjoyable.

Types of Depth

There are also different types of depth that I don't recall us calling out, and I'm not sure what the OP had in mind, but it seems to me there is depth created by:

1. Volume differences
2. Frequency response differences
3. Naturally recorded timing differences between sources and venue reflections captured in the recording
4. Reverb or ambiance
5. Phase and timing manipulations of originally mono tracked sounds (Qsound and other processing techniques)
6. And no doubt others I'm not aware of...

I'm primarily discussing types 1-4 which usually create depth behind the plane of the speakers (the stereophile track above being an exception with sound well out into the room at one point). While type 4 creates a sense of depth behind the speakers, type 4 can also project reverb out into the room which I lump in with envelopment, or depth projecting into the room. <--Not saying this the correct way to speak of these experiences, it's just context for my comments.

Depth types 1-4 are all clearly in the recording, but they seem to be accentuated by late reflections and the corresponding envelopment.

Type 5--while enjoyable--can also be perhaps a bit gimmicky. Type 5 also tends to be somewhat less room dependent, again, in my experience.

 

[Emlin]

No apologies needed! I (almost) feel self-conscious engaging on this topic so much, but hearing the transformation brought about by room differences has been one of the more fascinating discoveries on my journey. How the room, binaural hearing, and cognition interacts to create such a beautiful [IMO] illusion is a fascinating topic.



CD: Stereophile Test CD, Vol. 3.
Track: Soundstage Maps & Microphone Techniques

The first microphone configuration, and the last two are the configurations I listen to the most. The first configuration has a distinct "U" shape for depth starting deep in the front left corner of the room, moving forward to about the left speaker, across the front of the room (not a smooth pan due to microphone configuration) over to the right speaker, then receding deep into the front right corner of the room. With eyes closed you can pretty well convince yourself the person is 40-50' away in the corners, but it's a delicate illusion.

On the the last two configurations I listen to the segments starting at the back of the church, walking past the microphone, and up to the front of the church. The banging starts deep in the front of the room, well beyond the front wall, moves out into the room to where it feels like you can reach up and touch the cowbell (whatever he's banging on) as it passes the microphone, then recedes back to plane of the speakers or slightly beyond.

The ambiance/reverb throughout the track also serves as a pretty good reference for "watching from outside" verses "you are there" based on how enveloping the reverb becomes. Though even in the direct dominate sound field I find the reverb spacious and enjoyable.

Types of Depth

There are also different types of depth that I don't recall us calling out, and I'm not sure what the OP had in mind, but it seems to me there is depth created by:

1. Volume differences
2. Frequency response differences
3. Naturally recorded timing differences between sources and venue reflections captured in the recording
4. Reverb or ambiance
5. Phase and timing manipulations of originally mono tracked sounds (Qsound and other processing techniques)
6. And no doubt others I'm not aware of...

I'm primarily discussing types 1-4 which usually create depth behind the plane of the speakers (the stereophile track above being an exception with sound well out into the room at one point). While type 4 creates a sense of depth behind the speakers, type 4 can also project reverb out into the room which I lump in with envelopment, or depth projecting into the room. <--Not saying this the correct way to speak of these experiences, it's just context for my comments.

Depth types 1-4 are all clearly in the recording, but they seem to be accentuated by late reflections and the corresponding envelopment.

Type 5--while enjoyable--can also be perhaps a bit gimmicky. Type 5 also tends to be somewhat less room dependent, again, in my experience.

I think that 3 and 4 are the most important, and, if I have understood properly, the same thing. However they are produced.

 

[Bugal1998]

I think that 3 and 4 are the most important, and, if I have understood properly, the same thing. However they are produced.

Agreed that they are ultimately produced the same way (also true for 1 and 2 in live, non-close mic'd recording). The difference in my head being that 4 is consciously audible as a reverb or reflection, and 3 (and to a large extend 1 and 2) is only experienced primarily by the way it changes the perceived location of the source sound. Examples could include inaudible floor, wall, and ceiling reflections that impart frequency response and timing cues but aren't heard as distinct audible reflections or reverb.

Again, not in anyway saying this the correct way to speak of these experiences, it's just context for my comments.

 

[Emlin]

Agreed that they are ultimately produced the same way (also true for 1 and 2 in live, non-close mic'd recording). The difference in my head being that 4 is consciously audible as a reverb or reflection, and 3 (and to a large extend 1 and 2) is only experienced primarily by the way it changes the perceived location of the source sound. Examples could include inaudible floor, wall, and ceiling reflections that impart frequency response and timing cues but aren't heard as distinct audible reflections or reverb.

Again, not in anyway saying this the correct way to speak of these experiences, it's just context for my comments.

I can't agree that inaudible things of any kind can make an audible difference. But maybe something's been lost in translation...

Just because you can't isolate an audible sound consciously, it doesn't mean that your brain is ignoring it. This is especially true when it comes to spacial cues.

 

[Bugal1998]

I can't agree that inaudible things of any kind can make an audible difference. But maybe something's been lost in translation...

There's nuance here, which is why I said, "only experienced primarily by the way it changes the perceived location of the source sound."

Case and point, thousands and thousands of times everyday sound sources translate left or right, or up and down--literally every time we move our head--and there is a frequency shift due to the influence of the head and pinna; unless we actively listen for the frequency shift, we don't consciously hear it. We just don't. What we hear is the source of the sound changing location. We CAN hear the frequency change, but 99.99% of the time we DON'T, we simply experience it as a spatial change. There are numerous other examples of spatial cues when sitting, standing, entering or leaving a room, etc. We process the frequency response changes and reflections as spatial cues about the environment, but we don't experience them as reflections and frequency response changes consciously, again, unless we actively listen.

Hopefully that clarifies.

 

[Bugal1998]

I can't agree that inaudible things of any kind can make an audible difference. But maybe something's been lost in translation...

Just because you can't isolate an audible sound consciously, it doesn't mean that your brain is ignoring it. This is especially true when it comes to spacial cues.

I think we cross posted while you were making your edit; but yes, that was exactly what I was implying and we're in alignment here.

Do I hear an explicit reflection/echo/reverb (type 4), or do I primarily experience a spatial perception (type 3).

 

[Bugal1998]

What a nice room you have! I'm jealous, as I have to make compromises. My listening room is just an ordinary living room with neighbors above, below, and on the other side of the speaker wall.

Thank you! I never imagined a dedicated space when I set out on the journey 15-years ago! I had neighbors on the other side of walls for years, and it's challenging.

But still, with fairly small floor-standing loudspeakers (ATC SCM40) and two subwoofers (in stereo configuration) about 1 meter out from the wall, 2.03 meters apart, and the listening position just inside the equilateral triangle at 2.01 meters from each speaker, I have found the sweet spot for hearing "into" the recordings thanks to a rather high ratio of direct sound. The room (5.08m wide, 4.28m length, 2.6m height) has a moderate level of acoustic treatment with a few absorption panels. Still, the fairly short listening triangle is probably the main reason for the high ratio of direct sound.

I'm sure your system sounds great. I've experienced wonderful depth and envelopment in a friend's completely untreated living room; with a thoughtful approach, a bit of knowledge, and little setup creativity, exceptional sound can be achieved almost anywhere.

The main thing I focus on when setting up the speakers and the listening positioning is to get the phantom center to sound (almost) as distinct as if a real center speaker is in place. The phantom center is like the "anchor" for the stereo image and when you get that distinct-sounding center image, everything else falls into place like finer details such as depth and width in the recording.

Agreed, if the phantom image isn't right, the foundation is missing. Casual living room setups aside, I've always had a rock solid phantom image. I know there's research that says we hear the angle of the left and right speakers, so a 2-channel phantom image is a significant compromise, but for whatever reason I just can't hear the difference at the MLP. The exception being my movie surround configuration where the L/R speakers are wider than 60deg; in that case the center fill isn't great and the center speaker makes a nice difference.

In other words, the two speakers must work in tandem to create a unified stereo image, and the easiest way of knowing when that is achieved is when the phantom-centered objects sound equally as distinct and equally as wide as when the same objects are hard-panned to one of the loudspeakers.
I have created a sound file containing a snare drum, a kick drum, and a female voice which are panned Left, 50% Left, Center, 50% Right, and Right. They should sound about the same in width and distinctiveness wherever they are panned in the stereo field.

Dropbox

www.dropbox.com

I'll test my setup with your file, thank you for sharing! So far I've used commercial content or panning pink noise to setup the system. I used to emphasize a maximally focused and precise phantom image, but now I prioritize envelopment, spaciousness, and depth since it contributes more to my overall enjoyment. I find that opening up the 'ball of pink noise' to the size of a beachball (instead of a softball) produces a more compelling sense of scale for vocalists and instruments while also increasing the 'holographic' sense of space. The placement of sounds is softened slightly, but it in no way impedes the ability to precisely localize where sounds are positioned across the front stage. In fact, I find it makes the presentation more organic and natural.

You can use a wheeled office chair when you are checking the sound in the centerline, that way you keep your ears at the level of your loudspeakers' acoustic center vertically, but don't fall off the cliff while doing so.

That's a good idea! Haha fortunately the cliff goes up, so no risk of falling if I roll backwards from the speakers. Unfortunately I wouldn't be able to roll past the MLP due to the rear of the room being raised right behind the MLP... and that's where the center image starts to become more diffuse. And now the row of five seats are in the way and connected to each, making them a real hassle to move.

 

[5-pot-fan]

What a nice room you have! I'm jealous, as I have to make compromises. My listening room is just an ordinary living room with neighbors above, below, and on the other side of the speaker wall.

But still, with fairly small floor-standing loudspeakers (ATC SCM40) and two subwoofers (in stereo configuration) about 1 meter out from the wall, 2.03 meters apart, and the listening position just inside the equilateral triangle at 2.01 meters from each speaker, I have found the sweet spot for hearing "into" the recordings thanks to a rather high ratio of direct sound. The room (5.08m wide, 4.28m length, 2.6m height) has a moderate level of acoustic treatment with a few absorption panels. Still, the fairly short listening triangle is probably the main reason for the high ratio of direct sound.

The main thing I focus on when setting up the speakers and the listening positioning is to get the phantom center to sound (almost) as distinct as if a real center speaker is in place. The phantom center is like the "anchor" for the stereo image and when you get that distinct-sounding center image, everything else falls into place like finer details such as depth and width in the recording.

In other words, the two speakers must work in tandem to create a unified stereo image, and the easiest way of knowing when that is achieved is when the phantom-centered objects sound equally as distinct and equally as wide as when the same objects are hard-panned to one of the loudspeakers.
I have created a sound file containing a snare drum, a kick drum, and a female voice which are panned Left, 50% Left, Center, 50% Right, and Right. They should sound about the same in width and distinctiveness wherever they are panned in the stereo field.

Dropbox

www.dropbox.com

You can use a wheeled office chair when you are checking the sound in the centerline, that way you keep your ears at the level of your loudspeakers' acoustic center vertically, but don't fall off the cliff while doing so.

Thank you for your thoughtful contributions, and also for the soundfile.

 

[audiofooled]

In other words, the two speakers must work in tandem to create a unified stereo image

Sorry for being late to the discussion but I couldn't agree more with what you said here.

I also agree with what @Duke said earlier in the thread, that soundstage depth has to do with getting the spatial characteristics of the recording to dominate over the spatial characteristics of the playback room, which in turn can mean suppressing or manipulating the early in-room reflections.

I highly recommend this presentation from J_J, where he talks about distance perception on multiple occasion, that is if you have an hour or so to watch the whole presentation:

So it seems that we can tell the distance from primary cues, namely ITD and ILD, also elevation when we take into account reflections from our shoulders. But first and foremost, the brain works on loudness. Not SPL, but perceived loudness. What is also important is that the ears are non linear, everything is frequency dependent and for low frequencies we get waveform lock, we can also detect phase and in doing so we can tell time.
And then there's auditory compression mechanism which effects are profound in detecting how loud things are and for perceived dynamic range.

So, in room with a stereo setup we may find out that reflections (some of which would be frontal) will have a role in affecting the performance of the loudspeakers. FR would be wrong as compared to the anechoic response, just how wrong would be highly dependent on loudspeaker radiation pattern, setup and room acoustics.

In the hopes of getting a result of a unified stereo field, IMHO it is very important that left and right channel are to some extent equally wrong about FR and that they are time and phase aligned. And then there's low frequency extension which is more challenging the more extension you would want. Because of auditory system's abilities when it comes to low frequencies, they can provide yet another distance cue with regards to the rest of the spectrum. How you deal with room modes and stored energy is very important because it can make or brake the cues in the recording. By that I mean masking them.

At the risk of mixing up seemingly entirely different concepts, here's how things work with my setup:

- From the ground up, this is the raw magnitude response for low frequencies:

This is stereo bass setup, L+sub in purple, R+sub in pink, vector average of the two in red and finally, FR summation of L+R+sub in cyan.

- Trace arithmetic for the FR and phase differences in between the channels (VAR smoothing):

As you can see, the transition region is messed up both in magnitude and phase. It turns out that it's because of the frontal reflections, the coffee table would be the main culprit. More about that later.

- This is showing how many dB's of summation in between the stereo channels I get across the audible range, in comparison to just the vector average of them:

So, for the phantom center, time/phase alignment implies that I get mostly full 6 dB of summation across the board, exception being high frequencies due to the listening distance of about 10 feet and the air absorption.

Now, the reflections. Both horizontal and vertical directivity control makes it easier to keep things in check:

This shows what happens in time after the initial impulse. Left (purple) and Right (pink), occupying their own acoustic space (which is asymmetrical BTW), are showing measurable spectral differences, but still of questionable audibility, especially in stereo (cyan). Practically, listening tests confirm that, when either left or right channel is on, there are some clues from the reflections which make it possible to discern where each of the loudspeakers is located (but none of the actual drivers). It sounds like some kind of halo around them which Amir often describes when reviewing some of the larger speakers.

But, when both channels are playing, any chance of localizing the actual loudspeakers goes down the drain. And this is true for any distance and any location in the room I sit or stand.

Standard view of the impulse responses:

Now for the frontal reflections from coffee table. They are inaudible most of the time, because it's being positioned symmetrically with regards to the speakers. The case in favor of audibility, it's responsible for some mild tonality changes only when I move my head closer or further away from the MLP, but no change in localization or timbre at all when I'm rotating my head for listening to dynamic changes in ITD and ILD.

What and when table does can be seen here:

The same, but not being normalized to peaks:

Some of the stuff come from a TV, some from the table. My point is, with a good polar response and setup, frontal reflections, at least in my case are not even remotely detrimental to the sound quality or image localization and perception of depth. When I remove the table, the only difference is that there are no perceptible spectral changes when moving front to back. So I leave it be, the environment being just a normal living room.

 

[goat76]

In the hopes of getting a result of a unified stereo field, IMHO it is very important that left and right channel are to some extent equally wrong about FR and that they are time and phase aligned. And then there's low frequency extension which is more challenging the more extension you would want. Because of auditory system's abilities when it comes to low frequencies, they can provide yet another distance cue with regards to the rest of the spectrum. How you deal with room modes and stored energy is very important because it can make or brake the cues in the recording. By that I mean masking them.

These are main points as I see it to maximize our hearing “into the recording”:

• Start with getting a pair of loudspeakers with a really good matching response as possible. This will make sure that the finer details in the recordings will not get lost when the two loudspeakers reproduce the unified stereo image.

• The positioning and distance between the two loudspeakers should never create “a hole in the middle” of the stereo field, so it's very important to find the correct spacing between the loudspeakers that works well with the particular loudspeakers of choice which is often dependent on size and dispersion characteristics.

• The single most important aspect to get right is to get the phantom centered sounds as distinct sounding as if there was physical center speaker in place. This will act as an “anchor” to everything in the music mix right down to the finer details putting together the whole puzzle, which are extra important for hearing the room information of the recorded venue.

• I have found out that I prefer to have my listening position just a few centimeters inside of an equilateral triangle, which is also a fairly small one of just about 2 meters. This makes it possible to have both the loudspeakers and listening position far off the boundaries in my room which reduces the early reflections down to -19 dB and lower, also thanks to some absorption panels in the room.

• As I believe in David Griesinger’s theories about how to create the sensation of envelopment, I have my two subwoofers set up in a stereo configuration to create some amount of phase differences which I think helps with that. I also think I’m lucky as the asymmetrical positioning of my loudspeakers and subwoofers helps creates enough differences for a lot of phase differences in the bass region. This has of course very little to do with the recorded information and is what I believe just a diffuse field of sound as I previously named as “a mess of bouncing reflections” in the listening room.



In short: A high ratio of direct sound vs reflected sounds and properly positioned loudspeakers will give a pure insightingo the recorded venue and the recorded reverb, and some late reflections from the listening environment will give the sensation of development even if that diffuse sound field probably doesn't bear much resemble of what is actually on the recordings.

I think that pretty much sums up most of the things I think is important for a convincing reproduction of the sound of a stereo recording. Maybe I missed something.


P.S. I believe I have seen that video with JJ but it was some time ago. Maybe I refresh my memory and watch it again when I have some spare time.
Thank you.

 

[dualazmak]

In short: A high ratio of direct sound vs reflected sounds and properly positioned loudspeakers will give a pure insightingo the recorded venue and the recorded reverb, and some late reflections from the listening environment will give the sensation of development even if that diffuse sound field probably doesn't bear much resemble of what is actually on the recordings.

I have quite similar feeling (and my personal subjective rationales) in my audio setup, as I shared here #502 on my project thread.

Spoiler

In my post #502 on my project thread, I wrote:

As you could find in my above posts # #493 and #494, I set 16 ms delay in ST+TW+SQ+WO against SW. This corresponds to that I pushed my SWs (sub-woofers) 5.5 m ( 345 m/s x 0.016 s) ahead towards my listening position.

At present, since I like the listening feeling as if I am sitting on the best S-class center seat in Concertgebouw Amsterdam, I do not like to have my SWs just beside me at my listening position; even with the "complete and perfect" time alignment of SPs I achieved, the orchestral big drum sound should be coming from the stage direction in front of my eyes and ears!

Of course, I know well that in real Concertgebouw Amsterdam (I have been there several times), I hear the entire hall tone (including the reflections, resonances, standing-waves) from all the directions surrounding me, from behind, above (ceilings), side (side walls) and floor. This is the main reasons that I insists we need suitable and preferable reverberations also in our home audio listening room which somewhat "simulates" the real hall tone, but never never to be perfect.

In my listening environment, I have fairly nice and big open spaces/rooms behind the SPs and also behind my listening position, and these (I designed so when I built my present house) are very nice for the room acoustics I fully enjoy now. Fortunately, many of the semi-professional audio enthu friends well agree me on the nice acoustics even in my present setup.

As you kindly suggested, I may try some fine tuning adjustments of physical alignment of SPs and/or Furnitures, as well as putting acoustic treatment material(s) near to the portion of "the wall"; I mean the wall which I identified causing very subtle amount of standing waves in certain frequency, but only very faintly audible with our ears with unusually high volume/gain sounds. (Measurement microphone ECM8000 is really sensitive enough to hear and see the standing waves in 35 - 45 dB below the main sound, though.)

I feel/assume that I and my wife, as well as my audio enthu friends, even unconsciously "love" these kind of faint standing waves in our listening environment, audible or inaudible, while enjoying music.

These posts would be also of your reference, I assume.

- Perfect (0.1 msec precision) time alignment of all the SP drivers greatly contributes to amazing disappearance of SPs, tightness and cleanliness of the sound, and superior 3D sound stage: #520

- Not only the precision (0.1 msec level) time alignment over all the SP drivers but also SP facing directions and sound-deadening space behind the SPs plus behind our listening position would be critically important for effective (perfect?) disappearance of speakers: #687

- A new series of audio experiments on reflective wide-3D dispersion of super-tweeter sound using random-surface hard-heavy material: Part-4_Provisional conclusion to use Case-2 reverse reflective dispersion setting in default daily music listening: #929

- The latest system setup of my DSP-based multichannel multi-SP-driver multi-amplifier fully active audio rig, including updated startup/ignition sequences and shutdown sequences: as of June 26, 2024: #931

 

[audiofooled]

The positioning and distance between the two loudspeakers should never create “a hole in the middle” of the stereo field, so it's very important to find the correct spacing between the loudspeakers that works well with the particular loudspeakers of choice which is often dependent on size and dispersion characteristics.

Absolutely, along with the rest of the good points. Even with the same loudspeakers, this would require different setups in different rooms. One of the signs of correct spacing would be the mid-bass. I also believe in Griesinger's theories and I think of it as the icing on the cake if one is able to hear the effects.

In my room the correct spacing between the channels turned out to be the half of the wavelength of 110Hz, about 1,6 meters (with listening distance being about 3,5 meters!). Sub is in the middle. Room modal behavior also implied that I get rid of the ports:

- This is a bunch of measurements at subwoofer cone level with increasing distance starting from 0.5 meters up to 3 meters, with mains ports open:

Phase:

Energy at ear level, MLP:

It's pretty obvious that, although phase alignment down to 110Hz is quite good, when modes start to kick in, pressure starts to behave quite unpredictably, leaving me with zones of high pressure where I don't want them, also a nasty dip down low which says everything bellow 30hz will be rolled off quickly. Room does what it wants.

- Ports closed, and now it's completely different:

Phase:

Ear level, MLP:

This sounds profoundly different. There's much less delayed and stored energy, there are zones of low pressure and high velocity exactly at offending modal frequencies. SPL drops as it normally should with increasing distance. I actually want that because I sit close to the back wall, where particles have nowhere to move so the pressure is quite high. The dip now shifted to 14Hz. I'll take that.

- Now when modes are much less supported, transient response is improved dramatically. Bass behaves quite predictably and is localized at the phantom center, for correlated transient signals as low as 35Hz. The more it reaches steady state, for longer duration signals, the more it starts to get enveloping. And it's not mutually exclusive, so bass transients are perceived over there as being roughly basketball sized, in the midst of enveloping sub bass over here.

My point is that people should be encouraged to experiment with their setups outside the convention if the room or loudspeakers require so. I mean, my setup is much less wide than the equilateral triangle, speakers and sub close to the front wall, asymmetric to the lateral boundaries, also MLP close to the rear wall, big no no as far as setup goes. Yet this doesn't prevent me to hear into the recordings, with sonic images localized anywhere from "up your nose", over head, top of the ceiling, to far beyond the front wall for the phantom center, also laterally up to more than twice the loudspeaker spacing outside the channels. Also early and continuous spatial impressions are no problem. It all depends on how they do decorrelation, and I want to hear that. If the recording is narrow, let it be so, if it's wide and deep, sit back and enjoy.

In the end, I want a system that makes me want to explore new genres of music as well as behave well and predictable when listening to the test tracks I'm already sick and tired of.

Admittedly, my speakers are quite oversized for my room, they don't drop SPL with distance in conventional way, them having a "longer throw" if you will. Basically, high D/R ratio is something I cannot escape inside the room boundaries even if I wanted to, once the setup is right. I also have loads of headroom with drivers barely moving for any listening levels that are normal, so I can do lots of things in low frequency department in the hopes of creating some directivity in that region as well.

One more thing I'd like to mention is this:

Time intensity trading | Auditory Neuroscience

auditoryneuroscience.com

Setting the system up for this is non trivial and it requires some special characteristics of the loudspeakers for it to sound natural. But is certainly another thing to consider.

IMO, another cool thing form the same site is this https://auditoryneuroscience.com/spatial-hearing/virtual-acoustic-space

You can experiment and have fun in comparing how this sounds on the headphones and in room.

 

[Willem]

With respect to dipoles, I think there is more than just being a dipole. For many years my main speakers were Quad 57 electrostats. I liked them, and I still do, even though they have their weaknesses. When the new 2805 came out and bought those, and they are a lot better in all respects, even though audiophiles keep lamenting that the 57's sound better. One of the things the 2805 excels in is spatial representation, and, if you like, invisibility. Both are dipole electrostats, but the modern one has a far more realistic soundstage. This is due, I think to Peter Walker's innovation of using delay lines to make sure that the sound seems to originate from the same point.

 

[dualazmak]

In my room the correct spacing between the channels turned out to be the half of the wavelength of 110Hz, about 1,6 meters (with listening distance being about 3,5 meters!). Sub is in the middle. Room modal behavior also implied that I get rid of the ports:

This is very interesting at least for me since I too have settled in similar triangle after intensive try-and-error.

The main SP cabinets are sealed ones, and hence no port in the rear of the SP cabinets (ref. #931).
The large-heavy L&R subwoofer YAMAHA YST-SW1000 has ports in front which I prefer very much.

I briefly observed the room modes of my listening acoustic environment using REW in very early stage of my audio project (ref. #22):

The latest (and almost best tuned) Fq-SPL at my listening position is like this (ref. #931).

 

[audiofooled]

This is very interesting at least for me since I too have settled in similar triangle after intensive try-and-error.

The main SP cabinets are sealed ones, and hence no port in the rear of the SP cabinets (ref. #931).
The large-heavy L&R subwoofer YAMAHA YST-SW1000 has ports in front which I prefer very much.
View attachment 414442

I briefly observed the room modes of my listening acoustic environment using REW in very early stage of my audio project (ref. #22):
View attachment 414445

The latest (and almost best tuned) Fq-SPL at my listening position is like this (ref. #931).
View attachment 414446

I can see you are a man of strong will and determination when it comes to this hobby!

My system is built for convincing mid-bass and low frequency extension amongst other things. But, when setting it up in just a normal living room I quickly realized that it won't be easy to deal with real problems of bunch of waves that are bouncing and standing. Ports firing at the front wall just had to be eliminated.

A picture may be worth a thousand words, so here are wavelets (with the same settings as in your previous post), depicting the response at MLP:

Ports:

No ports:

But this is not the whole picture. If you're interested, in attachment you may find the wavelets of measurements I previously posted, taken low to the ground with increasing distance from close field up to the MLP. What you may notice is that with ports, most of low bass energy accumulates where I don't want it, namely underneath my seat. Elsewhere in the room, this translated to unpredictable energy vs. frequency, wanting to bring the house down at higher SPL.

No ports, and it shifts to the ear level, with very little energy underneath that gets partially absorbed by the sofa, and quite benign elsewhere in the room. It's all about balancing the energy dictated by the room modes and the floor bounce. Also, with no ports, I can have as little as 0.5 dB difference for low frequencies in between the channels.

One may ask, then, what does this have to do with soundstage depth? Well, all the bass transients are now localized at the phantom center and stay put when listening from anywhere in the room, with sustained bass decaying more quickly and "following" me anywhere I walk around. This gives me another depth cue.