Properties of speakers that creates a large and precise soundstage

[MattHooper]

I found the Griesnger work illuminating, one of the most helpful on this subject. I think it might have been suggested by Duke. Checked—yep—January 9 from Duke. It was a YouTube lecture/presentation:

I need to watch again for a refresher. I suspect everyone here has seen it or studied his other works, but if not, well worth the 20 minutes or so for sure.

That was interesting (though at sometimes a bit rushed IMO)...

Griesnger's talk about "proximity" of sound being "engaging" meshes pretty well with my experience or tastes. Though his specific talk wasn't exactly concerning the following, it does remind me of the point made about advantages I get from my 2 channel speaker set up vs surround:

I think it's quite common that, as in my own case, for sheer practicality surround systems are often not placed in the positions many audiophiles prefer for 2 channel listening.
So my surround speakers are on the walls, my L/C/R speaker flank my projection screen, whereas my 2 channel speakers are pulled out "audiophile style" in to the room to form the perfect listening triangle.

One of the benefits of the closer positioning of the stereo speakers is "proximity" to the sound, IMO. The speakers are capable of casting a VERY deep soundstage that can seem to go waaaay back beyond the room borders, for sounds in a soundstage that are placed in the distance. But on the other hand, sounds that are mixed right "up front" occur at the front plane of the speakers, and so their "proximity" is very close to me, and attention grabbing. This means that not only is there a wide variety of depths and layers of imaging that happens when switching between recordings, but recordings that contain a wide layering of sound - e.g. from right up close say sounds panned left and right at the speaker plane, to sounds waaaay back in the soundstage, can therefore create a tremendous impression of layering and depth, from "really close right next to me" to "very far away."

I don't get this from my surround system if only because, of necessity, the L/C/R speakers are significantly further away and even the closest mixed sounds start at quite a distance from me, and thus are never as attention grabbing and I don't get the same sense of the whole room opening up in space, from close to me to far away.

(And that's also why, despite having heard plenty of surround music, I still find my 2 channel set up can sound more convincing and enveloping in some respects. For some live performances the imaging actually feels more like being among the audience, hearing through a real acoustic space, to the performers. Where there are other times, other tracks, where I might prefer the surround immersion of the surround system).

 

[program2000]

When I wrote "sounds materialize", I meant the impression that, for example, the vocals and snare drum are so close to the listener that we see these sounds next to the face, literally a few centimeters from the face. Other sounds are meters away. And it sounds so realistic, as if each sound had its own separate speaker somewhere in the space in front of us. The surround systems that I know simply surrounds us with sound, but none of the sounds are close to us, they are only distant, just like the speakers that surround us. I'm not an audiophile, just a music producer, and after 30 years of working with music, I was able to experience this materializing sound from a stereo set only once, which was shocking to me, I didn't know it was even possible.

 

[Duke]

One of the benefits of the closer positioning of the stereo speakers is "proximity" to the sound, IMO. The speakers are capable of casting a VERY deep soundstage that can seem to go waaaay back beyond the room borders, for sounds in a soundstage that are placed in the distance. But on the other hand, sounds that are mixed right "up front" occur at the front plane of the speakers, and so their "proximity" is very close to me, and attention grabbing. This means that not only is there a wide variety of depths and layers of imaging that happens when switching between recordings, but recordings that contain a wide layering of sound - e.g. from right up close say sounds panned left and right at the speaker plane, to sounds waaaay back in the soundstage, can therefore create a tremendous impression of layering and depth, from "really close right next to me" to "very far away."

I don't get this from my surround system if only because, of necessity, the L/C/R speakers are significantly further away and even the closest mixed sounds start at quite a distance from me, and thus are never as attention grabbing and I don't get the same sense of the whole room opening up in space, from close to me to far away.

I have a friend who could easily be an amplifier or loudspeaker manufacturer if he wanted to, who has worked with me on product development many times. He used his multichannel system for two-channel and for multichannel, using a trinaural processor in two-channel mode. His (identical) front three speakers were maybe five or six feet in front of the front wall, and he had a motorized screen that would drop down in front of them. So there was no penalty in multi-channel mode from unfavorable speaker positioning.

Watching music videos with the center channel speaker engaged, the soundstage depth was pretty much the same from one recording to the next and did not seem to go much (if any) deeper than the location of the center channel speaker. In phantom center mode, the soundstage depth varied significantly from one recording to the next and could be very deep if that's what was on the recording. Likewise with two-channel program material and using his trinaural processor and the center-channel speaker, the soundstage depth was significantly constrained relative to when only the left and right speakers were engaged.

He sold his center-channel speaker and his trinaural processor and never looked back. This was over ten years ago.

My point being, what you observed might not be entirely due to the unfavorable positioning of the speakers in your multichannel system.

 

[jim1274]

That was interesting (though at sometimes a bit rushed IMO)...

Griesnger's talk about "proximity" of sound being "engaging" meshes pretty well with my experience or tastes. Though his specific talk wasn't exactly concerning the following, it does remind me of the point made about advantages I get from my 2 channel speaker set up vs surround:

I think it's quite common that, as in my own case, for sheer practicality surround systems are often not placed in the positions many audiophiles prefer for 2 channel listening.
So my surround speakers are on the walls, my L/C/R speaker flank my projection screen, whereas my 2 channel speakers are pulled out "audiophile style" in to the room to form the perfect listening triangle.

Exactly—front stage for surround needs to be on roughly the same plane as the TV display—for 2 channel in almost any situation they need to be well forward into the room. My Front towers are so big, I’d never move them out into the room. Did that once during this testing adventure and said never again.

One of the benefits of the closer positioning of the stereo speakers is "proximity" to the sound, IMO. The speakers are capable of casting a VERY deep soundstage that can seem to go waaaay back beyond the room borders, for sounds in a soundstage that are placed in the distance. But on the other hand, sounds that are mixed right "up front" occur at the front plane of the speakers, and so their "proximity" is very close to me, and attention grabbing. This means that not only is there a wide variety of depths and layers of imaging that happens when switching between recordings, but recordings that contain a wide layering of sound - e.g. from right up close say sounds panned left and right at the speaker plane, to sounds waaaay back in the soundstage, can therefore create a tremendous impression of layering and depth, from "really close right next to me" to "very far away."

I don't get this from my surround system if only because, of necessity, the L/C/R speakers are significantly further away and even the closest mixed sounds start at quite a distance from me, and thus are never as attention grabbing and I don't get the same sense of the whole room opening up in space, from close to me to far away.

Not sure if I entirely understand what you mean here.

Isn’t the point of setting exact speaker distance and level for surround channels using the AVR set-up function “solve” that? Distance setting is critical for proper delay time and sound level adjustment for proper volume leveling.

(And that's also why, despite having heard plenty of surround music, I still find my 2 channel set up can sound more convincing and enveloping in some respects. For some live performances the imaging actually feels more like being among the audience, hearing through a real acoustic space, to the performers. Where there are other times, other tracks, where I might prefer the surround immersion of the surround system).

How discrete multichannel recordings are mixed is all over the map. A true discrete multichannel recording is not a fair or proper comparison to DSP’d 2 channel source played through however many surround speakers you deploy.

I’ll be using DSP of 2 channel source material to create a version of “fake” surround to compare to soundstage aspects vs 2 stereo speakers alone.

 

[jim1274]

When I wrote "sounds materialize", I meant the impression that, for example, the vocals and snare drum are so close to the listener that we see these sounds next to the face, literally a few centimeters from the face. Other sounds are meters away. And it sounds so realistic, as if each sound had its own separate speaker somewhere in the space in front of us. The surround systems that I know simply surrounds us with sound, but none of the sounds are close to us, they are only distant, just like the speakers that surround us. I'm not an audiophile, just a music producer, and after 30 years of working with music, I was able to experience this materializing sound from a stereo set only once, which was shocking to me, I didn't know it was even possible.

I get what you mean, but probably would describe it as a wide and deep soundstage with good instrument spread and spatial localization. Or something like that.

 

[jim1274]

I have a friend who could easily be an amplifier or loudspeaker manufacturer if he wanted to, who has worked with me on product development many times. He used his multichannel system for two-channel and for multichannel, using a trinaural processor in two-channel mode. His (identical) front three speakers were maybe five or six feet in front of the front wall, and he had a motorized screen that would drop down in front of them. So there was no penalty in multi-channel mode from unfavorable speaker positioning.

Watching music videos with the center channel speaker engaged, the soundstage depth was pretty much the same from one recording to the next and did not seem to go much (if any) deeper than the location of the center channel speaker. In phantom center mode, the soundstage depth varied significantly from one recording to the next and could be very deep if that's what was on the recording. Likewise with two-channel program material and using his trinaural processor and the center-channel speaker, the soundstage depth was significantly constrained relative to when only the left and right speakers were engaged.

He sold his center-channel speaker and his trinaural processor and never looked back. This was over ten years ago.

My point being, what you observed might not be entirely due to the unfavorable positioning of the speakers in your multichannel system.

This made me think of the Hughes unit from back in the 80’s:

https://gearspace.com/gear/hughes/ak-100

Not what you describe, but still a soundstage “enhancer” gizmo.

 

[Duke]

When I wrote "sounds materialize", I meant the impression that, for example, the vocals and snare drum are so close to the listener that we see these sounds next to the face, literally a few centimeters from the face. Other sounds are meters away. And it sounds so realistic, as if each sound had its own separate speaker somewhere in the space in front of us. The surround systems that I know simply surrounds us with sound, but none of the sounds are close to us, they are only distant, just like the speakers that surround us. I'm not an audiophile, just a music producer, and after 30 years of working with music, I was able to experience this materializing sound from a stereo set only once, which was shocking to me, I didn't know it was even possible.

What was different about that particular stereo set, and/or about how it was set up? Or was it the recording that was different?

 

[Duke]

This made me think of the Hughes unit from back in the 80’s:

https://gearspace.com/gear/hughes/ak-100

Not what you describe, but still a soundstage “enhancer” gizmo.

I think I owned something conceptually similar (but probably much less sophisticated) in the early 80's. Mine was called the "Omnisonic Imager". I think it was a knock-off of the Carver Sonic Hologram Generator, and probably the cheapest such unit on the market. The settings for realistic-sounding results varied enormously from one recording to the next. Once the gee-whiz novelty wore off, it seemed like more trouble than it was worth. The unit also boosted the bass too much, presumably as a first-approximation compensation for making the difference signals louder.

Next I tried the "Haflter Hookup", using two rear speakers wired to play only the difference signal. I made mine level-adjustable by driving the two difference-signal speakers with their own dedicated integrated amp. It was a learning experience, once again more trouble than it was worth.

 

[jim1274]

Once in my life I experienced holographic imaging and out of four pairs of speakers placed next to each other, only one played in 3D. The rest is a flat wall of sound. So I know speakers, room and positioning matters.

So it was 8 speakers all aligned in the same plane in front of you? I’m not getting the one played in 3D part.

 

[jim1274]

I think I owned something conceptually similar (but probably much less sophisticated) in the early 80's. Mine was called the "Omnisonic Imager". I think it was a knock-off of the Carver Sonic Hologram Generator, and probably the cheapest such unit on the market. The settings for realistic-sounding results varied enormously from one recording to the next. Once the gee-whiz novelty wore off, it seemed like more trouble than it was worth. The unit also boosted the bass too much, presumably as a first-approximation compensation for making the difference signals louder.

The few times I heard any of these “enhancers”they seemed to do more harm than good to the overall Sonics, and the variation in effect between recordings was significant as you said, sometimes not in a good way.

Next I tried the "Haflter Hookup", using two rear speakers wired to play only the difference signal. I made mine level-adjustable by driving the two difference-signal speakers with their own dedicated integrated amp. It was a learning experience, once again more trouble than it was worth.

I had a table radio some years back that had some type of soundstage enhancement feature, but can’t recall the name. It didn’t last very long on the table…

 

[jim1274]

What was different about that particular stereo set, and/or about how it was set up? Or was it the recording that was different?

I’m wondering if he was referring to the 4 stereo pair configuration in his post shortly before this one you replied to? The last sentence about only hearing it once with a stereo set confused me. Or am I just slow on the uptake?

I’m more than curious about the comment on sounds right by his face. I’ve experienced instrument placement extremely wide and deep, but have not found a speaker or placement yet that extended well in front of the plane of the speakers.

Ok program2000, cough up the “secret sauce” here—this is “holy grail” stuff. I’m truly not being sarcastic at all—just a seeker who wants another piece of the puzzle.

 

[MattHooper]

Not sure if I entirely understand what you mean here.

Isn’t the point of setting exact speaker distance and level for surround channels using the AVR set-up function “solve” that? Distance setting is critical for proper delay time and sound level adjustment for proper volume leveling.

No. Speaker distance settings are more helpful for flattening out room problems and maintaining sonic coherence among the speakers. It's not about making speakers you've placed 14 feet away from the listening position sound like they are only 5 feet away from you. If you place your speakers 14 feet away, that's how far away the "closest" mixed sound can be to you (unless either some perceptual tricks are used like Qsound, or some object-type mixing can sometimes seem to adjust the proximity of sounds more forward, but that's not really getting at what I'm talking about).

So in other words, the speaker distance info for the AVR isn't about saying "make the speakers I placed 15 feet from my listening seat fool me in to sounding like they are placed 5 feet from my listening seat." The speakers will sound as far away as you place them, essentially. You are deciding where the front plane of your sound stage will start in how you place your speakers.

 

[audiofooled]

I was able to experience this materializing sound from a stereo set only once, which was shocking to me, I didn't know it was even possible.

I'm not surprised you experienced this only once, achieving this in room is non trivial. IME, if a system does that, there are tracks that do this, but the opposite is rarely true.

Since you brought this up, and I can see people are interested, I'd like to propose a couple of short videos, just that people understand the complexity:

and part 2, where part 1 is actually explained:

As I stated earlier, I'm not trying to start a debate about phase, there's other thread about it. But, in addition to my post #734, I can share my unsmoothed and unwrapped phase plot from 2 measurement points, 0,5 meters subwoofer cone level (brown), and at 3,5m (ear level, MLP), both the sub and the mains, no EQ or any DSP:

Whatever system you have (I designed and built mine from scratch), the starting point in setting it up in room would be how the loudspeakers would interact with each other. IME, equilateral triangle is not always the best choice. Triangle, yes, but loudspeakers themselves would have a dispersion pattern that would pretty much dictate how far apart you place them, also to where you point them. First thing to be solved would be fixing your phantom center. In stereo, I think that's what's most important. The most obvious thing to listen for when spacing and pointing your mains would be the mid-bass. It will also affect vocals with regards to coherency, speech intelligibility, clarity and definition.

When you solve the phase for low up to mid bass frequencies, there's a pretty good chance that your system now would respond well to EQ.

For higher up in frequency you have to solve where the heck in the room you place your system and your listening position. Oh, did I mention that the first step would require knowing how on earth would you solve the second one, because the room would mess everything up anyways? Ha! See how complex things may get?

There are reflections and comb filtering effects. One tool that I may propose would be:

https://amcoustics.com/tools/amroc (for identifying your room modes and get a general idea where to start)

Also this:

https://amcoustics.com/tools/amray

This is a ray tracing sketchpad where you can model your room, place your loudspeakers and listening position, enter the angles to where you plan to point them and see where are your reflection points and how far apart in ms they are. This is where Griesinger's work most certainly comes into play. Unfortunatelly, it seems that tool allows you to model only one loudspeaker at the time.

In conclusion, there's nothing to conclude, really. Different speakers, different rooms, so I guess everyone needs to figure this out for themselves. Just don't smooth your measurements. IMO they are most helpful when viewed as ugly as they are. Our hearing works differently, and one measurement for a single point in space doesn't tell you the whole thing.

 

[tmuikku]

I haven't listened to just one of my speakers, with the exception of the occasional track which might begin with a sound hard panned to one side.

But I certainly do move my head around to note imaging differences. For instance I have plenty of tracks where the image seems to extend well beyond the side boundaries of the speakers, like some soundtracks with tympani that sounds like it's coming from a distant corner of the soundstage beyond the right speaker. If I sway my head more towards the right speaker, putting myself more on axis to it, some of that illusion collapses a little bit, sounding more direct in line from the speaker position.

Idea behind single speaker noise experiment is to simplify as much as possible, the noise has no spatial information, nor does single speaker, so you'd mainly hear just how you perceive the sound in the room originating from the speaker, basically to gauge your own perception. Of course using the transition to A/B the effect is required to make any sense of it, toggle the sound with auditory system.

The experiment can give insight how to tune your system. For example: I knew the location in my room the stream separation happens, now playing noise from one speaker I could go standing there and listen what changes when I move bit closer of further. Interestingly most notable change was that opposing side wall reflection seemed to mute, very weird thing to hear as there is nothing in the room or speaker that would mute it but it just the auditory system doing a trick. Now that I had exact location marked on the floor where I have to stand the mute happens, I could just adjust toe-in and listen again, try to move the point as far into the room as possible. After adjusting toe-in I could get it even further by placing a makeshift acoustic treatment to mute it "manually" so to speak My hope was to move the transition point with stereo sound bit further into room, and I think it moved some, at least the transition got mellowed out a bit.

I've yet to do the same test with snare, band limited noise and so on, would be again a lot of nice information to process on. Honestly sound is pretty good, has been, so I haven't been tweaking since christmas, of to other things for now.

 

[gnarly]

Isn’t the point of setting exact speaker distance and level for surround channels using the AVR set-up function “solve” that? Distance setting is critical for proper delay time and sound level adjustment for proper volume leveling.

I would say yes.
Because the time-of-flight of the physical distance from each speaker summed with the delay given to each speaker, should all be equal for every speaker when measured at the listening position. It's making sure all arrivals occur simultaneously that counts.

A simple example using LCR given 3 identical speakers being used......
The Center channel needs to be placed on a physical arc with L & R speakers such that all three are equidistant, or the Center channel needs to be time delayed by the difference in its physical distance vs L or R (which are assumed to be the same)

Ime, if the Center channel doesn't have either equal physical distance or time-delay alignment, it really messes up any soundstage comparisons vs stereo.

 

[jim1274]

No. Speaker distance settings are more helpful for flattening out room problems and maintaining sonic coherence among the speakers. It's not about making speakers you've placed 14 feet away from the listening position sound like they are only 5 feet away from you. If you place your speakers 14 feet away, that's how far away the "closest" mixed sound can be to you (unless either some perceptual tricks are used like Qsound, or some object-type mixing can sometimes seem to adjust the proximity of sounds more forward, but that's not really getting at what I'm talking about).

So in other words, the speaker distance info for the AVR isn't about saying "make the speakers I placed 15 feet from my listening seat fool me in to sounding like they are placed 5 feet from my listening seat." The speakers will sound as far away as you place them, essentially. You are deciding where the front plane of your sound stage will start in how you place your speakers.

I’m only talking discrete multichannel surround sound—-This explains it better than I did:

Speaker Delay: How to Set Up Surround Sound Speakers?

Wondering how speaker delay impacts your surround sound speaker set up? This article explains the human eccentricity that makes this a necessary step.

www.axiomaudio.com

The reason you must set this correctly involves compensation for a peculiarity of human hearing--our tendency to locate a close-up sound before one that’s a little farther away.

The receiver digitizes all the incoming signals for each channel and inserts an appropriate degree of digital speaker delay so that the sounds radiated by the two surround channels reach your ears after the sound from the center channel and the main speakers. Since in many installations, the surround speakers will be closer to the listening area than the main front speakers or center channel, our ears would normally detect the sound from the surround speakers before that from the main speakers, which would totally throw off the illusion. So by entering the distance from your chair to the surround speakers, the Yamaha inserts enough digital delay so you will hear the sound from the main front speakers first, then that from the surround speakers. (Sound travels at about 1 foot per millisecond, so it's easy for a digital circuit to apply the appropriate amount of delay.)

It’s a kind of precedence effect that's important in real life because it helps protect and alert us to nearby events that might be threatening. (If a car horn beeps to your right as you step off a curb, you immediately glance there and take steps to avoid being run over.) But with multi-channel surround sound, the illusion only works if the sounds from nearby surround speakers reach your ears after the sound from the main front speakers

 

[jim1274]

I would say yes.
Because the time-of-flight of the physical distance from each speaker summed with the delay given to each speaker, should all be equal for every speaker when measured at the listening position. It's making sure all arrivals occur simultaneously that counts.

A simple example using LCR given 3 identical speakers being used......
The Center channel needs to be placed on a physical arc with L & R speakers such that all three are equidistant, or the Center channel needs to be time delayed by the difference in its physical distance vs L or R (which are assumed to be the same)

Placing the LCR on the same arc would probably be best option (exactly what Dolby shows on their layouts) and likely what most folks do. Sometimes it’s not as easy to do in practice, so the distance setting corrects for the time delay difference. I just looked at the Dolby layout spec for my 7.1.4 overhead ATMOS speaker layout—shows the front stage in arc you describe:

https://www.dolby.com/siteassets/about/support/guide/setup-guides/7.1.4-dolby-atmos-enabled-speaker-placement/7_1_4_atmos_enabled_speaker_setup.pdf

Ime, if the Center channel doesn't have either equal physical distance or time-delay alignment, it really messes up any soundstage comparisons vs stereo.

Total agreement.

 

[goat76]

Whatever system you have (I designed and built mine from scratch), the starting point in setting it up in room would be how the loudspeakers would interact with each other. IME, equilateral triangle is not always the best choice. Triangle, yes, but loudspeakers themselves would have a dispersion pattern that would pretty much dictate how far apart you place them, also to where you point them. First thing to be solved would be fixing your phantom center. In stereo, I think that's what's most important. The most obvious thing to listen for when spacing and pointing your mains would be the mid-bass. It will also affect vocals with regards to coherency, speech intelligibility, clarity and definition.

Yes, the loudspeakers should be placed so that the phantom center comes close to sounding as if a physical center speaker is in place. When the correct distance between the particular speakers in use is found, the phantom-centered sounds will have a distinct pinpoint sound and will act as an "anchor" to the whole stereo image, and pretty much everything down to the smallest details in the stereo image will then fall in the right place including depth and width. So yes, that's one of the first things that needs to be figured out with the particular choice of loudspeakers.

But in my opinion, an equilateral triangle is the way to go if we want to approach the most common studio standard, which in turn will most likely give us the intended width of the mix as it was heard by the engineer in the studio. So when we have figured out what distance works well for our speakers to get that distinct sounding phantom center, we can almost be sure that 'about' the same listening distance will also work pretty well (as in an equilateral triangle). So now that we approximately know how large this listening triangle should be, we can move this triangle around inside the space of the listening room to find the best compromise when it comes to both the listening position and the placement of the loudspeaker's distance to the walls.

There are many parameters to adjust, and there will unavoidably be a lot of compromises to consider when setting up a stereo system in a room.

 

[jim1274]

Has anyone else following along here tried the DSP multichannel creation of 2 channel stereo source discussed in recent posts? Well, I just did. Used a live Amsterdam hall recording since I have that particular hall DSP setting that was noted earlier as a top 3 acoustics venue. Waaaaay too much reverb/echo with really no improvement in soundstage aspects (depth, width, or instrument placement) and a HUGE hit on clarity. I was only getting signal output from the front stage and the side surrounds, the sides being at such a low level you had to put your ear up to them. I’d give this a big thumbs down from this first comparison.

Suppose I should try some other DSP modes with shorter delay times in the interest of thoroughness.

I see heads nodding out there, thinking “I told you so”…. Well, this isn’t any different than what I expected, recalling trying it a long time ago and abandoning it almost instantly.

 

[audiofooled]

Has anyone else following along here tried the DSP multichannel creation of 2 channel stereo source discussed in recent posts? Well, I just did. Used a live Amsterdam hall recording since I have that particular hall DSP setting that was noted earlier as a top 3 acoustics venue. Waaaaay too much reverb/echo with really no improvement in soundstage aspects (depth, width, or instrument placement) and a HUGE hit on clarity. I was only getting signal output from the front stage and the side surrounds, the sides being at such a low level you had to put your ear up to them. I’d give this a big thumbs down from this first comparison.

Suppose I should try some other DSP modes with shorter delay times in the interest of thoroughness.

I see heads nodding out there, thinking “I told you so”…. Well, this isn’t any different than what I expected, recalling trying it a long time ago and abandoning it almost instantly.

Some years ago I tried setting up a 7.1 with my Yamaha DSP a-1000. I was actually quite impressed by how much engineering had been put to this and the various effects it offered. The sound was truly immersive on some tracks, especially live performances. On majority of other tracks the effects were less then enjoyable and convincing. Reverted to 2.1 and never looked back.