What speaker measurement references how good imaging will be?

[Mike Up]

I have recently tested side by side 5 pair of speakers I have and imaging goes from poor to good with most having a very narrow sweet spot for imaging and only 2 having a decent sweet spot before the image breaks up to either speaker on the sides.

I've seen reviews here, at Erin's audio corner and Audio Science Review and don't know which dictates how well imaging will be. As far as I know, it has a lot to do with time alignment with the drivers and crossovers but don't recall a measurement for that.

Thanks

 

[TLS Guy]

I have recently tested side by side 5 pair of speakers I have and imaging goes from poor to good with most having a very narrow sweet spot for imaging and only 2 having a decent sweet spot before the image breaks up to either speaker on the sides.

I've seen reviews here, at Erin's audio corner and Audio Science Review and don't know which dictates how well imaging will be. As far as I know, it has a lot to do with time alignment with the drivers and crossovers but don't recall a measurement for that.

Thanks

That is difficult to say. My strong impression is that a good polar response is a prerequisite, in that the off axis response should mirror the axis response closely until about the 10K range. I do think excess bass and especially poor quality bass upsets the image. I tend to be agnostic about wave guides. They are in vogue now, but too narrow a dispersion from any driver I think is a negative. I tend to favor the shallowest of wave guides. Then you can't forget the room. Room acoustics are probably as important as the speaker here.

I guess again, I'm a lucky chap with my designs, they have good dispersion and imaging. The center is purposely controlled to the listening area. This is mainly not to comb filter the mains.

Trouble is, that every designer will give you a different opinion about this. I have just found over the years the balance I like. That is another bonus of DIY. If I am happy, I just leave well alone, and be content.

As an aside if you have a good room and can do a setup to speck, that Atmos technology improves 3D imaging greatly. I have a couple of BDs now in Atmos, and the BPO are now starting to get the hang of it. Having said this, it is not a technology to rush into. It is complex and requires precision on multiple fronts.

 

[Swerd]

I have recently tested side by side 5 pair of speakers I have and imaging goes from poor to good with most having a very narrow sweet spot for imaging and only 2 having a decent sweet spot before the image breaks up to either speaker on the sides.

I've seen reviews here, at Erin's audio corner and Audio Science Review and don't know which dictates how well imaging will be. As far as I know, it has a lot to do with time alignment with the drivers and crossovers but don't recall a measurement for that.

It certainly does have a lot to do with crossovers, but I don't think time aligning the drivers is needed. At least, time alignment isn't required to create good imaging. I've heard plenty of speakers that imaged well but were not built with time alignment.

I don't think there's a single measurement for imaging. But it does show up in graphs of a speaker's loudness vs. frequency response made when the test microphone is on-axis (0° off-axis) as well as a number of angles off-axis, such as 30°, 45°, 60°, 75°, and 90°.

As far as I understand things, a speaker's ability to create a good image (in stereo) depends on dispersing sound as widely as possible. I'm talking about speakers where sound emanates from drivers mounted on the front of a cabinet, a monopole speaker. Not a bi-polar design where sound emanates from both front and rear. A number of different features in speaker design can affect this:

• In general, a driver disperses sound widely at frequencies where the wavelength is larger than or equal to its diameter. At wavelengths shorter than the driver's diameter, the driver begins to beam sound instead of dispersing it widely.
• Is the speaker 2-way or 3-way?
• What is the diameter of certain drivers in the speaker? In a 2-way, what is the mid-woofer's diameter? And in a 3-way, what are the woofer and mid-range diameters?
• At what frequency is each crossover point?
• The designer should carefully select crossover frequencies so they prevent the mid-woofer (in a 2-way), or the woofer and mid-range (in a 3-way) from operating at frequencies where they beam sound in a narrow pattern.

This topic was discussed a few years ago on AH. See this link:

Controlling speaker directivity

Can someone point me to a good source of information for learning to understand and manipulate this through crossover design? I think this will help broaden my understanding of speaker design, and maybe put a few other things into context.

forums.audioholics.com

 

[Eppie]

There is no measurement that is specific to imaging so there is no imaging rating per se. Erin describes it pretty well in some of his videos, where he goes on to explain what the measurements mean and how they affect the sound profile. In general, speakers with narrow dispersion tend to have tighter imaging, or placement of objects. They also tend to have a smaller sweet spot as the response varies more off axis. Speakers with a wide dispersion will tend to yield a wider sound stage as they interact with the side walls more and some will say that this leads to less precise imaging. Those are some of the general objective comments I have heard.

This is probably easiest seen from the polar map where a large even red area indicates a wide dispersion speaker. The Philharmonic BMR is a good example of a wide dispersion speaker with very even off axis response. You can also look at the phase angle graphs which are often overlayed with the speaker impedance graph. I'll refer to Philharmonic again as the recently reviewed BMR Tower HT has a nicely behaved phase angle.

It also depends on other factors though. Are you listening near field or at a distance? How close are the speakers to the side walls? What is the vertical dispersion of the speaker like and how are the floors and ceiling interacting with the sound? Were you always aiming the speakers directly at you or did you try different degrees of toe-out or toe-in. Extreme toe-in for example can improve imaging for those sitting off to the side. If the speakers are facing straight ahead then naturally sitting to the side in front of one speaker will make the speaker more predominant.

 

[Matthew J Poes]

What people call good imaging is really a very vague description and the research correlating Spenser parameters and perceived imaging isn’t that good.

instead we can point to some specific attributes and the characteristics that go along with that. However the speaker and room both contribute to this. One can alter the other.

envelopment is the sense of being surrounded by the soundscape. The notion of feeling there at the live musical event. Envelopment is largely a result of strong lateral reflections with good phase randomization. Speakers can’t randomize the phase of reflections. That’s a room property. They can direct a lot of sound toward the side walls. But the side walls could absorb or redirect those. So a wide dispersion speaker and bare parallel walls tends to give very good envelopment and spaciousness. The latter is a related parameter.

imaging or the ability to locate sound objects on a point in space from multiple speakers is a result of a trick. Normally we locate by direct sound sources. A discrete source gives the ear an ITD and ILD signal that we use to locate it. Add a visual cue and we are mostly set. HRTF is often cited but has almost no impact on imaging in the lateral plane. Only vertical.

a stereo system reproduces these cues by giving the brain two cues with offset timing and level. The Brian splits the difference and that creates phantom images. We only do that with ITD and ILD and is the brain receives co flouting ITD and ILD, it will usually go to ITD cues over ILD.

what that means is that object placement is a result of strong direct signals from the speaker. Conflicting early reflection cues from walls wash that out. As such we get sharper and more precise imaging from narrow dispersion, at the expense of envelopment. We get vague imaging but better envelopment from wide dispersion.

now what about polar response? Well a bad polar response will mostly impact timbre, not imaging. But if the directivity collapses as it goes higher and becomes very narrow you can get both strong coloration and weird shifts in imaging. At least in theory. In practice most people don’t seem to notice this since this is how B&Ws operate and people often cite amazing imaging.

reality is this wasn’t studied that much. Toole did a low budget small study comparing directivity and preference. It favored wider dispersion but it wasn’t really a good study of directivity preference and it wasn’t a study of imaging. Some others did studies of imaging accuracy for surround and found narrow dispersion was preferable. It’s also preferred for technical reasons with really high tech surround systems like Ambisonics (the narrower the better in fact). I kind of wish some better studies were done with this. I don’t think we know that much about how this all works in small rooms. It’s mostly conjecture.

lastly I would say that pair consistency is probably helpful. It would make sense that if a speaker differs from its mate too much that this could impact imaging accuracy in some way. This is conjecture on my part, but I could imagine certain cues being harder to track accurately depending on what their bandwidth is if the speakers are very different. There are some tracking perceptions tied to changes in Timbre. That’s how HRTF works. The brain memorizes the error and when it detects specific shifts in timbre it knows the object must be moving in a specific direction.

 

[Mike Up]

Thanks everyone! Thanks for the advise in what to look at. Doesn't sound like there is really just one measurement and even then, it's only a starting point.

I guess you have to audition the speakers to really hear how they image obviously. Problem in today's world there is no longer audition rooms unless you are lucky enough to find a Best Buy but their rooms are poorly set up. Plus many internet companies want to charge you to audition their speakers, through return delivery fees and some even have restocking fees.

 

[lovinthehd]

Thanks everyone! Thanks for the advise in what to look at. Doesn't sound like there is really just one measurement and even then, it's only a starting point.

I guess you have to audition the speakers to really hear how they image obviously. Problem in today's world there is no longer audition rooms unless you are lucky enough to find a Best Buy but their rooms are poorly set up. Plus many internet companies want to charge you to audition their speakers, through return delivery fees and some even have restocking fees.

Your room itself has a great deal to do with it (as well as placement within the room) so auditioning isn't always going to be a great route (i.e. in other rooms and perhaps with limited placement opportunities). You should be able to get good imaging with a variety of speakers, though.

 

[Swerd]

For what it's worth, I have first hand experience with several small internet speaker makers whose products create images quite well. They are Philharmonic Audio and Ascend Acoustics.

I've heard these Philharmonic Audio models, the BMR Monitor, the BMR Tower. In my opinion, they're the champions among commercially available speakers at creating good stereo images and convincing soundstage. Their designer, Dennis Murphy is well-known for his ability & efforts at this. Click on the links for those speakers and you'll see a whole series of on- and off-axis frequency response curves. They've also been reviewed on Audioholics where there are complete measurements confirming what Philharmonic shows.

I've heard the more recent Phiharmonic Audio HT Tower but not under good enough conditions as the other two. It was a single prototype – could not listen in stereo. I also haven't heard either of those two smaller 2-way speakers, but knowing who designed them, I don't doubt that they are also very good at imaging.

The Ascend Acoustics model I'm familiar with is the Sierra. It's also excellent at imaging. Probably the whole Sierra series is good at this.

 

[MrBoat]

To the contrary, I have found excellent imaging to be pretty much inescapable with every speaker I own. Figured it must have something to do with the knowns of computer sim'd designs. The rather budget JBL S38 that I got from a forum member here image like crazy. So much so, coupled with relatively low audible distortion, makes them much more valuable to me than their price would suggest. All of the speakers have the star image directly in the center to where you can reach out and touch it. Some people actually reach out and try to.

All of my waveguide speakers, whether it be multiple pairs of JBL, the Tempests with the SEOS waveguides, and even the Classix 2 budget kit really throws an immersive image. I thought perhaps it must be even the most subtle waveguided tweeters causing it, since most of my speakers have that in common.

Only thing I do different than most other people now perhaps being, I don't use any auto room correction and I use speaker positioning to set up a sweet spot where I want it, which is much closer than most people like theirs. You would think being point blank like this would make it more difficult, but it is still intact even 4-5 ft out from the drivers on my large system, and as little as 2ft with the table top systems I set up frequently. This is in great part why I rarely look for upgrades. With such neutral, immersive, and low distortion setups, what else is left?

 

[Mike Up]

I only have "budget speakers" but they most were considered to be good speakers previously. The speakers I have compared back to back are the Infinity Primus P143, Infinity Entra Point Five, Elac Debut 2.0 B6.2, Polk Blackstone TL1 and Polk TSI100. I have them all but just gave my step son the Polk TSI100 bookshelf speakers and matching CS10 center channel speaker.

The best imaging speakers by far are the Elac Debut 2.0 B6.2, followed by the tiny Polk Blackstone TL1s, substantially worse is the Polk TSI100, and the worst imaging speakers were a tie between the Infinity Primus P143 and Infinity Entra Point 5s.

The best speakers were the Elac B6.2s and Tiny Polk TL1s as these didn't really show loudness compression as their voice stayed the same at low and high volumes. The rest became quite shouty at loud levels even though they are much bigger than than the little Polk TL1s.

The Polk TL1s and the Elac B6.2s also sounded realistic on acoustic guitar and had the best tonality over the rest.

I think it was a toss up between the TSI100 and Infinity P143. While imaging on the Infinity was worse, it was slightly less bright, better tonality and sounded more realistic on acoustic guitar. The Polk was louder, and newer also.

 

[Mike Up]

I should point out all image great when you are on axis and disappear. The differences are the size of the axis sweet spot and how quickly the image shifts to either of the side speakers when you are off axis. A good imaging speaker, as the Elac, will keep the center image between the speakers even if you're slightly off axis.

 

[Mike Up]

Only thing I do different than most other people now perhaps being, I don't use any auto room correction and I use speaker positioning to set up a sweet spot where I want it

I don't use room correction either, I hate that Audyssey XT and normal Audyssey on my Denon receivers, it just destroys the music with a harsh, distorted sound character. Plus after several tries, the settings are always screwed up and wrong. After I disable Audyssey, the sweet sound comes back after my manual setup.

I do tow in my speakers so their center axis' line up at my listening spot. I get the most focused imaging that way.

 

[MrBoat]

On the SEOS waveguide speakers, you can go face up to either speaker and still not have it show it's position with regard to left vs right, at least not without really focusing on it. It's the most uncanny thing. I discovered it right after I first assembled them and was trying to do initial assessments of the build. This was my first waveguide speaker so I didn't know what to expect. I actually like to listen to my individual speakers at times. Sometimes I want to hear how a lone woofer is performing without the support from the other.

I can even set them up like this and stand in between them and get an arms reach image at the full 180 relative to front and sides, which may as well be infinite distance. This trait also tends to keep me from going surround, or even a center. The subs fill in the rest of that 3rd dimension.

I suppose in my case, it's more of a headphonic experience, than a stage about the room. This was just to test my newly refurbed amp but I was still taken by the incredible imaging across multiple pairs of speakers, and me just standing getting hammered between them grinning like a possum eating sandspurs. From here, my living room is almost no challenge comparatively, pretty much relying on how much I toe them in/out.

 

[TLS Guy]

What people call good imaging is really a very vague description and the research correlating Spenser parameters and perceived imaging isn’t that good.

instead we can point to some specific attributes and the characteristics that go along with that. However the speaker and room both contribute to this. One can alter the other.

envelopment is the sense of being surrounded by the soundscape. The notion of feeling there at the live musical event. Envelopment is largely a result of strong lateral reflections with good phase randomization. Speakers can’t randomize the phase of reflections. That’s a room property. They can direct a lot of sound toward the side walls. But the side walls could absorb or redirect those. So a wide dispersion speaker and bare parallel walls tends to give very good envelopment and spaciousness. The latter is a related parameter.

imaging or the ability to locate sound objects on a point in space from multiple speakers is a result of a trick. Normally we locate by direct sound sources. A discrete source gives the ear an ITD and ILD signal that we use to locate it. Add a visual cue and we are mostly set. HRTF is often cited but has almost no impact on imaging in the lateral plane. Only vertical.

a stereo system reproduces these cues by giving the brain two cues with offset timing and level. The Brian splits the difference and that creates phantom images. We only do that with ITD and ILD and is the brain receives co flouting ITD and ILD, it will usually go to ITD cues over ILD.

what that means is that object placement is a result of strong direct signals from the speaker. Conflicting early reflection cues from walls wash that out. As such we get sharper and more precise imaging from narrow dispersion, at the expense of envelopment. We get vague imaging but better envelopment from wide dispersion.

now what about polar response? Well a bad polar response will mostly impact timbre, not imaging. But if the directivity collapses as it goes higher and becomes very narrow you can get both strong coloration and weird shifts in imaging. At least in theory. In practice most people don’t seem to notice this since this is how B&Ws operate and people often cite amazing imaging.

reality is this wasn’t studied that much. Toole did a low budget small study comparing directivity and preference. It favored wider dispersion but it wasn’t really a good study of directivity preference and it wasn’t a study of imaging. Some others did studies of imaging accuracy for surround and found narrow dispersion was preferable. It’s also preferred for technical reasons with really high tech surround systems like Ambisonics (the narrower the better in fact). I kind of wish some better studies were done with this. I don’t think we know that much about how this all works in small rooms. It’s mostly conjecture.

lastly I would say that pair consistency is probably helpful. It would make sense that if a speaker differs from its mate too much that this could impact imaging accuracy in some way. This is conjecture on my part, but I could imagine certain cues being harder to track accurately depending on what their bandwidth is if the speakers are very different. There are some tracking perceptions tied to changes in Timbre. That’s how HRTF works. The brain memorizes the error and when it detects specific shifts in timbre it knows the object must be moving in a specific direction.

I think we are in pretty close agreement. I do think however that your choice of program does influence this to a degree.

For one thing in the popular domain the ambience is largely artificial. In addition the productions are so often a series of panned mono sources. Not only that, but when I analyze them in WaveLab, they are center centric. The dominance of the center information compared to the right and left information is striking.

In the classical domain, you really want to hear the space. You are looking to a much more immersive sound. In my view this definitely favors accurate wide dispersion speakers, with very flat FR. I am amazed at how with Atmos and the new up-mixers, you really can hear the space. The BBC transmission of the coronation was amazing. I heard they were experimenting with mic position for days. Anyhow it really showed off the upmixer. You really could hear that the RAF trumpeters were in a high side balcony for instance. It was incredibly close to being in the Abbey. The concert prior on the Abbey that was only on Radio 3 was also astounding.

Another case in point was a concert on Medici TV, with the La Scala orchestra in the palace of Charles V in the Alhambra. That was very much a you are there experience, with the acoustics captured in a totally uncanny manner.

The BPO are getting the hang of their Atmos recordings. I have seen on occasions a modified Decca tree above the orchestra. There was a Mozart evening recently that was stunningly recorded giving a you are there rendition.

So things are improving, but as it does so, realizing the benefits becomes increasingly complex and therefore costly.

By the way thanks for that Trinnov Waveforming Technology Explained with Arnaud Laborie Video. That was outstanding, and I started a thread about it.

 

[highfigh]

It certainly does have a lot to do with crossovers, but I don't think time aligning the drivers is needed. At least, time alignment isn't required to create good imaging. I've heard plenty of speakers that imaged well but were not built with time alignment.

I don't think there's a single measurement for imaging. But it does show up in graphs of a speaker's loudness vs. frequency response made when the test microphone is on-axis (0° off-axis) as well as a number of angles off-axis, such as 30°, 45°, 60°, 75°, and 90°.

As far as I understand things, a speaker's ability to create a good image (in stereo) depends on dispersing sound as widely as possible. I'm talking about speakers where sound emanates from drivers mounted on the front of a cabinet, a monopole speaker. Not a bi-polar design where sound emanates from both front and rear. A number of different features in speaker design can affect this:

• In general, a driver disperses sound widely at frequencies where the wavelength is larger than or equal to its diameter. At wavelengths shorter than the driver's diameter, the driver begins to beam sound instead of dispersing it widely.
• Is the speaker 2-way or 3-way?
• What is the diameter of certain drivers in the speaker? In a 2-way, what is the mid-woofer's diameter? And in a 3-way, what are the woofer and mid-range diameters?
• At what frequency is each crossover point?
• The designer should carefully select crossover frequencies so they prevent the mid-woofer (in a 2-way), or the woofer and mid-range (in a 3-way) from operating at frequencies where they beam sound in a narrow pattern.

This topic was discussed a few years ago on AH. See this link:

Controlling speaker directivity

Can someone point me to a good source of information for learning to understand and manipulate this through crossover design? I think this will help broaden my understanding of speaker design, and maybe put a few other things into context.

forums.audioholics.com

How would time aligning not be important, especially in the crossover region? Slope is often used to 'steer' the output between the mid and tweeter in order to prevent issues with overlap, e.g., asymmetrical crossover design. If the drivers on either side of the crossover point aren't time-aligned, phase shift will be worse, assuming equal slope.

 

[highfigh]

I think we are in pretty close agreement. I do think however that your choice of program does influence this to a degree.

For one thing in the popular domain the ambience is largely artificial. In addition the productions are so often a series of panned mono sources. Not only that, but when I analyze them in WaveLab, they are center centric. The dominance of the center information compared to the right and left information is striking.

In the classical domain, you really want to hear the space. You are looking to a much more immersive sound. In my view this definitely favors accurate wide dispersion speakers, with very flat FR. I am amazed at how with Atmos and the new up-mixers, you really can hear the space. The BBC transmission of the coronation was amazing. I heard they were experimenting with mic position for days. Anyhow it really showed off the upmixer. You really could hear that the RAF trumpeters were in a high side balcony for instance. It was incredibly close to being in the Abbey. The concert prior on the Abbey that was only on Radio 3 was also astounding.

Another case in point was a concert on Medici TV, with the La Scala orchestra in the palace of Charles V in the Alhambra. That was very much a you are there experience, with the acoustics captured in a totally uncanny manner.

The BPO are getting the hang of their Atmos recordings. I have seen on occasions a modified Decca tree above the orchestra. There was a Mozart evening recently that was stunningly recorded giving a you are there rendition.

So things are improving, but as it does so, realizing the benefits becomes increasingly complex and therefore costly.

By the way thanks for that Trinnov Waveforming Technology Explained with Arnaud Laborie Video. That was outstanding, and I started a thread about it.

What mic techniques do they use for symphony recordings nowadays? I know they used figure of eight often, but are they close-micing many instruments or are they using more overheads?

What mics are most common, Nuemann, AKG, Schoeps, Telefunken...?

 

[TLS Guy]

What mic techniques do they use for symphony recordings nowadays? I know they used figure of eight often, but are they close-micing many instruments or are they using more overheads?

What mics are most common, Nuemann, AKG, Schoeps, Telefunken...?

Telefunken became Neuman about half a century ago, otherwise all of the above. The BBC do se some Coles ribbons I note still, that is the same as the old STC ribbons of he fifties and sixties. Coles bought that STC division.

Coincident techniques are not common, but should be. The spaced omnis with spots is still common. I'm noting a big resurgence of the Decca tree. I note the DSO streamed concerts are done pretty much exclusively with the Decca tree. I note a big reduction in the use of spot mics of late, which is a good thing. Having said that, I still see mics thrown about all over the place, which is bad technique. The BPO were ding that a few years back, but the mic count has gone down markedly over time.
You can not make a realistic recording peppering mics all over the place.

 

[highfigh]

Telefunken became Neuman about half a century ago, otherwise all of the above. The BBC do se some Coles ribbons I note still, that is the same as the old STC ribbons of he fifties and sixties. Coles bought that STC division.

Coincident techniques are not common, but should be. The spaced omnis with spots is still common. I'm noting a big resurgence of the Decca tree. I note the DSO streamed concerts are done pretty much exclusively with the Decca tree. I note a big reduction in the use of spot mics of late, which is a good thing. Having said that, I still see mics thrown about all over the place, which is bad technique. The BPO were ding that a few years back, but the mic count has gone down markedly over time.
You can not make a realistic recording peppering mics all over the place.

I only recorded two live performances- one was a demo for a band and the other was a live gig- the demo wasn't going to sound great, but it was more important for the band's playing to come through and the second wasn't going to sound great because of the club's acoustics. However, it sounded very realistic- I used two mics and listened to the sound before pressing Record and playback was a nice surprise because it was like being there. I used the new (at the time) Sony PCM-F1, so background noise was not an issue, at all.

 

[Swerd]

How would time aligning not be important, especially in the crossover region? Slope is often used to 'steer' the output between the mid and tweeter in order to prevent issues with overlap, e.g., asymmetrical crossover design. If the drivers on either side of the crossover point aren't time-aligned, phase shift will be worse, assuming equal slope.

I have heard Vandersteen 3A speakers, which are carefully built with time-aligned drivers and use carefully designed 1st order acoustic (not electrical) crossover slopes. They can cast wonderful sounding images, but only over a very narrow range. Even if I sat in the center of a sofa, if I slightly shifted my head to a different position, the image could vanish. I wondered where I could find an audiophile-grade head vise .

Not many commercial speakers have been built with cabinets that allow physical time-alignment, and I've only heard the Vandersteens. The Vandersteen crossovers were carefully designed to have 1st order crossover slopes when acoustically measured. This was done to achieve time and phase alignment of drivers They are far from the simple 1st order electrical crossovers commonly seen. Vandersteens have many more parts in a complex arrangement. I've seen photos of them, but never did see a schematic.

My old original speakers, JBL L-100As, designed without anything resembling modern speaker or crossover design, could cast images, only if I sat in the middle of my sofa, only with enough tow-in, and only if the volume was high enough. FWIW, they did have extremely simple 1st order (electrical) crossovers.

When I first heard some Dennis Murphy-designed speakers, they created images that astonished me at the time. They imaged over a much wider range than Vandersteens, they lacked any physical time alignment of drivers in the cabinet, and used 4th order Linkwitz-Riley crossovers. All drivers on either side of a 4th order L-R crossover are in phase with each other, but out of time by 360°, one full wavelength at crossover frequency. Dennis Murphy strives to modify the slopes on either side of a crossover so that the roll-off slopes are as symmetric as possible, given the drivers. (He also carefully selects drivers for this same reason.) He believes his efforts in phase alignment pay off, but time alignment makes little, if any difference, that listeners can hear.

As the OP @Mike Up said earlier in this thread, all of his speakers can create images with enough care in speaker placement & listening position. Some could do it only over a narrow area, and some could do it over a much wider area.

I should point out all image great when you are on axis and disappear. The differences are the size of the axis sweet spot and how quickly the image shifts to either of the side speakers when you are off axis. A good imaging speaker, as the Elac, will keep the center image between the speakers even if you're slightly off axis.

To invent an imaging metric, imagine how wide a seating area – how many sofa cushions wide – would allow a listener to hear images with a pair of speakers. Instead of sweet spot, let's call it the Sofa Cushion (SC) rating .

If those Vandersteen 3As had a SC rating of <1 (less than one sofa cushion wide), the JBL L-100As (with enough tow-in & enough listening volume) had a SC rating of about 1. When I replaced them with Salk SongTowers, their SC rating was at least 3 (my sofa is not wider), in the same room, with the same music. And good imaging was created regardless of listening volume. Obviously, time-alignment – alone – is not the key.

 

[highfigh]

I have heard Vandersteen 3A speakers, which are carefully built with time-aligned drivers and use carefully designed 1st order acoustic (not electrical) crossover slopes. They can cast wonderful sounding images, but only over a very narrow range. Even if I sat in the center of a sofa, if I slightly shifted my head to a different position, the image could vanish. I wondered where I could find an audiophile-grade head vise .

Not many commercial speakers have been built with cabinets that allow physical time-alignment, and I've only heard the Vandersteens. The Vandersteen crossovers were carefully designed to have 1st order crossover slopes when acoustically measured. This was done to achieve time and phase alignment of drivers They are far from the simple 1st order electrical crossovers commonly seen. Vandersteens have many more parts in a complex arrangement. I've seen photos of them, but never did see a schematic.

My old original speakers, JBL L-100As, designed without anything resembling modern speaker or crossover design, could cast images, only if I sat in the middle of my sofa, only with enough tow-in, and only if the volume was high enough. FWIW, they did have extremely simple 1st order (electrical) crossovers.

When I first heard some Dennis Murphy-designed speakers, they created images that astonished me at the time. They imaged over a much wider range than Vandersteens, they lacked any physical time alignment of drivers in the cabinet, and used 4th order Linkwitz-Riley crossovers. All drivers on either side of a 4th order L-R crossover are in phase with each other, but out of time by 360°, one full wavelength at crossover frequency. Dennis Murphy strives to modify the slopes on either side of a crossover so that the roll-off slopes are as symmetric as possible, given the drivers. (He also carefully selects drivers for this same reason.) He believes his efforts in phase alignment pay off, but time alignment makes little, if any difference, that listeners can hear.

As the OP @Mike Up said earlier in this thread, all of his speakers can create images with enough care in speaker placement & listening position. Some could do it only over a narrow area, and some could do it over a much wider area.To invent an imaging metric, imagine how wide a seating area – how many sofa cushions wide – would allow a listener to hear images with a pair of speakers. Instead of sweet spot, let's call it the Sofa Cushion (SC) rating .

If those Vandersteen 3As had a SC rating of <1 (less than one sofa cushion wide), the JBL L-100As (with enough tow-in & enough listening volume) had a SC rating of about 1. When I replaced them with Salk SongTowers, their SC rating was at least 3 (my sofa is not wider), in the same room, with the same music. And good imaging was created regardless of listening volume. Obviously, time-alignment – alone – is not the key.

Well, that's the thing about time alignment- it works for a limited space and only for that space. There's not a great way to cover a large area AND have the sound from various drivers reach the ears at the correct time. Thiel speakers are/were time-aligned, too- they don't sound like most speakers but again, the listener can't expect to move around the room and hear the same sound at all locations.

The part about carefully selecting drivers and other components- some companies don't seem to be as interested in sound quality when the bean counters pipe up with "That's good enough!".