Home Theater Forum and Systems banner

How do you define good Imaging? and good Soundstage?

18K views 46 replies 18 participants last post by  JoeESP9 
#1 ·
In another thread recently there was some discussion about imaging and the definition of good imaging. I had always assumed there was only one way of looking at it, but of course, as with most things, that is not true.

So now I am really curious to hear how others define good imaging. What is it? How do you define it? How do you describe it? How do you achieve it? And since imaging and soundstage seem to go hand in hand, let's throw it in there, too.

There is a whole list of important qualities that come to mind for me, but here is my first one:

Good imaging is convincingly natural, like the vocalist or the instrument is right there in the room with you, reach-out-and-touch real.

What is your definition/description?
 
#2 ·
Right or wrong, here's how I look at it in very general terms.

Soundstage is the perception of sound sources located in three dimensions, beyond the plane(s) of the physical speaker drivers, i.e. near/far, wide, deep, behind, in front, etc...

Imaging has to do with how focused each source is, i.e. can you pinpoint it's location if, in reality, it should be originating from a single point?
 
#3 ·
Soundstaging is one of those elusive definitions, because it requires everything to be just so to be convincing. By everything, I am talking about proper toe-in of the loudspeaker for localization and size of the image, correct distance for sense of scale, timbral accuracy, proper room treatment for image specificity and on and on. When all of this is correct, the soundstage will be 180 degrees... i.e. wall to wall, depth beyond the wall behind the speakers (front wall) with individual instruments and performers in their own defined space with "air" (hard to describe, but when you hear it you know it) surrounding them. Instruments should be proper in size and fixed in space with no wander. All of this is dependent upon the recording obviously. Many modern pop recordings use digital processing that messes mightily with phase to get a sense of space, so beware. QSound and other spatial correlators will mess with your head when the image is proper. Check out Suzanne Ciani's recordings, Pink Floyd etc.

Ain't easy but fun when we get there. Here are a couple of pictures of my quickie listening room setup right after we moved into this house a few years ago. Had to have the toons. I threw up the corner and side wall treatments to make the room work. The room size is really good for mode distribution. That was one of the reasons we bought this "House in a Kit". Right size, right place, needed remodeling. I was up to it a couple of years ago. Not so much anymore, but I'm getting there. Even with all of the junk in the room, including the R/C HelicopterMan, this space works with minimal treatment. Wall to wall, 10' deep and no image wander. This room supports that "they are here" realism that we all seek.

Rooms don't need a lot of treatment for 2-channel if the dimensions are right. I see too many over treated rooms that just suck the life out of the recordings.

This is not my present system, either. My current system is a pair of GR-Research Kinda SuperV's ( I can't leave anything alone), Dodd Audio Buffer and 15WPC Stereo Amp (see pic), Mac Mini music server and cabling by me. I am currently slowly doing the remodel of the room that I wanted to do a couple of years ago.
 

Attachments

#13 ·
...When all of this is correct, the soundstage will be 180 degrees... i.e. wall to wall, depth beyond the wall behind the speakers (front wall) with individual instruments and performers in their own defined space with "air" (hard to describe, but when you hear it you know it) surrounding them...
...

Rooms don't need a lot of treatment for 2-channel if the dimensions are right. I see too many over treated rooms that just suck the life out of the recordings.
Hi dBe,

Do you happen to have a top-view sketch of your listening room ? Could you post that, putting on top the sound sources of a recording? Suzanne Ciani, Pink Floyd or whatever suits you best : )

What I hope for is somth like , "this is the room, that is the position of the speakers inside the room, and in this recording of XYZ here and there and overthere is where the musicians -or other creatures- 'live' ..."

Particularly interested in the "beyond the (front) wall" presentation, never could get that sense of the walls "not being there". And I've tried various speaker positions in empty / furnished / absorption-panel treated rooms...

Thanks,

Charlie
 
#4 ·
While I agree that the near perfect room acoustics will improve the imaging and soundstage, The biggest way to achieve great imaging is with well made speakers. Ive had many speakers in my lifetime and non have ever sounded as good for imaging as my current EVs. A speaker thats got good clean natural mids and highs that have good off axes response is a must.

Good recordings that have left and right imaging is so important. I have recordings that are amazing to say the least and have to be heard to believe. The issue is that a bad recording sounds really bad with my speakers so thats the downside in a way.
To achieve good imaging several things must be in place.
one, the recording master must have been well engineered. EQ at the source for both channels (2ch) needs to be identical particularly in the upper frequencies and there must be good channel separation.
two, Introducing delay and even some reversal of phase can give a very wide sound.
 
#10 ·
This makes a lot of sense to me. Imaging starts with the recording, mixing, & mastering, all with a mind toward preserving the frequency and phase information between channels for a given instrument or sound. Then the speakers have to be designed right to deliver that information and deliver it faithfully. So for good imaging, I would say 40% is recording, 40% is speakers, 20% is room.

Soundstage is more room related, but also depends on a good recording and good speakers.

BTW, we often talk about soundstage being 3-dimensional, and I agree that on occasion the third dimension (vertical) comes into play. But don't we normally mean 2-dimensional, with width and depth?
 
#5 ·
Imaging is easy. Take two speakers feed them a stereo signal, spread them a couple of feet apart and you've got imaging. Imaging is mostly left to right spread and positioning. While any two speakers will give you imaging many will not produce any type of sound stage. Having a sound stage means you also have the depth that makes the sound three dimensional.

Getting a good sound stage is 50% speakers and 50% recording. Some speakers just don't do a very good job producing a sound stage. Studio albums which are multi-tracked, overdubbed and "sweetened" rarely if ever have any sound stage. Classical, Jazz, live recordings and recordings made with a minimal number of microphones are usually the ones with a sound stage.

In short: You can have imaging without a sound stage but not the reverse. Ideally you want both
 
#6 ·
IME the soundstage/imaging thing is 50% room, 30%, recording and 20% speakers.

I've gotten absolutely killer soundstage out of a boombox set up in a good acoustical environment.

Mostly it comes down to personal preference. How much of "they are here" is important to the listener. Some people could care less. For a recording engineer it is paramount. Next is the timbral accuracy and then the effects used.
 
#7 ·
Soundstage (width, depth, height) is the stage the performers are on. It can be large, medium or small, akin to sitting front row, middle or back of the hall at a live performance.

Imaging is where the performers are located on that stage, and the distance between them (air).
 
#9 ·
JoeESP9 said:
Imaging is easy. Take two speakers feed them a stereo signal, spread them a couple of feet apart and you've got imaging. Imaging is mostly left to right spread and positioning. While any two speakers will give you imaging many will not produce any type of sound stage. Having a sound stage means you also have the depth that makes the sound three dimensional.
fmw said:
...soundstage is adjusted with speaker placement and room acoustics. Imaging is adjusted by the recording engineer by using a pan control for each channel during mixing.
Interesting viewpoints, seeming to suggest that imaging is easy to achieve and soundstage is not, if I understand these posts correctly.

Everyone so far seems to agree on one thing, at least implying that imaging and soundstage are related. And from there it becomes a matter of definition, how you divide the listening qualities between one and the other.

I think of imaging as the qualities of a single sound source in the mix, such as the lead singer, for instance. With good imaging, the voice is precisely located, does not wander or smear on different notes or different vocal sounds. My own view is that this is not easy to achieve, in fact is quite a feat. I think of Soundstage as the spacial relationships between the different sounds in the mix, ideally a natural 2-dimensional array that convinces the listener that it is independent of the speakers, each component having its own distinct placement in space. Also not easy to achieve.
 
#15 ·
First off, room treatment is sooooo important for imaging and soundstage. It almost doesn't matter what speakers you have until you treat your room to minimize early reflections.

As for the OP, there is so much variance in this imaging/soundstage idea.

Regarding imaging, some people feel you need to be able to put an orchestral recording on and be able to point to the concertmaster or first clarinetist and say, "He's (she's) right there," and pinpoint the musician's exact location. Others (notably Peter Snell) feel that this is not anything like what you'd experience in the concert hall, so why try to do that? You need a sense of depth and space (soundstage?), but the pinpoint imaging is an audiophool conceit.

There is the side-to-side 'imaging,' but then there's the front-to-back 'depth' (is that 'soundstage'?). Then there's the pinpoint *localization* of sound sources in the recording. (What is that? Also 'Imaging?')

I find that horn speakers (Klipsch, Tannoy, etc.) excel at pinpoint localization and side-to-side imaging, but are sorely lacking in depth and atmospherics. They put the musicians right in yer face, or in your lap. Like, 'they are here.'

Many cone-and-dome speakers diffuse the sound more to create an atmospheric spray of sound, creating that 'wide soundstage.' ('You are there.') I have a pair of Snell Type C (early version) that play a wide, atmospheric soundstage, but don't give you a whole lot of front-to-back 'imaging.' Many of the high-end cone-dome speakers can portray a great sense of depth, but aren't as dynamic as the horn type speakers.

Planar speakers often have spectacular depth (imaging?) but are hard to drive and can sound dynamically constricted. Some are among the best ever designed (like Quad ESL-63, imo).

Speakers are so imperfect that you basically have to decide what you prefer in the sound, and choose accordingly. I guess that's why some prefer Magnepan and others choose Klipsch. The two are wildly different experiences from each other, yet both are 'good.' How can that be?

--
 
#20 ·
First off, room treatment is sooooo important for imaging and soundstage. It almost doesn't matter what speakers you have until you treat your room to minimize early reflections.

As for the OP, there is so much variance in this imaging/soundstage idea.
For sure!! Even in these few posts, all from clearly observant listeners, the definitions and perceptions vary quite a bit.

One exception to your room treatment comment is with near-field monitoring, where speaker frequency and phase response matching seem to carry 80-90 per cent of the weight for good imaging & soundstage. With a more typical arrangement, room treatment is certainly paramount.

Speakers are so imperfect that you basically have to decide what you prefer in the sound, and choose accordingly. I guess that's why some prefer Magnepan and others choose Klipsch. The two are wildly different experiences from each other, yet both are 'good.' How can that be?
Totally agree.
 
#18 · (Edited)
Hi tesseract, very welcome!

I tend to think of the room dimensions much like looking at a drawing board: XYZ = Width, Depth, Height. But yes, the Height component is more of a whim, coming from our hearing mechanism, and speaker-room interaction.

I wonder though, if a loudspeaker has a sloping front bafle -and the drivers are not paralel to the floor|ceiling axis- that would probably increase the ratio of vertical to horizontal reflections and possibly the perception of height in a recording. By what factor, I can not say, just making assumptions here...

dBe, thanks for your interest, looking forward to your sketch!

rongon, thanks for the input, interestin you should mention an orchestral recording! You, and all, know how S.Linkwitz has a strong position on soundstage, or the "auditory scene", as he names it. In his "Hearing spatial detail" paper, he includes this interesting scaled diagram of how the recording of an orchestra is expected to sound in his room...:

Text Pattern Diagram


I attempted a re-scale showing how the same orchestra might be placed if full sized and if the walls in SL's room would actually suddently "drop". And I feel I may have been rather restrained in upscaling .... Keeping S.L.'s assumption that the "... The phantom source will ... essentially not be closer than the distance to the real sources of sound, the two loudspeakers... ", I have placed the first violin at the position of the left loudspeaker : )


post Tree Diagram Illustration


The question therefore arises, is there any kind of room treatment -be that mere absorption/diffusion panels, or better yet overall room design and construction including splayed walls, wideband resonators and similar goodies- that will deliver that much sence of spaciousness inside a small -by measure of the original event- actual living room ? Pray tell, especially if you have achieved this in your listening room :unbelievable: :clap: , or even witnessed it elsewhere : |

Or should we all aggree that listening to the full depth and breadth of an orchestra is to be reserved for either the consert room or for similarly huge living spaces ?

Best,

Charlie

PS. Please note, I am not including the loudness capabilities of the stereo system itself in the equasion. Quoting SL, " . . . Distance and size of the auditory scene are playback volume dependent. The auditory scene moves closer with increasing volume, it becomes larger and more detailed . . ." So, for the benefit of this discussion, let's take as a given that the speakers can deliver 105-110 dB to the listening seat (~ what the front rows in a concert hall may experience? ) without compression.

EDIT : just realized, I typed S.R. instead of S.L. for Siegfried Linkwitz... sorry about that :sad:
also, forgot to thank tesseract back for his kind words, double sorry :sad: :sad:
 
#21 ·
I will read the articles linked to above. There is certainly room for some clarity of definitions, although I have a feeling that most of us will continue using the definitions we are accustomed to.

Also there will be differences in expectation. When I listen to an orchestra, for instance, if I hear a soundstage that seems natural, clear, and separate from the speakers, the actual size, position, perspective relative to the original recorded area are of little concern to me. Of course, now I will be listening in a different way. The thought of being able to "tune" the soundstage to a certain width seems like a stretch, but a fascinating possibility to consider.

Which room is more likely to have a wide soundstage, a live-but-well-controlled room or a dead room?
 
#22 ·
As I understand it, the brain uses sound reflections to make an estimate of the size of a certain: if reflections arive later rather than sooner, and from all over the place, the brain sais "I am now in a large space". So, I would say between the two choices, a live-but-well-controlled room would sound larger than the same room made to sound dead.

Placing the ammount and type of room treatmen on a Likert scale, somthing like this might come through:
very live > treated but live > average > treated to dead > very dead
with each room state measured RT60/30 at various fs

Then again, there is a different aproach that is rather beyond the normal live vs dead discussion: make the room dead so that it has as little contribution as possible, run the input signal through a convolver to emulate the impulse response of larger spaces, then play back the now manipulated signal through the system, close your eyes and, voila, you are now in the Notre Dame, or at the Met Opera house, or in anyplace you can have an IR from. Best done with many speakers, spread arround the room much like in surround setups. Now, that is one for my bucket list : p

Ranking possible setups from bottom to top, I would say:

untreated room: tried it, worked so and so. better with speakers and listener away from walls
dead room: have not tried, need too much absorption which I do not have. Am guessing, biger is better here too
treated room leaning to dead: tried it, worked ok, easier to pick up on recorded details
treated room leaning to live: have not tried it. need many diffusors, which I do not have.
dead room + signal manipulation: have not tried it. need multitude of speakers & amps, plus somewhat of an expertise on how to do signal manipulation on a computer, far beyond my reach.

Still learning, though : )
 
#23 ·
This is a very useful discussion, and timely considering the listening that we did this weekend at Sonnie's. Much of what he is after in his 2 channel system is a good image and sound stage.

So just for the record, and to contextualize the comments that I will be posting about the eight pairs of speakers we auditioned, here are my views.

First, imaging is the ability to produce sound that to some degree convinces the listener that the reproduced sound is not coming from the speakers but from other locations. This is something like suspension of disbelief. We KNOW that the speakers ARE producing the sound, but if we listen for where the sound is coming from it SEEMS like it is coming from between, behind, or outside the speaker positions. I won't get into debates about what effects good imaging, but most stereo systems are capable of it to some degree. Better imaging means more likely to be convincing to me. It also means that the image is robust with respect to the position of the listener, for me. That might not be important to others, but much of my listening is not sitting in the sweet spot. I might be in the kitchen or moving about the room listening casually but would still like to pretend that I have a real performance going on in the house.

Sound stage to me is the totality of the locations that I perceive though imaging. It can have dimensions of depth, width, and height. Depth is pretty easy, relatively speaking. Fewer systems will locate sounds outside the speakers, and still fewer will distinguish the height of instruments. Imaging and sound stage can both depend upon placement of speakers in a room, more for sound stage, which is the detail of where things seem to be. For instance, in the Nickel Creek song that we used in Sonnie's listening sessions, there was a part where there is very clearly an the acoustic guitar in the same vicinity as a stand up bass. On most of the well placed speakers, sound of the bass was slightly behind and center relative to the guitar and both were behind the speaker in my perception. The bass sounded like it was lower as well. Some speakers located these with more precision than others, and until the best location for the speakers was found, most had trouble with the image at this level of detail of sound stage placement.

Certainly, the quality of the recording, the original hall that the performance was in, the mix, the equipment in the recording and playback chains, the room, speaker placement, listening position, and the listener all play a role in imaging and sound stage. I won't debate the relative importance of each, but the ones that we have the most control over are speaker selection and placement.
 
#29 ·
A seasoned pair of SF Concerto I had for nearfield listening benefited immensely from a slight rack angle. Tedious work involved applying quarter turns to the front spikes and listening on a good vocal track for that "sudden" moment - center vocal image with natural weight and density hanging in clear 3D space.
 
#36 ·
In my experience/opinion a sound stage cannot be altered by the end user. The sound stage was created by the studio it was recorded in and the techniques used by the engineer and recording artist. It is a sonic signature.

Imaging, is a representation/replication of the sound stage the artists were trying to capture in the studio. We can control speaker setup and selection, hardware, software and acoustic treatments to try and emulate that.
If a person gets the setup right, you'll have a good "image" and the best possible recreation of the intended "soundstage". Just throwing it out there.:gulp:
 
#38 ·
Wow, great contributions and suggestions. You guys have made this a "keeper" thread.



Yeah, it is getting that recorded soundstage to show up in your room that is the hard part. With depth is harder. With tightly defined depth is the hardest.

IME/O most heavily processed, tweaked and massaged studio recordings (pop, rock) don't have much (if any) of a sound stage. Sure, they have great imaging and clear clean sound. However, that sense of musicians playing in a large room "over there" where your speakers used to be is lacking. Without that "sound" you don't really have any kind of a sound stage. Those same recordings always have great lateral imaging. It's easy to do that with a pan pot. OTOH no amount of level manipulation will give a sense of depth.
Probably largely true. I am biased, being attracted to the rock that IS recorded to have spaciousness and depth, and there is a fair amount of it out there, mainly in the progressive, experimental, alternative, indie genres, plus instrumental, bluegrass, singer-songwriter genres. Not so much in metal or straight pop areas.
 
#37 ·
IME/O most heavily processed, tweaked and massaged studio recordings (pop, rock) don't have much (if any) of a sound stage. Sure, they have great imaging and clear clean sound. However, that sense of musicians playing in a large room "over there" where your speakers used to be is lacking. Without that "sound" you don't really have any kind of a sound stage. Those same recordings always have great lateral imaging. It's easy to do that with a pan pot. OTOH no amount of level manipulation will give a sense of depth.
 
#39 ·
OTOH no amount of level manipulation will give a sense of depth.
Depth in studio recordings that are not a live event, tend to also be manipulated by the board. Everything pretty much is controlled by the engineer and a good recording engineer that has the artists approval can make or break the whole shebang.

Too much overdubbing and the sound space can be confusing, too little and sound can thin out and maybe even become a thing of beauty. Width, depth, multiple versions of each person, placement of everything is in the hands of the person working the board.
 
#40 ·
My music collection includes a very wide variety of recordings (3500+ LPs, 1600+ CDs) that have been collected over the past 46+ years. I have a system with dipolar speakers that is a standout for exhibiting a wide deep and spacious sound stage when one is there. In addition, I have experience with three aspects of recording, in front of the microphone, at the mixing board and in front of my stereo. It's my experience based on my system and recordings that most (almost all) heavily processed studio recordings have little if any type of real sound stage. At best they give the impression of cardboard cut out musicians that are in very close layers. Almost none of them have the kind of depth and width (height also) that a good Telarc, Mercury, old RCA recording or most especially any direct to disk recording has.

IME any and all the level adjustments that are made at a mixing board cannot substitute for a recording made "live" with a minimal number of microphones, no studio manipulation of the signal and all the musicians present and playing together in real time. Pan potting and level adjustments are simply incapable of providing a real sense of depth.
 
#41 ·
Very interesting opinion you have and something we may all wish to explore more in depth. It is even more fascinating that we are close to the same age, have been listening for about the same amount of time, both of us have worked on both sides of the mic, yet have similar and different opinions in these matters.

I do also appreciate when you mention telarc, mercury, rca and what have you as some of those are truly golden, especially when Wilma was at the help and Fritz was on the performance side. (I am a Chicago boy to be sure but not quite that old)

I hope you do not mind discussing further as we have such similar and yet diverse opinions and I truly wonder how that comes to fruition. I like very much dipole sound, I have had many in my life ranging from Magnepan to Martin Logan but for reasons not even clear to me, I found the sound space provided by a good moving coil type speaker to on occasion edge out dipole, although I think open baffle may be a slightly different animal but the Dahlquist DQ 10 still reigns strong in my speaker loves. (Yes I still have a pair made in '83)

Interesting, I look to learn a bit here.
 
#43 · (Edited)
DQ10's are one of the few non planar speakers that I could maybe live with. I've also heard some open baffles that sound quite good to me. I think a large part of what open baffle and high efficiency speakers bring to the table is the ability to use SET amplification. There is a certain seductiveness in the sound of SET's. I would probably use them except that all the speakers I really like and those I own need substantial amounts of power and high powered SET's are quite expensive.

I'm sticking to my guns about level adjustments in the mix being incapable of creating any sense of depth. I've never heard it either in the studio or on any system I've heard. OTOH direct to disk recordings and other recordings (Mercury, old RCA, etc) made live with a minimal number of microphones and no studio signal manipulation can and do have wide and deep sound stages. For those who don't do LP's a good CD example of this is any Groove Note recording. They are recorded live to a two track master using a minimal number of microphones, passive mixers and no post processing whatsoever.

IMO it's all about phase differences in the signals reaching the microphones. Those differences can't be duplicated in a mixing board.

Incidentally: I used Magneplanars from 1976 to 1983 (MG-1, 2 and 3). I then switched to Acoustats. I've tried most of their models and am quite happy with my current two pair. I've tried ML's but I think the crossover frequency for many of their models is high enough that there is a sound discontinuity at the crossover frequency caused by the different radiation patterns between the esl panels and the cone woofers. The same thing occurs when any planar is paired with cone type woofers. It's just a lot less audible with lower crossover frequencies. That's why my cone subs operate only below 85Hz.

More thoughts: Concerning headphones; I often read comments about the sound stage "X" headphones produce. I guess my ear brain combination is different. I've never heard any kind of sound stage from any headphone I've ever heard. All I ever get is sound in a line from one ear to the other. I wish it were different as I've used headphones more frequently in the last two and a half years while undergoing radiation and chemo therapy sessions. Even without a sound stage my Grado's and ClipZip full of FLAC files are indispensable.
 
#44 ·
How do you define good imaging & good soundstage ????? that's an easy one

All you need is PANELS - PANELS - panels -




























With ALL of these IMAGING & SOUNDSTAGE is in the bag.:):):)
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top