• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Properties of speakers that creates a large and precise soundstage

Bjorn

Major Contributor
Audio Company
Forum Donor
Joined
Feb 22, 2017
Messages
1,258
Likes
2,486
Location
Norway
I have posted some measurements of the KEF Q350 here:
 

ahofer

Major Contributor
Forum Donor
Joined
Jun 3, 2019
Messages
4,877
Likes
8,581
Location
New York City
I wish I could get my SVS Ultra Towers to have an even sound stage with off axis listening. From where I sit (just to the left of center) the phantom center from two channels is on point but from the right seat (this is theater seating with a small island in the middle so you can’t sit dead center) the right speaker is dominant and no matter how I point the speaker or how close or far from the wall it is it’s always dominate.

No my hearing is fine; both young and old listeners notice this as well.

My walls are treated but not overly so and I’m using dirac live which does help immensely to improve audio quality throughout the entire room; but that one right channel speaker (this happened with audyssey and it happens with direct stereo audio too), is always dominate and I can’t get a good phantom image from the right seating position

I’m wondering if maybe this is because I have my audio rack somewhat near the left channel speaker; it’s against the wall and about 3 feet in front of the left channel speaker.

I’ve tried toeing in at extreme angles too and still… still….

So, at this point I’m looking to buy some JBL horn style speakers or something with a wider stage in hopes that it’ll fix this. I love the SVS ultra towers, they have very good sound and blow my mind every time I listen to them with or without subwoofers and they play loud without any distortion (at least for as loud as I’m willing to listen hehehejeej)

But this sorta thing drives me insane and I need a solution.

I hope I’m not the only one who has been haunted by an audio issue like this before and that someone here can provide assistance or at least a hug and tell me it’s okay and that I’m not alone

Anyhow, thanks for reading and I hope everyone is having a nice week.
I think the wider, more even dispersion is your best bet, unfortunately. I did see somewhat better image balance with R/L convolution files in EQ, but I suspect that won't ultimately solve the problem. Obviously a little experimentation with toe-in might also be worth it.
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,354
Likes
2,366
Location
Sweden
I wish I could get my SVS Ultra Towers to have an even sound stage with off axis listening. From where I sit (just to the left of center) the phantom center from two channels is on point but from the right seat (this is theater seating with a small island in the middle so you can’t sit dead center) the right speaker is dominant and no matter how I point the speaker or how close or far from the wall it is it’s always dominate.

No my hearing is fine; both young and old listeners notice this as well.

My walls are treated but not overly so and I’m using dirac live which does help immensely to improve audio quality throughout the entire room; but that one right channel speaker (this happened with audyssey and it happens with direct stereo audio too), is always dominate and I can’t get a good phantom image from the right seating position

I’m wondering if maybe this is because I have my audio rack somewhat near the left channel speaker; it’s against the wall and about 3 feet in front of the left channel speaker.

I’ve tried toeing in at extreme angles too and still… still….

So, at this point I’m looking to buy some JBL horn style speakers or something with a wider stage in hopes that it’ll fix this. I love the SVS ultra towers, they have very good sound and blow my mind every time I listen to them with or without subwoofers and they play loud without any distortion (at least for as loud as I’m willing to listen hehehejeej)

But this sorta thing drives me insane and I need a solution.

I hope I’m not the only one who has been haunted by an audio issue like this before and that someone here can provide assistance or at least a hug and tell me it’s okay and that I’m not alone

Anyhow, thanks for reading and I hope everyone is having a nice week.
Have you measured the response from each position (L and R separate)?

There are several possible reasons
- early reflections differ L vs R (furniture in and around the speaker)
- floor reflection is different L vs. R
- side wall reflection is different L vs. R

Both above can cause timbral differences as well as reflections that reveal the source distance/position, making one speaker dominant over the other. Seating could also affect if e.g. ear level is different from the positions.
 
Joined
Jul 25, 2022
Messages
32
Likes
8
Hello again!

One more question

When is distortion really audible? I’ve been thinking about those ~10% at 3,5khz on the graph of the first post.
 

HarmonicTHD

Major Contributor
Forum Donor
Joined
Mar 18, 2022
Messages
3,326
Likes
4,809
There is no clear cut threshold, it is also frequency dependent and highly individual. Yes 10% should be audible. But you can take online listening ABX tests (see eg Benchmark or Klippel website) to find out where the threshold lies for you.

Are any of the speakers showing 10% at 3.5khZ? That seems high. At what SPL? Sorry didn’t have time to look.
 
Last edited:

AdamG

Follow the Science, no matter where it may lead.
Moderator
Forum Donor
Joined
Jan 3, 2021
Messages
4,569
Likes
14,686
Location
Reality
Joined
Jul 25, 2022
Messages
32
Likes
8
There is no clear cut threshold, it is also frequency dependent and highly individual. Yes 10% should be audible. But you can take online listening ABX tests (see eg Benchmark or Klippel website) to find out where the threshold lies for you.

Are any of the speakers showing 10% at 3.5khZ? That seems high. At what SPL? Sorry didn’t have time to look.
Sorry, I replied without noticing my original post was merged here :facepalm:


At 75db\ SPL (green B&W 805; red LS50) - might be a cross-over problem? (20 years old..)

LS50_805_DISTORTION_percentage.png
 

HarmonicTHD

Major Contributor
Forum Donor
Joined
Mar 18, 2022
Messages
3,326
Likes
4,809
Sorry, I replied without noticing my original post was merged here :facepalm:


At 75db\ SPL (green B&W 805; red LS50) - might be a cross-over problem? (20 years old..)

View attachment 326911
Ah I see. I am not an expert when it comes to speaker distortion measurements so it would be good if others chime in.

I think the high absolute values of up to 10% are not correct because
a) you are not measuring near field and or even better, quasi anechoically such as eg the Klippel does and therefore get also the „room distortion“ into your values.
b) Assuming you use UMIK1 or so. Here I am not sure how well suited these mics are for distortion measurements.

What you can also do is, if you find distortion measurements from Erin or Armir for those speaker and compare (usually these are done at 86 and 96dB SPL and for good speakers are generally an order of magnitude lower than what you show.
 

Curvature

Major Contributor
Joined
May 20, 2022
Messages
1,059
Likes
1,319
According to David Griesinger studies, or at least how I understand them and can associate to my listening situation, there cannot be engagement, nor envelopment, or sharp localization, if listening beyond the audible critical distance
Beyond the "limit of localization distance". Critical distance is something else entirely.
 
OP
sigbergaudio

sigbergaudio

Major Contributor
Audio Company
Forum Donor
Joined
Aug 21, 2020
Messages
2,543
Likes
5,138
Location
Norway
Sorry, I replied without noticing my original post was merged here :facepalm:


At 75db\ SPL (green B&W 805; red LS50) - might be a cross-over problem? (20 years old..)

View attachment 326911

Do you have the frequency response of this measurement as well?
 

tmuikku

Active Member
Joined
May 27, 2022
Messages
296
Likes
326
Hi Curvature, yes, hence calling it audible critical distance as I do not know proper word for it, and it's relation to actual critical distance. Perceptually it's kind of what critical distance means, as if direct sound or room sound dominates either side of it, and there is a transition between, where they seem perceptually equal. Critical distance is where they are equal by measurement (SPL) https://en.m.wikipedia.org/wiki/Critical_distance

I'm not sure how localization distance is defined, never heard of the term before. Sounds are quite localizable (as in angle from median plane) beyond the audible critical distance, just not as accurately as closer, it's more hazy kind of sound perceptually. "Bloated" or "enhanced* by local room, depends how you think and feel about it, could be good or bad thing.

I would think that if actual critical distance was measured with third octave bands for example, it would get further away with increasing frequency band. Very low frequencies have no direct sound as wavelengths are longer than room, and on the other end at very high frequencies brain has plenty of time to analyze before reflections come in. Logically there is some bandwidth in between the extremes that is responsible for the shift in perception. Haziness with localization further away could be then explained, close enough for localization, but too far for clarity.

So, I think audible critical distance doesn't necessarily coinside with localization distance, unless specific directivity and room acoustics. What do you think?

edit. Griesinger paper atyached, which identifies localization distance and proximity happen simultaneously, like you say.

I'm not sure if and how this stuff translates to stereo hifi. All i know is that my perception changes at certain listening distance with my hifi setup, and Griesinger work is I can associate with, haven't seen any other studies or attempts to explain other than his. I'd be glad if you / others have and shared experience or knowledge on this.
 

Attachments

  • p35.pdf
    368.1 KB · Views: 18
Last edited:

Curvature

Major Contributor
Joined
May 20, 2022
Messages
1,059
Likes
1,319
Hi Curvature, yes, hence calling it audible critical distance as I do not know proper word for it, and it's relation to actual critical distance. Perceptually it's kind of what critical distance means, as if direct sound or room sound dominates either side of it, and there is a transition between, where they seem perceptually equal. Critical distance is where they are equal by measurement (SPL) https://en.m.wikipedia.org/wiki/Critical_distance

I'm not sure how localization distance is defined, never heard of the term before. Sounds are quite localizable (as in angle from median plane) beyond the audible critical distance, just not as accurately as closer, it's more hazy kind of sound perceptually. "Bloated" or "enhanced* by local room, depends how you think and feel about it, could be good or bad thing.

I would think that if actual critical distance was measured with third octave bands for example, it would get further away with increasing frequency band. Very low frequencies have no direct sound as wavelengths are longer than room, and on the other end at very high frequencies brain has plenty of time to analyze before reflections come in. Logically there is some bandwidth in between the extremes that is responsible for the shift in perception. Haziness with localization further away could be then explained, close enough for localization, but too far for clarity.

So, I think audible critical distance doesn't necessarily coinside with localization distance, unless specific directivity and room acoustics. What do you think?

edit. Griesinger paper atyached, which identifies localization distance and proximity happen simultaneously, like you say.

I'm not sure if and how this stuff translates to stereo hifi. All i know is that my perception changes at certain listening distance with my hifi setup, and Griesinger work is I can associate with, haven't seen any other studies or attempts to explain other than his. I'd be glad if you / others have and shared experience or knowledge on this.
Griesinger's term for the collapse of subjective clarity in concert halls, related to where you are in the hall, is "limit of localization distance".

Critical distance is a purely physical ratio which has no subjective aspect.

Much of what Griesinger has written about is not applicable to small rooms because of their acoustic (physical) properties. He does, very usefully, draw on the psychoacoustic work of Bregman regarding stream separation. @Duke has already commented on this part.

As far as I know Griesinger's work is unique. Especially for small room acoustics, we have scant guides other than Toole's.
 
Joined
Jul 25, 2022
Messages
32
Likes
8
Do you have the frequency response of this measurement as well?

It is in this post :): https://www.audiosciencereview.com/...precise-soundstage.48542/page-29#post-1772324

index.php


Ah I see. I am not an expert when it comes to speaker distortion measurements so it would be good if others chime in.

I think the high absolute values of up to 10% are not correct because
a) you are not measuring near field and or even better, quasi anechoically such as eg the Klippel does and therefore get also the „room distortion“ into your values.
b) Assuming you use UMIK1 or so. Here I am not sure how well suited these mics are for distortion measurements.

What you can also do is, if you find distortion measurements from Erin or Armir for those speaker and compare (usually these are done at 86 and 96dB SPL and for good speakers are generally an order of magnitude lower than what you show.

Thank you!
a) makes sense
b) I compared my results with both the LS50 Meta measurements by Armir (and they do not hit the 5% I am getting) - same thing on the 805S (I know they are not exactly the same model)
 

tmuikku

Active Member
Joined
May 27, 2022
Messages
296
Likes
326
Griesinger's term for the collapse of subjective clarity in concert halls, related to where you are in the hall, is "limit of localization distance".

Critical distance is a purely physical ratio which has no subjective aspect.

Much of what Griesinger has written about is not applicable to small rooms because of their acoustic (physical) properties. He does, very usefully, draw on the psychoacoustic work of Bregman regarding stream separation. @Duke has already commented on this part.

As far as I know Griesinger's work is unique. Especially for small room acoustics, we have scant guides other than Toole's.
Hi, yeah his work is mostly concerning concert hall acoustics but the phenomenon he studies happens in hearing system, which is the interesting part. We carry the same hearing system no matter size of the room or sound source we are facing. You are right though, that he calls the transition "limit of localization distance", or LLD in short, and perhaps I should too. But, as it's small room acoustics and stereo setup concerning phantom image, I'm not sure what the relationship is to LLD, is it exactly the same thing?

I mean, there is a loosely described listening test in the paper I attached to previous post, where Griesinger tests LOC in a small room using two loudspeakers each playing separate violin part. His transition (LLD) happens there as well, the LLD in small room! For now I've been listening clarity of phantom center, mono signal from two sources, while his test in paper is listening localization of two violins from two loudspeakers, no phantom image involved. I wonder if both listening tests would make same transition distance?
small-room-LOC-test.png

I think audio files he uses in the test are these: https://research.cs.aalto.fi/acoust...ment-and-analysis/85-anechoic-recordings.html I need to try this, listen how the test in paper works out and if LLD distance is same where the phantom center clarity happens (what I've been calling audible critical distance). The paper does not say how far apart the speakers are in the experiment, nor what kind of directivity the speakers have, so if there is difference in location of LLD and "audible critical distance", it's due to this kinds of stuff I think.

Here, another paper of his has paragraph about the stream separation:
https://www.akutek.info/Papers/DG_Audibility_Direct_Sound.pdf
stream-formation.png

The snippet above very accurately describes perceptual effect what I've been hearing with my home stereo system and why I'm thinking it's the very same thing in home stereo situation, transition between "hazy frontal sound" vs. "enveloping clear sound" and with about one step difference in listening distance makes the transition happen. Although phantom center of stereo setup vs. two real sounds sources on a concert hall is not exactly the same thing, the change in perception is likely due to the very same reason (hearing system) because the perceptual effect seems to be the same, or similar at least. Well, perception is different due to very different acoustics, but the change in perception at certain listening distance seems to be the same.

Have you tried to experiment with all this with your stereo setup? I think anyone should be able to find the transition, LLD, just by changing listening distance.

ps. could you link to Duke's comment you refer to?
 
Last edited:

tmuikku

Active Member
Joined
May 27, 2022
Messages
296
Likes
326
Alright here Tapio Lokki recordings, the Mozart violin 1 on left channel and violin 2 on right channel. Boosted both 20db as they seem to be very quiet tracks.

I'm not sure if these tracks are what was used in his test on the paper, or how to actually listen to it while testing. Played back on my stereo system, the violins are at least partially localizable to different angles (left and right speaker) in horizontal plane at quite far distance, at least meter further than where "the transition" with phantom center is with my setup. Better localization, also clarity and sound shifting from frontal to more enveloping, happens about where it happens for mono / phantom center sound as well. Listening opposite side of room there is no separation in localization, it's just one blob of sound. I did not test if distance between speakers make any difference, or is it just function of room acoustics and speaker directivity, or any other stuff like other instruments whose bandwidth was different.

So, it's very likely Griesinger LLD what I've been listening to, wrongly naming and calling it "audible critical distance". I'm not sure why I haven't noticed it's clearly labeled LLD by him :D Perhaps, because the name is kind of unintuitive in stereo and phantom sound context, as left/right localization is quite apparent with longer listening distances as well, at least with pop/rock stuff I've been mostly listening to. It's just that the clarity of sound and sharpness of localization of phantom center happens at closer listening distance, which I've been calling the audible critical distance. In case of each instrument having it's own loudspeaker LLD as name makes a lot of sense.

Anyway, while the name is important to avoid confusion, the real source of confusion is if people do not know about the transition and how it affects perception. For example, there is no way one could get clarity if listening other side of domestic room by upgrading amplifier/ cables/ magic, just move your chair bit forward and save your money. Or, if one wants envelopment, just go closer than the transition. Not hearing envelopment? You should, so one now knows to experiment with toe-in, speaker directivity and acoustic treatment in order to trying to enhance it. Got too close to speakers so that listening distance just isn't practical? well, another situation where adjusting the system DI, positioning and acoustic could help. Lot's of stuff changes depending which side one is listening at, and it's great great listening excersize, how to connect your perception of your system in your room, to written concepts.
 
Last edited:

Curvature

Major Contributor
Joined
May 20, 2022
Messages
1,059
Likes
1,319
@tmuikku I apologize I'm quite busy but I will try to reply later. This topic is very significant for our understanding of perceptual phenomena in small rooms.
 

tmuikku

Active Member
Joined
May 27, 2022
Messages
296
Likes
326
No problem, no hurry, and thanks for your time anyway! :)

It's about a year since I first noticed there is such a transition, quite sudden change in perception, and it took several months before I found Griesinger papers and been trying to promote this stuff ever since, as I don't see it widely known. I think this is core to understand what to listen, and how my perception of sound relates to written concepts. To me it's really the key to unlock listening skills and understanding about what I'm actually hearing, how my room affects my perception of my stereo and how positioning affects perceived sound and what kind of speaker would perhaps work better in my situation, and most importantly to my liking, to know what I like with the music I tend to listen to. Also perspective to understand discussion online, apparently contradicting comments can all be true depending on which side of the transition they talk from, and confusion is only because they don't know. Knowing the transition and utilizing it at will is literally key to understanding, at least that's how I see it, so yeah, important stuff to stumble upon sooner or later :)

ps.
I'm in a hurry as well, didn't notice in the paper was more about the listening test, where short clips of single violin track was played alternating (panned) to either side, apparently the test is to listen at what distance one doesn't notice which speaker the sound comes from. I'll post such audio file later, busy now.
 
Last edited:

tmuikku

Active Member
Joined
May 27, 2022
Messages
296
Likes
326
Here are few more tracks for fun:

Two tracks have the Lokki Mozart Violin 1 playing alone, panned to left and then to right, alternating. Try to find a listening distance with your stereo setup where it's impossible to hear which way it's panned. There are two tracks, other one has full panning, and the other is only half way so a phantom image is involved, if it makes any difference. Third track is the one I posted earlier, violin 1 and 2 playing together, each in their own speaker.

Anyway, as disclaimer this is not anything official or in association with either professionals, Griesinger or Lokki and others, but just me a hobbyist trying to interpret what the test in Griesinger paper might have been. Hopefully its fine to use the tracks here, although this is not academic study use :) I hope these references are enough to allow me to post this stuff here.

Reference to the origin of the sound tracks I used in the samples.

And here is reference to their work:
Pätynen, J., Pulkki, V., and Lokki, T., "Anechoic recording system for symphony orchestra," Acta Acustica united with Acustica, vol. 94, nr. 6, pp. 856-865, November/December 2008. [Online IngentaConnect]

Here is the David Griesinger paper "Localization, Loudness and Proximity" with the listening test the tracks are supposed to kind of relate to. I attached the file to my earlier post, and here is reference where I found it:

His webpage here http://www.davidgriesinger.com/ links to a power point file with same title, Localization, Loudness and Proximity/ , which also has it.
 
Last edited:

jim1274

Active Member
Forum Donor
Joined
Dec 28, 2023
Messages
276
Likes
183
When I set up a pair of stereo speakers outdoors, it's quickly noticeable that all the soundstage is located between the speakers, other than hard-panned instruments or vocals coming directly from the speakers.
Makes me realize that without lateral room reflections, it's not possible to have a soundstage wider than the speakers.
So if I want an indoor soundstage wider than the speakers, it boils down to how strongly do I want lateral reflections to widen the soundstage.

Which makes a speaker's radiation pattern the key determinant.
And I think the idea of what makes for a good speaker has to first specify what is the radiation pattern desired.

I know we throw out the idea of a point source for example, but is that at all really descriptive? A true point source is omni.
Don't all speakers have increasing directly as frequency increases ? (except maybe ball-type designs like a dodecahedron used for room acoustics measurements )

I think a dome tweeter is probably par for the course in terms of a HF/VHF pattern preference, for most folks.
Then we have waveguides to narrow pattern some. Then horns for further narrowing, etc. And we can't forget planars, dipoles, and on and on, and their particular radiation control.

Having chosen a radiation preference, I certainly believe there are principles of design that make for "best speakers"
But I think those principles need to be considered apart from the idea of what makes the best soundstage....as soundstage is much more about speaker & room interplay.

Or to say the same thing another way......I think soundstage optimization as part of speaker design is more about satisfying envelopment preference, than good fundamental principles of speaker design that apply to any speaker. ....and that speaker and room must be considerd together.

The second realization/observation from outdoor stereo listening, is significantly improved imaging in-between the speakers, with the phantom center becoming relatively rock solid.
It's a sound I prefer. I pretty much say to heck with envelopment, when i can achieve glorious direct sound. So my speakers are built to minimize lateral reflections and hold pattern control as low in frequency as possible. my2c and preference

You brought up a lot of good points here. I’ve been experimenting with different combinations to get good full range sound across a long wide pool area outside, and got some Duevel Omnis last summer which would probably have stayed if not for insufficient SPL capability and low end response. They were the small Planets. I looked for threads about outdoor sound set-ups, but didn’t find much beyond more limited goals and typical outdoor speakers. I’ll probably start one later to share some of my experiences, all using indoor speakers.
 

jim1274

Active Member
Forum Donor
Joined
Dec 28, 2023
Messages
276
Likes
183
FWIW, I was skeptical of a direct firing “box” speaker even coming close to an Omni (or even to a dipole), on the soundstage front, but changed my view today when trying some different “box” speaker placements. I was previously comparing the box soundstage too far from the front wall in a more near field listening arrangement. After some adjustments in location, the box soundstage became much larger. The soundstage has a different character with the box speakers, maybe preferable to many with better spatial localization of instruments in the mix.
 
Top Bottom