What determines imaging?

[Joe Schmoe]

I have noticed that, even among monopoles, some speakers have significantly better imaging/soundstage than others. Since virtually all modern speakers have a 1" dome tweeter, and most have a narrow baffle with rounded edges, this seems counerintuitive. What are the design factors that contribute to imaging? Do bookshelves always image better than towers, and if so why? Also, what are some makes/models that are particularly outstanding in this regard?

 

[fmw]

I have noticed that, even among monopoles, some speakers have significantly better imaging/soundstage than others. Since virtually all modern speakers have a 1" dome tweeter, and most have a narrow baffle with rounded edges, this seems counerintuitive. What are the design factors that contribute to imaging? Do bookshelves always image better than towers, and if so why? Also, what are some makes/models that are particularly outstanding in this regard?

Imaging results from reflections of the sound from nearby objects. It is the audiophile way a defining room acoustics. The contents of the room and the speaker placement control it. I don't believe speaker systems themselves have such a parameter in an anechoic chamber. If you've heard imaging differences in speaker systems themselves I would suggest that the differences are more likely in frequency response or distortion or speaker placement. All of those things will change the behavior of the sound in its acoustic environment. I think trying to define "imaging" without both the speaker and the acoustic environment is probably meaningless.

Why would baffle edges and dome tweeters seem counterintuitive?

 

[3db]

I have noticed that, even among monopoles, some speakers have significantly better imaging/soundstage than others. Since virtually all modern speakers have a 1" dome tweeter, and most have a narrow baffle with rounded edges, this seems counerintuitive. What are the design factors that contribute to imaging? Do bookshelves always image better than towers, and if so why? Also, what are some makes/models that are particularly outstanding in this regard?

Its a good question.. I think imaging has to do with the relative phase and amplitude of the signal, relative meaning between left and right speaker. From that, I'm thinking a speakers crossover plays an important role from a phase point of view in how well a speaker images given that speakers have narrow baffle design. Remember the THX optomizer on some DVDs to check to see if a speaker was wired in or out of phase. I would think it falls the same prinicple.

The question about bookshelves vs towers? Are you speaking of identical quality?

 

[Joe Schmoe]

Why would baffle edges and dome tweeters seem counterintuitive?

I meant that these are two things that could effect imaging, but that most speakers have them in common and should therefore not image differently.

 

[Joe Schmoe]

The question about bookshelves vs towers? Are you speaking of identical quality?

I don't know how one would determine "identical quality" in such a comparison.
It is a common belief (though not necessarily true) that "small bookshelf monitors have superior imaging".
I have considered placing my speakers very far from the walls (meaning a minimum of 3' in any direction) to see how this effects imaging. Unfortunately, this is rather inconvenient, and also likely to reduce the bass response.

 

[3db]

I don't know how one would determine "identical quality" in such a comparison.
It is a common belief (though not necessarily true) that "small bookshelf monitors have superior imaging".
I have considered placing my speakers very far from the walls (meaning a minimum of 3' in any direction) to see how this effects imaging. Unfortunately, this is rather inconvenient, and also likely to reduce the bass response.

For identical quality, I meant from the same series of the manufacturer; ie PSB Image series, Paradigm's monitor series... comparing within the series.

It would be unfair to compare say my PSB Image T45s to PSB's Alpha series bookshelfs or Paradigm's atoms. But it would be fair to compare the ImageT45s to PSB's Monitor bookshelf or PSBs Image bookshelves.

 

[fmw]

I meant that these are two things that could effect imaging, but that most speakers have them in common and should therefore not image differently.

What these things affect is dispersion not "imaging." You can measure dispersion. You can't measure "imaging" unless you define it as the same thing as dispersion. The term is too tied up in the acoustic environment in which the speakers operate. Perhaps we need to get down to what the definition of imaging is. Perhaps I'm not clear on what you meant by the term in the first place.

 

[Joe Schmoe]

Perhaps I'm not clear on what you meant by the term in the first place.

1) The sound extends beyond the speakers left and right, and sometimes even beyond the walls of the room.
2) Sounds move up/down and front/back as well as side/side. (Occasionally, some sounds will even appear to originate behind the listener.)
3) A large instrument (eg a drum kit) has a 3D "shape" within the soundstage.
4) Sounds with a specific point of origin have a precise location (and don't drift around unless it is intentional.)
5) Everything seems to be taking place within the original recording venue, not the listening room (especially live recordings.)

All of these things together make up what I mean by soundstage/imaging. When I upgrade, it is usually an improvement in one or more of these areas that I am seeking. (All speakers achieve them to some extent, of course.)

 

[fmw]

That's what I thought you meant. My word for it is room acoustics. It is not a characteristic of speakers. It is characteristic of speaker performance within an acoustical environment. The "imaging" one would hear in one room could be very different from that heard from the same speakers in a different room. And, yes, it could vary from one speaker to another in the same room because of differences in frequency response and distortion. These things won't exist as you describe them in an anechoic chamber.

 

[Joe Schmoe]

Seated between the speakers in an anechoic chamber, would you not hear imaging because of the stereo effect, much like when listening to headphones? (I thought that the goal of room treatment was to approximate conditions in an anechoic chamber, atleast to some extent.)

 

[3db]

That's what I thought you meant. My word for it is room acoustics. It is not a characteristic of speakers. It is characteristic of speaker performance within an acoustical environment. The "imaging" one would hear in one room could be very different from that heard from the same speakers in a different room. And, yes, it could vary from one speaker to another in the same room because of differences in frequency response and distortion. These things won't exist as you describe them in an anechoic chamber.

So you won't hear stereo in an anoechic chamber?

 

[3db]

That's what I thought you meant. My word for it is room acoustics. It is not a characteristic of speakers. It is characteristic of speaker performance within an acoustical environment. The "imaging" one would hear in one room could be very different from that heard from the same speakers in a different room. And, yes, it could vary from one speaker to another in the same room because of differences in frequency response and distortion. These things won't exist as you describe them in an anechoic chamber.

Hmmmm. What about THX optimizer tests for ensuring your speakers are wired in phase relative to one another. When the sound is in phase, it appears that the sound is coming from between the speakers. When the sound is out of phase, the sound field becomes diffused and its impossible to localize the sound between teh speakers. Surely, that effect isn't primarily room acoustics as it is phasing between the speaker pair.

 

[WmAx]

The 'imaging' as Joe defines it, is highly dependant on 3 primary things, in order of importance: (1) room acoustics (2) symmetry of response between speakers in relation to listener position (3) treble response curve.

(1) As an example, no speaker will have good imaging, if placed in a highly reverberant area 1' near the boundaries, unless that speaker has a specific mechanism to prevent substantial off axis output. Such an example would be a large lens full range horn speaker system. But this is a double edge sword. As, if you place such a speaker in an ideal room acoustic, it will have less of a realistic sound stage effect as compared to a speaker with superior dispersion, all else being equal. The room reflections act as such to provide a more enveloping acoustic effect, if used properly. Of course, too much of anything is never ideal, and appropriate acoustic controls/treatments should be used in any application where optimal sound quality is the primary objective. A speaker should also strive to have as even a frequency response at all dispersion angles as possible. As, if the off axis/reverberant sound field has differences from the direct sound, your brain will detect this is a lowered sound quality, according to the credible perceptual research published to this date.

(2) If the frequency/phase is different between channels, it will affect the stability of the center anchored illusion. Room placement symmetry and angle/distance to the listener must be considered.

(3) Assuming a standard 2 channel stereo set-up: Merely altering treble(example, boost energy over 6khz or cut it over 6khz) will affect the perceived definition of a centered image. Generally, more treble will be perceived as a narrower, more distinct image, where as less treble may be perceived as a duller, wider center image.

-Chris

 

[Joe Schmoe]

(1) As an example, no speaker will have good imaging, if placed in a highly reverberant area 1' near the boundaries

My current speakers (conventional dome tweeters) are 18" from the side walls (and several feet from the wall behind them.) Would moving them even farther from the side walls (hence closer together) improve the imaging, or is 18" enough? If I did move them farther in, how much?

 

[WmAx]

My current speakers (conventional dome tweeters) are 18" from the side walls (and several feet from the wall behind them.) Would moving them even farther from the side walls (hence closer together) improve the imaging, or is 18" enough? If I did move them farther in, how much?

The optimum distance/placement depends on several things. One is the specific off axis behaviour of the speaker in question. Another is the property of the room itself. On most speakers, you can, for example, modify the soundstage and center image definition/stability to an extent by simply changing the angle of the speaker. For example, try towing the speakers in 10 degrees and out 10 degrees. A clearly audible change in the center image should result. As another result, as you tow them out, the standard dome tweeter will now reflect more energy off of the side wall, and if the distance is correct(18" should be sufficient for a side wall), the illusion of sound cues that extends outside of the speakers themselves should increase by a certain degree, assuming the source material is not mono, of course. This would be most apparent in recordings with strong ambiance reverberation effects. If, however, the off axis response is very different from the on axis response, then almost always, a distinct benefit can be had by moving the speaker farther from the side wall(this reduces the amplitude of the first reflection) and/or by using effective absorption panels at these points.

-Chris

 

[Joe Schmoe]

It might be worth experimenting a bit more. Mind you, I have no complaints about the imaging I am hearing now, as it is quite good. I am always in favor of improving it even more if I can, though.
(In all rooms and with all speakers, I tend to prefer a lot of toe-in. That seems to be "hard wired" into my personality, so to speak. In fact, I generally like the baffles facing directly towards my ears, or nearly so.)

 

[krabapple]

It might be worth experimenting a bit more. Mind you, I have no complaints about the imaging I am hearing now, as it is quite good. I am always in favor of improving it even more if I can, though.
(In all rooms and with all speakers, I tend to prefer a lot of toe-in. That seems to be "hard wired" into my personality, so to speak. In fact, I generally like the baffles facing directly towards my ears, or nearly so.)

I would expect this more closely emulates a 'near field' listening environment at the sweet spot -- by toeing the speakers in to aim them at your ears, you are beaming treble directly at them (enhance the treble effect WmAx noted) and reducing ear input from off-axis/first reflection sound. I would guess it enhances imaging between the speakers possibly at some expense of 'outside the speakers' images (though in an actual two-channel near-field setup, I recall hearing an amazing degree of both)

 

[Swerd]

What effects imaging?

Others have already covered:

Room reflections
Speaker toe-in
Baffle size (narrow vs. wide and large vs. small)

In my experience the off-axis dispersion of a speaker, especially across the midrange, has a very large affect on a speaker’s ability to create the illusion of imaging. Obviously, to interact with room boundaries, a speaker must be able to disperse sound widely. The crossover has a large effect on this.

There is an article at Speakerbuilder.net that discusses choosing a crossover frequency so that the woofer’s off axis response is not too low compared to its on axis response. In a 2-way speaker for example, when a woofer produces sound with wavelengths larger than its diameter, the sound is widely dispersed. At wavelengths near the woofer’s diameter and shorter, the sound tends to beam directly ahead. This can be seen in the frequency response curves of a woofer that show several curves measured on-axis and off-axis. If the crossover point is low enough so that the off-axis response is not lower than about 3 dB than the on-axis response, then the whole speaker will have better off-axis performance through the critical midrange.

 

[fmw]

Hmmmm. What about THX optimizer tests for ensuring your speakers are wired in phase relative to one another. When the sound is in phase, it appears that the sound is coming from between the speakers. When the sound is out of phase, the sound field becomes diffused and its impossible to localize the sound between teh speakers. Surely, that effect isn't primarily room acoustics as it is phasing between the speaker pair.

Yes, but I don't believe that is what the OP was talking about.

 

[Joe Schmoe]

I am going to try moving them an additional 6" in from the side walls, just to see what difference (if any) it makes. Of course, this will also require moving them forward to maintain an equilateral triangle. If I like what I hear, I will go with it (though having them that far out into the room is a bit visually awkward.)