Imaging

[josko]

I attended a live performance (Shostakovich 5th) and noticed how very well I coud image (estimate direction to) various players in the orchestra. When I played the piece on my 2.0 system, yes, I could still distinguish some directionality, but it was nowere near as definitive as in the live performance.
What can I do to improve imaging on my home stereo system? I tried playing with equalization curves, and it seems they don't help, and some distinctly made things worse. Should I be trying to deaden reflections in the room? Do I want a 'dead' room for optimal imaging or a 'live' room?
My speakers are B&W 802D's, and they are carefully positioned equidistant from the wall and from the the listening point. I also have a center speaker (B&W HTM1D), and one of my recordings is in Dolby TrueHD 5.1, but it seems the center speaker detracts from imaging, rather than helps. It almost seerms like my CD 2-channel recording has better imaging than the 5.1 recording - seems totally counterintuitive.
Anyway, I would welcome any help in setting up my 2.0 system for better imaging. I do a lot of listening to orchestral music.

 

[bikemig]

Signing on as I want to read the posts in this thread. I think using living music as a reference point is the right way to think about setting up a music system and orchestral music is just tough to reproduce right in one's house. You might want to describe your room and music set up since those are both key variables.

 

[josko]

Ok, here goes:
Music setup: Oppo BDP-93 and Dual 1219 to Emo UMC-1
Levinson 334 (bridged) to each B&W 802D (fronts)
Emotiva XP5 to HTM1D and 803D (center and rears)
Sub SVS SB-13 Ultra sub

Room is 18' x 24', with a complex ceilning 9'-12' high, with the speakers symmetrically arranged. There's a line of French windows on the left side which is typically curtained during critical istening. Two 4' passageways to the dining room in the corners of the room (18' side, behind the listener). Sub is in one corner with a bass trap in the adjacent corner (both behind and to the side of fronts), but is not used much for music listening, as it seems to detract from imaging, and the 802's go down as low as most orchestral music.

 

[Irvrobinson]

Imaging is very dependent on the capabilities of the speakers, and I've never heard the B&W 800 series throw an awesome image. I'm not saying they can't, but I've had several extended demo sessions with 800Ds and 802Ds at two different dealers and the imaging didn't impress me. The dealer set-ups were pretty good too. It's always been a mystery to me, because the 800 series looks like it should throw a great image, all that narrowness and roundness.

You didn't mention the exact positioning of the speakers in your room, but they should be at least four feet from the back wall, and three or four feet from the side walls. With the 802D you'll also want to experiment with toe-in. For maximum imaging you want your listening seat to complete an equilateral triangle with the speakers. An isosceles triangle works too, but narrows the soundstage. Eliminate any large hard surfaces, like a coffee table, between your listening seat and the speakers. Make sure your listening seat is several feet from the back wall.

I recommend getting the sub out of the corner and placing it somewhere between the 802Ds and using it, even at a low level just for fill-in to reduce room modes. It's not a matter of bass volume, it's about in-room smoothness. In my attempts at sub placement the corners were never good for imaging, and either were room modes. What you really want is a room measurement system like the Parts Express OmniMic to see what's going on at your listening position for sub placement.

You don't want a dead room. I have a wall of glass behind my speakers, and that doesn't seem to harm anything. Symmetry doesn't seem to matter that much, because my room is asymmetric, and the imaging I get is awesome.

 

[Steve81]

One other factor I'd toss out there (beyond the speaker design itself and the room/placement which have already been mentioned) with respect to imaging is matching of the right and the left speakers. Tolerances at the driver production level, XO parts, etc can and will make small measurable differences between two speakers in a pair. Just to give an idea, Ascend notes on their Sierra tower page:

Left / right pairs are matched by hand to within ± 1dB and we are including a printout of the actual production line response measurements of each speaker.

Even a +/- 1dB difference can potentially be a pretty big difference, and that's with the manufacturer actually taking the time to match by hand and include the response measurements for each speaker!

 

[Irvrobinson]

One other factor I'd toss out there (beyond the speaker design itself and the room/placement which have already been mentioned) with respect to imaging is matching of the right and the left speakers. Tolerances at the driver production level, XO parts, etc can and will make small measurable differences between two speakers in a pair. Just to give an idea, Ascend notes on their Sierra tower page:



Even a +/- 1dB difference can potentially be a pretty big difference, and that's with the manufacturer actually taking the time to match by hand and include the response measurements for each speaker!

Good point, but I suspect the B&Ws are more than adequate in this regard. Their quality control appears to be outstanding.

 

[Steve81]

Good point, but I suspect the B&Ws are more than adequate in this regard. Their quality control appears to be outstanding.

One would certainly hope (and I'm not trying to imply it's the case here).

 

[shadyJ]

The recording plays a big role in imaging. You will have to find a recording that you know has the imaging that you are looking for before you can even begin to worry about the speakers. That ideal recording may not exist.

 

[AcuDefTechGuy]

1. Find a very high quality recording. Most recordings are crap compared to the original live concerts.
2. Move the 802D 3-4 ft from all walls.
3. Toe them in (experiment with different angles from 20-45 degrees).
4. Turn off/ bypass all Room Correction (like Audyssey, etc), EQ, DSP.
5. Set speakers to Small, Crossover to 80Hz.

 

[endless ent]

Tweeter and midrange differences

I noticed the same spacial differences over time due to so many topics but B&W specific just last week. A friend showed me why he had changed from his to Sonus Faber Cremona for just this reason. What if a phase change or reflecting both into a corner at various angles, changing the incidence of reflection broadened the sound stage. So many brands tout reflective sound like Mirage, Bose, wave guides a plenty and never forget the dispersive Klipsch Horn. I'm a big B&W fan but I don't get hugs from an elephant. Sound staging may be stronger in another brand. Have you tried all but neural processing with digital effects, also not a bad idea.

 

[josko]

I have a 5.1 DolbyTrueHD recording (SF Orchestra) on Blu-ray and a London Philharmonic CD (2.0); I'm very surprised that the 2.0 recording seems to have better imaging than the fancy 5.1. I tried the 5.1 with and without the center speaker and sub, and just can't get the imaging as clear as with the older CD.
So I guess, yeah, quality of the original has much to do with imaging ability; I am surprised that the much-touted Michael Tilton Thomas Blu-ray doesn't seem to image (on my system, anyway) as well as the older CD.
I also have a (60's vintage vinyl) Deutche Grammophon recording of Karajan and the Berlin Philharmonic, and the imaging quality there is right on par with the CD.

Is there some kind of a reference recording for calibrating imaging?

 

[AcuDefTechGuy]

The recording plays a big role in imaging. You will have to find a recording that you know has the imaging that you are looking for before you can even begin to worry about the speakers. That ideal recording may not exist.

Man, you beat me to it.

 

[AcuDefTechGuy]

I have a 5.1 DolbyTrueHD recording (SF Orchestra) on Blu-ray and a London Philharmonic CD (2.0); I'm very surprised that the 2.0 recording seems to have better imaging than the fancy 5.1.

I'm not too surprised. I also usually prefer 2.1 over 5.1. That's why I hardly ever listen to 5.1 music anymore.

One exception I've found is the Dave Mathews/ Tim Reynolds Live at Radio City Music Hall BD. The 5.1 TrueHD here is awesome.

 

[3db]

I suspect the biggest culprit deals with the recording and not so much the speakers. I get the same kind of thing happening with my PSBs and they are known for their imaging capabilities. I would think the recording blends the orchestra together much more than our ears in a live concert. A much better venue would be a string quartet. In such a venue, it would be much easier to get the soundstage you are looking for.

 

[zieglj01]

Even something like the cheap Pioneer speakers can image
pretty good - and yes, there are better speakers around.
A lot depends on the recording, quality of the speaker with
the room and placement. >>> I do not use any type of EQ
above 200 hz, and I do not depend on the likes of Audyessy
or Mcacc to fine tune imaging. >> Overall I prefer 2.0 or 2.1
for the overall sound experience in music.

 

[Irvrobinson]

I suspect the biggest culprit deals with the recording and not so much the speakers. I get the same kind of thing happening with my PSBs and they are known for their imaging capabilities. I would think the recording blends the orchestra together much more than our ears in a live concert. A much better venue would be a string quartet. In such a venue, it would be much easier to get the soundstage you are looking for.

I'm not on this page. On my system many recordings, even the ones consisting just of studio mixes, have 3D imaging. Unfortunately, all of the speakers I've actually heard imaging like this with have been expensive. For example, the YG Carmel (I know someone that recently picked up a used pair), various Magicos, the Sound Lab electrostatics, the Linkwitz Orion, the Revel Ultima 2s, and some real oldies, the Martin Logan Monolith, the Dunlavy SC-V, and the a/d/s M30. I'm not saying other speakers can't also throw a 3D image on many recordings, I just haven't heard them. A friend across the country claims his Triton GoldenEar Two is so good in this regard he ditched his surround speakers (which made his wife ecstatic, apparently).

 

[3db]

I'm not on this page. On my system many recordings, even the ones consisting just of studio mixes, have 3D imaging. Unfortunately, all of the speakers I've actually heard imaging like this with have been expensive. For example, the YG Carmel (I know someone that recently picked up a used pair), various Magicos, the Sound Lab electrostatics, the Linkwitz Orion, the Revel Ultima 2s, and some real oldies, the Martin Logan Monolith, the Dunlavy SC-V, and the a/d/s M30. I'm not saying other speakers can't also throw a 3D image on many recordings, I just haven't heard them. A friend across the country claims his Triton GoldenEar Two is so good in this regard he ditched his surround speakers (which made his wife ecstatic, apparently).

That's ok.. We won't always agree. I have classical recordings that image a lot better than others but even the best recording won't approach a live concert for imaging. Every quartet I have really stands out however. I tend to believe that the recording itself is far more in play than you give credit for. I know this is an extreme but listen to mono and stereo ...see which images better. On some blues albums, I can hear the singer move from left to right and cross back again and settle back down in center stage. I'm not talking about extreme left or right but just moving off of dead center. I can also pick out the two Allman brother's guitarists move towards one another and away from one another as they play. Very cool.

 

[GranteedEV]

Let's start with the basic assumption that imaging is present in the source. I'm not sure this is a safe assumption, but i'm making it anyways.

Imaging cues are masked by 800hz to 8khz reflections in the first 10 - 15 ms. In physical space that's about 12 to 17 feet of travel distance. Generally this means giving your speaker 5 to 8 feet distance from boundaries. It ALSO means keeping yourself this distance from boundaries.

If the reflected sound is not delayed this long, the image will be smeared.

Diffraction off of nearby objects will smear the image as well. It might be as something as basic as a coffee table or a television that is creating an unwanted phantom image.

Side wall reflections also contribute to perception of spaciousness, so you may not want to do away with them. However they can affect timbre perception if they are not similar in frequency response to the the forward radiated sound.

Floor and ceiling reflections do not necessarily have positive contributions - so absorb or diffuse them where possible.

If your room is too narrow to have the adequate boundary distance, then you need to consider a different sort of loudspeaker. The Audiokinesis Planetarium Beta comes to mind if you can pull the speakers out into the room, else consider the Gedlee Abbey. Obviously you can't rush into a new speaker purchase, but auditioning isn't a bad idea. Maybe your speakers aren't capable of what you want.

 

[Swerd]

This thread reminds me of a very interesting but long post by an accomplished DIY speaker designer named Jeff Bagby.

My old article on Loudspeaker Imaging

Some of it is rambling and hard to read, but overall it’s the best collection of thoughts on the subject I’ve read. He started with a definition of imaging:

"Imaging” is the loudspeaker’s ability to recreate a sense of the original localization information that was present in the recording environment in such a way that vocals and instruments seem to be placed in space, even to the point of creating a sense of three dimensional depth.

Now, let’s take a look at what this definition implies. First, it implies that this localization information is present, which is not necessarily a given when you consider how many recordings are mixed. But when they are done correctly this information should be there. Second, it implies that there are aspects of loudspeaker design that influence, constructively or destructively, a speaker’s ability to reproduce this information. My theory will only deal with implication number two and will assume that the criteria for implication number one have been fully met.
​

He talked at length about possible differences between analog and digital recordings. I had trouble following some of that and probably don’t agree with some of his conclusions. But after that, he finally got down to details about what good speakers can do to create an image. I’ll try to summarize what he said about features of speakers important for good imaging:

1. First and foremost, I think having the correct tonal balance and near perfect balance between the Left and Right speakers has to be near the top because frequency response dominates our perceptions of speaker performance.
   
   
2. How well diffraction is controlled is also a factor in the smoothness of frequency response so it will be a part of this one too.
   
   
3. Third, I believe the mechanical inertness in the enclosure comes next based on the systems I have heard and the characteristics that they shared.
   
   
4. And forth, I think are the time domain issues. This is less important than the first three issues because its audibility is lower. However, if everything is done right I still believe that a transient perfect speaker will offer advantages over the speaker that has a lot of error in the time domain.

If you find any of this interesting, go read Bagby's whole post as he says it better.

 

[CSG]

The recording plays a big role in imaging. You will have to find a recording that you know has the imaging that you are looking for before you can even begin to worry about the speakers. That ideal recording may not exist.

Absolutely this. So many recordings have terrible imaging; it's just left/right stuff.

The imaging I get from my speaker sets (Celestion DL-8's and NHT SuperOnes) is dependent on placement, toe-in, and most importantly, the recording itself.