The quantification of stereo imaging

[dshreter]

One of the most rewarding attributes of a stereo system that is setup well in a room is the presence of clear stereo imaging. For me, this is the sensation that sound is emanating directly from the space between the speakers, but others even go as far as to describe a 3d sound stage that can extend in front or behind the speakers.

I've listened to high quality speakers in poor rooms or located poorly, and that sense of imaging is lost and seems more like a wall of sound or even worse, sensing that the sound is emanating from the speakers. The same happens if drifting too far from the sweet spot in a well constructed system too.

Given the importance of imaging (at least to me), are there methods for measuring by microphone(s) how well a system is performing in this regard? If not, are there ideas for how this could be achieved? I'd love to be able to measurably optimize this attribute when dialing in a room.

 

[restorer-john]

Imaging is hit and miss. Mostly miss.

If you get it on one recording, 50 others will just be normal L-R with no "3D" space. Hundreds of speakers later, many different rooms, houses and tens of thousands of recordings later I have some holographic recordings that could sell any speaker to anyone but as a pursuit, it's chasing dragons if you ask me.

You'll get some wet-behind-the-ears audiophools going on about how their system resolves depth and height better than everyone else's and when you listen to their carefully picked "demo" music, it's just pinpoint L-R placement and nothing more.

But when it comes to marketing, wow, imaging is everywhere from the speakers wire to passive volume pots. Wait until someone chimes in with "your system isn't resolving enough to achieve good imaging"...

 

[dshreter]

If you get it on one recording, 50 others will just be normal L-R with no "3D" space.

I suggest something less ambitious as a starting point, to be able to measure just the L-R imaging quality. It's something you clearly sense when a system's sweet spot is locked in, and the speakers seem to disappear. Given it is so obvious to the ears, I would bet it is easily measurable.

 

[Blumlein 88]

I don't have the magic answers either. I have an interesting example of what can happen. I've a video 5.1 setup in a moderately large room. If listening to just stereo at the main listening position, things sound good, pin point imaging, and more sense of space if the recording has it. However, move back further about 8 feet deeper into the room from the front, and things change for the better. Much better sense of 3D space, and images esconsed within that space on any pretty decent recording. I suppose the amount of side wall reflections are better proportioned there and lacking closer in.

 

[Blumlein 88]

The LEDR test gives you some idea if you speakers are setup well. Here it is online. Was on early Stereophile test CD's and Chesky demo CD's.
https://www.audiocheck.net/audiotests_ledr.php

Also the recently discussed JA interview, he mentions playing pink noise on both channels and seeing if you get a very thin pinpoint location in the middle with no sense it is coming from the sides.

 

[dshreter]

Also the recently discussed JA interview, he mentions playing pink noise on both channels and seeing if you get a very thin pinpoint location in the middle with no sense it is coming from the sides.

Cool stuff. I wonder what you could measure though to say how thin that pinpoint location is so you could seek out optimization.

 

[Pio2001]

The first requirement for such a measurement, if it exists one day, would be a dummy head with two microphones.

 

[NTK]

There are these devices called acoustic cameras. They are microphone arrays used to locate sound source for applications such as noise source identification. I had wondered if these devices can be used to evaluate the imaging capabilities of stereo speakers.

Then I realize that these acoustic cameras are way smarter than us. They are not going to be fooled by the phantom images produced by a pair of speakers. They will have no problem telling us that the sound came only from two locations (i.e the speaker on the left and the speaker on the right), and there isn't a singer between the speakers in the room.

So why are these acoustic cameras invented in the first place? Precisely because the ability of our ears to locate sound source is poor. When we pop open the hood of our cars and listen to the engine, we have a hard time telling which components generate which sound. This is what these devices were invented to do.

Here is an interesting video (to me at least) from B&K.

 

[RayDunzl]

When we pop open the hood of our cars and listen to the engine, we have a hard time telling which components generate which sound.

On my Maxima (R.I.P) I stuck a mic under the hood, and slowed things down a bit...

The three injectors on the front bank of the V6.

https://soundcloud.com/ray-880875693%2Ffuel-injectors

 

[Doodski]

On my Maxima (R.I.P) I stuck a mic under the hood, and slowed things down a bit...

The three injectors on the front bank of the V6.

https://soundcloud.com/ray-880875693%2Ffuel-injectors

Sounds like two metronomes going at it together. I had no idea fuel injectors could create such a sound.

 

[direstraitsfan98]

Having a system that has precise stereo imaging is absolutely a thing and exists.

From my experience its a room and placement thing, not a speaker or equipment thing. All speakers can pull the neat trick of having a stereo image. I've noticed that the horns on my JBL 4367 offer a more 'hanging in the air' sort of voice then other speakers I have owned. It's kind of a spooky thing, like the voice is hanging in the air in front of you. You can close your eyes and hear the voice move, again, depending on the recording.

I love the track Paultalk from the album Peter, Paul & Mary: Live in Japan 1967. There is a strong sense of venue space in the recording and as both artists, Paul and his translator talk on stage you can hear their placements, as well as when Paul starts moving around. You can tell when he moves the mic closer to his mouth, and when he's stepped off to the side. You can hear and feel the rumble of the footsteps on the stage as they walk around.

https://tidal.com/browse/track/18236272

 

[pozz]

Cool stuff. I wonder what you could measure though to say how thin that pinpoint location is so you could seek out optimization.

There are guidelines on listening room and studio design for imaging, but the endpoint, measuring the result, hasn't been done as far as I know.

Even Trevor Cox, one of the main authorities on acoustic treatment, only has general comments in his book on absorbers/diffusors.

Very interesting proposal by the way.

 

[NTK]

On my Maxima (R.I.P) I stuck a mic under the hood, and slowed things down a bit...

The three injectors on the front bank of the V6.

That's really cool. Was there any problem with your Maxima at the time?

Come back to the topic of this thread, I think our brains do a lot of improvisation, in addition to the auditory signals from our ears, to make up the sound image we experience. Without a good understanding of how our brains do the processing, I doubt we will get very far with just a few mics.

Even our brains don't think our ears are very reliable. We hear about the ventriloquism effect. When there are visual cues, our brains trust our eyes over our ears.

 

[pozz]

Even our brains don't think that our ears are very reliable. We hear about the ventriloquism effect. When there are visual cues, our brains trust our eyes over our ears.

Griesinger has done some very excellent work on localization and attention (his interest is specifically in concert hall acoustics). One thing he's claimed over and over is that the eye aids localization.

Example: My girlfriend was watching TV and on her phone. I was puttering around nearby with my head down. At some point the program went quiet and then music played. I couldn't localize the music accurately, but I tried hard to without looking up. Maybe ten seconds passed, which is a long time subjectively when concentrating, but I had no idea if it was coming from the TV or her phone. I looked up, figured out it was her phone, then looked back down. Thing is, I couldn't then unlocalize the music. It was snapped to where she was sitting on the couch where before it was diffuse and nebulous.

Griesinger reports similar stories to do with the ability to localize instruments on stage. He actually brings this up as a criticism because actual aural localization requires better hall design, but is, a lot of the time, compensated for in an unacknowledged way by listeners watching the performances.

 

[Kvalsvoll]

It can be measured, and you can tell from the measurement if a system is likely to provide good imaging - or most likely will be hopeless. However, it is not so easy to see exactly how this imaging - or soundstage, will be perceived by a listener. Add to this, that there are significant differences between us listeners, in how we perceive this soundstage.

First level is l-r placement and a solid center image. Then there is depth - then you get instruments and sounds in a 3D-like scene, some are closer, some appear to be farther away.

But things get to a different level once you experience instruments and sound objects that appear like they have physical size and shape, completely separated form each other, some are huge, some are smaller, some are tiny and very precise, some are diffuse and kind of fill the room. You can also hear the size and boundaries of the venue where the recording took place.

The magic works best exactly at the center point between the speakers. But some speakers can work better across a larger area, and some present a soundstage that is completely dislodged from the speakers across the whole listening room, but it will not be as precise and real as for the listener in center.

A good system will present some kind of soundstage on all recordings. But, it is of course the recording that determines where instruments and sounds appear and how real those objects are.

It is mostly the speakers and placement of speakers and listener that determines the quality of all this. A lesser speaker can be helped by improving room acoustics, but the speakers will limit how good the images of the instruments can be.

 

[pozz]

It can be measured, and you can tell from the measurement if a system is likely to provide good imaging - or most likely will be hopeless. However, it is not so easy to see exactly how this imaging - or soundstage, will be perceived by a listener. Add to this, that there are significant differences between us listeners, in how we perceive this soundstage.

First level is l-r placement and a solid center image. Then there is depth - then you get instruments and sounds in a 3D-like scene, some are closer, some appear to be farther away.

But things get to a different level once you experience instruments and sound objects that appear like they have physical size and shape, completely separated form each other, some are huge, some are smaller, some are tiny and very precise, some are diffuse and kind of fill the room. You can also hear the size and boundaries of the venue where the recording took place.

The magic works best exactly at the center point between the speakers. But some speakers can work better across a larger area, and some present a soundstage that is completely dislodged from the speakers across the whole listening room, but it will not be as precise and real as for the listener in center.

A good system will present some kind of soundstage on all recordings. But, it is of course the recording that determines where instruments and sounds appear and how real those objects are.

It is mostly the speakers and placement of speakers and listener that determines the quality of all this. A lesser speaker can be helped by improving room acoustics, but the speakers will limit how good the images of the instruments can be.

At least from what you described, there's no measurement as such. There's no computation of apparent source width and no way to establish precision besides listening.

We talk about soundstage a lot and we can derive that information from recorded signals, but there's no established method that I know of to predict or evaluate the soundstage of an acoustic environment ahead of time with a given pair of speakers.

 

[Newk Yuler]

Great thread. When the 3D is happening the sound projects in thick layers way into the room and the speakers disappear. Magnificently cool thing about stereo imaging when it's happening. Jaw dropping coolness. One of the reasons I love electronic music that's crafted with spacey 3D effects. It sounds so damn awesome.

Easy test if/when it's working: Can you close your eyes and find the speakers in your mind's eye? IME it's often easy to see the speakers but it's gratifying to find a system can do it and it's the program material when it isn't. I'm convinced many systems can't and it's a frustrating obstacle when I can't make something I've assembled do it. Maddening with so many variables to sort. One of which (thanks to ASR) is wondering if a crucial component is crappy, failing, or noisy enough to ruin the resolution of the sound.

On my Maxima (R.I.P) I stuck a mic under the hood, and slowed things down a bit... The three injectors on the front bank of the V6.

I put my newish Mazda 6 in the dealership's shop before the 3 year warranty ran out because I was convinced I was hearing rocker arms lightly rattling under acceleration when the engine was cold. They kept it overnight and then told me it was the normal sound of the injection system. I was happy to find it wasn't a problem but it's apparently a benefit to know the difference before thinking the engine is falling apart. If I had done it after the warranty expired I would have probably paid to be told I was ignorant. That would have been a harder lesson.

 

[Kvalsvoll]

At least from what you described, there's no measurement as such. There's no computation of apparent source width and no way to establish precision besides listening.

We talk about soundstage a lot and we can derive that information from recorded signals, but there's no established method that I know of to predict or evaluate the soundstage of an acoustic environment ahead of time with a given pair of speakers.

Acoustic measurement of the system response at the listening position gives an indication of accuracy performance. For measurements on the same system, those measurements can be compared to give very good information about differences when something is changed.

To be meaningful, this requires some kind of reference for which the measurement can be compared against. For different systems in different rooms, one can use a known system as reference. Still, it is not easy to see exactly how good the performance is, there is no single performance number. It is not easy to see how the soundstage appears - is it wide, how deep. But it will give some indication of what to expect, and you can see things like whether low frequency instruments will have the same accuracy in placement as higher frequency instruments.

 

[pozz]

One of the reasons I love electronic music that's crafted with spacey 3D effects. It sounds so damn awesome.

Very much the case for me too.

 

[LDKTA]

The most accurate (two channel) stereo playback system is one that puts all the soundstage at a constant height. Any perceived effect of depth, width or height is purely an artifact of the loudspeakers, the loudspeakers interaction with room reflections (delay), and response anomalies mimicking the ears spectral modification that encodes the source (height, width and depth). A cool illusion nonetheless.

One of the most common "subjectivist" questions is, how can we measure the soundstage? I ask, why would you want to measure something that isn't actually there and how meaningful would those measurements be?