A stereo listening experience without inter-ear crosstalk can be quite amazing, and can be archieved quite easily within a really simple, but effective setup. Just grab any board and sandwich/squeeze it between your two speakers. Then, listen with your nose as close as possible to the outer rim of the board. Despite the quite owerwhelming listening experience you get from this one, nobody does want to listen like this for a longer time. Neither did I. I tried this one out with the mounted-away bathroom door. No kidding.
In a more normal stereo setup, typically you need two direct auditive paths/sectors, and for general convenience also one optical path/sector to listen to stereo in a home environment. Especially the audio paths/sectors must not be obstructed by any kind of structures, and also the visual one in case of a central screen.
Then, within this setup, there are these nasty xtalk-paths/sectors, inevitably making stereo to become xtereo. In order to theoretically avoid any interaural stereo crosstalk to the contralateral side, the sound energy within these paths/sectors to the contralateral ears should be either completely absorbed, or completely deflected away from the direct path to the contralateral ears. This might be performed by inserting obstructive structures right into these pathways:
By geometrically intersecting all the necessary and all the unwanted pathways, two aeras (or, more precisely two room volumes) will become evident, where the direct contralateral x-talk-path/sector can be obstructed, without obstructing the direct isolateral audio and optical paths/sectors. And only there you may place absorbing and/or deflecting structures, in order to be effective against xtalk without disturbing stereo. As the dimension/space of this intersection aera/volume is very limited, these structures necessarly will be acoustically small and ineffective at low frequencies. This means, that only higher frequencies will benefit from these anti-xtalk-structures:
The obstructing anti-xtalk-structures could be e.g. a hard board for deflecting sound away from the straight path to the counterlateral ear, or a woven PET acoustic absorbing board for absorbing sound energy, a combination, or anything else which might be suited. Also the shape of these structures might be quite variable and would have to be empirically optimized within a trial-and-error cycling process:
To define the the sectors in the above graphs, I arbitrarly chose to define them by the horizontal boundaries of the speaker baffle. And as said, because of the limited possible size of the x-path obstructing structures, only a frequency-dependent and partial x-talk minimizing effect will be archieved by this kind of setup. The higher the frequency, the more effective it might act against acoustical x-talk. And also, this is a first draft without proof of concept: I did not yet set up any experiments with this one for my own yet. I think an optimizing process should begin with a clean stereo geometry variant especially adapted in order to make the obstructable aera as big as possible. To allow for acoustically relevant obstructors.
If anybody already has own data about this approach, I would greatly appreciate to hear from these real-world-findings.
In a more normal stereo setup, typically you need two direct auditive paths/sectors, and for general convenience also one optical path/sector to listen to stereo in a home environment. Especially the audio paths/sectors must not be obstructed by any kind of structures, and also the visual one in case of a central screen.
Then, within this setup, there are these nasty xtalk-paths/sectors, inevitably making stereo to become xtereo. In order to theoretically avoid any interaural stereo crosstalk to the contralateral side, the sound energy within these paths/sectors to the contralateral ears should be either completely absorbed, or completely deflected away from the direct path to the contralateral ears. This might be performed by inserting obstructive structures right into these pathways:
By geometrically intersecting all the necessary and all the unwanted pathways, two aeras (or, more precisely two room volumes) will become evident, where the direct contralateral x-talk-path/sector can be obstructed, without obstructing the direct isolateral audio and optical paths/sectors. And only there you may place absorbing and/or deflecting structures, in order to be effective against xtalk without disturbing stereo. As the dimension/space of this intersection aera/volume is very limited, these structures necessarly will be acoustically small and ineffective at low frequencies. This means, that only higher frequencies will benefit from these anti-xtalk-structures:
The obstructing anti-xtalk-structures could be e.g. a hard board for deflecting sound away from the straight path to the counterlateral ear, or a woven PET acoustic absorbing board for absorbing sound energy, a combination, or anything else which might be suited. Also the shape of these structures might be quite variable and would have to be empirically optimized within a trial-and-error cycling process:
To define the the sectors in the above graphs, I arbitrarly chose to define them by the horizontal boundaries of the speaker baffle. And as said, because of the limited possible size of the x-path obstructing structures, only a frequency-dependent and partial x-talk minimizing effect will be archieved by this kind of setup. The higher the frequency, the more effective it might act against acoustical x-talk. And also, this is a first draft without proof of concept: I did not yet set up any experiments with this one for my own yet. I think an optimizing process should begin with a clean stereo geometry variant especially adapted in order to make the obstructable aera as big as possible. To allow for acoustically relevant obstructors.
If anybody already has own data about this approach, I would greatly appreciate to hear from these real-world-findings.
Attachments
Last edited:
Carver did a lot of work on this. Check out his sonic holography stuff. There are speakers that use properly spaced opposite channel out of phase drivers in each speaker to create binaural sound without headphones. The left ear only hears the left channel etc.. The drivers are a heads width apart and out of phase cancels the other channel to the ear. It only works over a limited sweet spot. Some love it, some hate it, some get sick from the sound messing with their brain!
Matthew Polk believed that using actual signals from the speakers to cancel out crosstalk was the way to go. He brought his ideas to life in his SDA speaker series.
Another approach came from Bob Carver, who believed that an IC could be programmed to eliminate crosstalk through signal alteration.
Carver Sonic Hologram Generator Model C-9.
Attachments
Last edited:
For interest, details of Polk SDA Technology: https://auth.polkaudio.com/en-gb/discover/sda-technology
Back wall FAIL. Do whatever you have to do to avoid being up against the back wall....
And NO, don't add extra vanes, reflectors, baffles, or whatever.
I read in early 90's someone did this with a mattress to block out the crosstalk, but I can't remember the conclusion.
I say point the speakers differently, obsess over the L-R symmetry of first reflection points, and enjoy the stereo. They worked all this stuff out back in the 40's...
And NO, don't add extra vanes, reflectors, baffles, or whatever.
I read in early 90's someone did this with a mattress to block out the crosstalk, but I can't remember the conclusion.
I say point the speakers differently, obsess over the L-R symmetry of first reflection points, and enjoy the stereo. They worked all this stuff out back in the 40's...
Attachments
The photos look crazy now, but before the invention of radar we had the acoustic location of aircraft.
^ German sound location, 1917. Combined acoustic/optical locating apparatus.
^ The Dutch personal parabola, 1930s.
More here: https://rarehistoricalphotos.com/aircraft-detection-radar-1917-1940/
Interesting mechanical options. 
You might also be interested in this thread:
https://www.diyaudio.com/community/threads/fixing-the-stereo-phantom-center.277519/
Be sure to read the attached paper.
You might also be interested in this thread:
https://www.diyaudio.com/community/threads/fixing-the-stereo-phantom-center.277519/
Be sure to read the attached paper.
I have completely neglected this really relevant thread while I was opening this new thread - what a shame! Otherwise, I would have added my post there. I think it would be best to stop this discussion here and continue it inside the fixing-the-stereo-phantom-center thread mentionned by pano.
Last edited:
OK, then. So let's go on here. And I will refer to your thread as the FTSPC thread (FTSPC will stand for Fixing the stereo phantom center) whenever it is appropirate to do so.
Thank you for all these pictures sent, which are some kind of proof of concept. Except that this approach here would place the obturator gear into the room, and not necessarily onto the listener's head.
In the FTSPC thread there is a discussion about the tonal changes caused by notching in the frequency response because of the inter-aural delays between the ipsilateral and the contralateral signals. And in the context of this FTSPC thread, how to best cope with these artefacts.
Instead, here exactly this notching will be very welcomed. Welcomed, because of every notch is a sensible indicator while tweaking for the best compromise for these acoustic obturator devices inside the obturation zone: You simply may make a simultaneous L+R (mono signal) measurement at the room location of one ear of you listening position, and then shift the obturator of the contralateral signal path around. If there were any effect, then these notches would become less deep. The placement at which the notches are the shallowest would then be the best compromise location for the chosen obturator.
I am really looking forward to get two 500Hz ... 5kHz horn drivers in arrival, to set up a bare-bone prototype xtereo-to-stereo measurement environment. We will see ...
Thank you for all these pictures sent, which are some kind of proof of concept. Except that this approach here would place the obturator gear into the room, and not necessarily onto the listener's head.
In the FTSPC thread there is a discussion about the tonal changes caused by notching in the frequency response because of the inter-aural delays between the ipsilateral and the contralateral signals. And in the context of this FTSPC thread, how to best cope with these artefacts.
Instead, here exactly this notching will be very welcomed. Welcomed, because of every notch is a sensible indicator while tweaking for the best compromise for these acoustic obturator devices inside the obturation zone: You simply may make a simultaneous L+R (mono signal) measurement at the room location of one ear of you listening position, and then shift the obturator of the contralateral signal path around. If there were any effect, then these notches would become less deep. The placement at which the notches are the shallowest would then be the best compromise location for the chosen obturator.
I am really looking forward to get two 500Hz ... 5kHz horn drivers in arrival, to set up a bare-bone prototype xtereo-to-stereo measurement environment. We will see ...
Well observed!
You are perfectly right within the real-world paradigm.
A person talking belongs to the real world.
A symphonic orchestra also is a real group of real musicians.
A thunderstorm is a real meteorological state also, no doubts.
And there is no need, and no way to eliminate or minimize xtalk.
You will hear what happens around you.
In real time in a real world.
Agreed.
Agreed.
Agreed.
And while listening to stereo material, you are listening to sound emanating from two real-world speakers.
So there is xtalk also.
And you still might be in the real world paradigm while doing so.
Audio reproduction in our rooms is physics and belongs to the real world.
So no need to eliminate or minimize xtalk.
Agreed.
But wait a moment ...
What's about the nature of audio reproduction?
I mean the second nature.
There never has been a person talking in your room.
Neither a symphonic orchestra.
And hopefully no thunderstorm.
Audio perceptions in our rooms does not correspond any longer directly to normality.
Now think further, about the nature of stereo, and maybe iterate a bit about the meaning of this very first sentence of this competent FTSPC thread:
So stereo definitively takes place in another paradigm.
It takes place in a world of auditory illusions.
In fact you have to deal with two different paradigms.
Reality and illusion.
Which do not exactly match.
And finally, have the freedom to choose between these two paradigms.
If you want to listen to your speakers, do not cancel xtalk.
Welcome to nice audio happening in the real world, then.
If you want maximum immersive stereo effect instead, you better do try to cancel or minimize xtalk.
Then be welcomed to be psychoacoustically taken away elsewhere from your acutal normality by a potentially mighty, evoked stereophonic illusion.
If you dare.
And afterwards, you will have to find a way back to your normal, physical reality and real-world surrounding.
But beware: Not everybody succeeds in trying to do so.
Agreed?
Last edited:
If I want to "play around" with the recorded stereo image, I use my SRS (Sound Retrieval System) which gives enhancement to the sound.
It makes the speakers dissapear, can widen and vary the stereo image, and give increasing concert-hall depth to the music and vocals.
CD's or tapes recorded with it can make a tabletop boombox sound like its 6 feet wide, you can imagine the effects in a car too.
It makes the speakers dissapear, can widen and vary the stereo image, and give increasing concert-hall depth to the music and vocals.
CD's or tapes recorded with it can make a tabletop boombox sound like its 6 feet wide, you can imagine the effects in a car too.
This is exactly what the described approach here is meant for. To play around with the stereo image. If you like the result, then the bette for you.
While doing so, be aware, that ...
... and therefore any sound engineer checking his mix along with speakers will also be subject to xtalk. So the most part of the available audio material is certainly optimized (as for the gusto of the sound engineer) to be listended to along with the ubiquitous xtalk you normally will encounter at home.
A different story are all these sound productions which are aimed at an earbud customership. These might have been produced along with earphones control. So the advice to ...
... is especially well suited in these cases.
There is neither "right" nor "wrong" in this contex. I's really a matter of ...
.... and having fun while doing so. And the simpler and cheaper the toys to play around with, the better.
In this terms, I think that such a simple setup as shown in the very first picture of this thread might be very well worth to try out. Simply to make the experience or how xtalk does influence the stereo illusion. Nearly everybody has a temporarly spare bathroom door in his home to take his chance.
To xtalk or not to xtalk. That's the question.
Last edited:
Binaural recordings might sound better through such a set-up than through headphones. Headphones disturb the resonances of the pinnae/auricles/external parts of the ears while loudspeakers with crosstalk cancellation do not, as long as there is nothing right next to your ears.
I never tried anything like this, but I did experiment with electrical crosstalk compensation networks in the past, that is, with a simple minimum-phase filter based on an article in Wireless World from decades ago. The results were not impressive. I think the filter was too simple.
I never tried anything like this, but I did experiment with electrical crosstalk compensation networks in the past, that is, with a simple minimum-phase filter based on an article in Wireless World from decades ago. The results were not impressive. I think the filter was too simple.
"Stereo inter-ear crosstalk minimizing for typical stereo loudpeaker setups - can it work?"
Hi,
To be honest I can't answer your question, also because I 've never done any experiment of that kind.
However, FWIW what I can say is just the following.
The fact that the stereo effect is defined as an illusion (by whom?) does not mean that it really is an illusion.
The hearing is a sense, and also one of the most complex and refined.
Extremely complex sense with various psychic implications and not at all completely known.
What generally concerns the senses I do not think it can be defined as illusion, but rather than as a perception.
Since human beings are extremely complex and unique beings then also their perceptions are such: extremely complex and unique,
Shareable perceptions in many cases, but still unique.
I liked your approach and I read your post with interest.
Recently I simply angled the loudspeakers so that the axis of the tweers falls outside the imaginary midline that separates them and the stereo effect has dramatically increased and improved.
On an almost regular basis I've the compulsive impulse to modify in different way some element in my audio system hearing for any improvements, but that position of the speakers (which I have naturally marked on the floor) remains a fixed point.
I can't change that positioning and even if I do the result it is never better so far.
Also the "holographic" imaging has improved, with the perception of individual instruments or voices of singers who are more precisely and concretely positioned in their virtual space.
Please continue your research because you could also achieve some unexpected success, some perhaps even by chance.
P. S.: I've not yet listened to an audio system that in a typical domestic room reproduces the recording of a musical event in a "real" way.
And IMO it's just here that the human auditive perception plays its role, because often even if a sound event is not reproduced in a "real" way it can be considered "realistic" and satisfy just enough to believe that one has well spent his money and his time and his commitment spent to achieve "the" purpose.
Speaking with some guys on this forum I learned they have succeeded in this intent (that is, to reproduce exactly the reality of a musical instrument which therefore was indistinguishable from the one played live in the same room), but deepening the matter then you realize that it succeeded only with a single and very simple acoustic instrument.
I no longer remember which instrument, but it was a very simple acoustic one.
However, I've never listened an audio system able to play a rock concert recording, or a symphonic concert, or any other musical event that has only remotely approximate the "reality".
I think that for that, one has to add his imagination, and perhaps only then the whole "thing" really become an illusion.
Hi,
To be honest I can't answer your question, also because I 've never done any experiment of that kind.
However, FWIW what I can say is just the following.
The fact that the stereo effect is defined as an illusion (by whom?) does not mean that it really is an illusion.
The hearing is a sense, and also one of the most complex and refined.
Extremely complex sense with various psychic implications and not at all completely known.
What generally concerns the senses I do not think it can be defined as illusion, but rather than as a perception.
Since human beings are extremely complex and unique beings then also their perceptions are such: extremely complex and unique,
Shareable perceptions in many cases, but still unique.
I liked your approach and I read your post with interest.
Recently I simply angled the loudspeakers so that the axis of the tweers falls outside the imaginary midline that separates them and the stereo effect has dramatically increased and improved.
On an almost regular basis I've the compulsive impulse to modify in different way some element in my audio system hearing for any improvements, but that position of the speakers (which I have naturally marked on the floor) remains a fixed point.
I can't change that positioning and even if I do the result it is never better so far.
Also the "holographic" imaging has improved, with the perception of individual instruments or voices of singers who are more precisely and concretely positioned in their virtual space.
Please continue your research because you could also achieve some unexpected success, some perhaps even by chance.
P. S.: I've not yet listened to an audio system that in a typical domestic room reproduces the recording of a musical event in a "real" way.
And IMO it's just here that the human auditive perception plays its role, because often even if a sound event is not reproduced in a "real" way it can be considered "realistic" and satisfy just enough to believe that one has well spent his money and his time and his commitment spent to achieve "the" purpose.
Speaking with some guys on this forum I learned they have succeeded in this intent (that is, to reproduce exactly the reality of a musical instrument which therefore was indistinguishable from the one played live in the same room), but deepening the matter then you realize that it succeeded only with a single and very simple acoustic instrument.
I no longer remember which instrument, but it was a very simple acoustic one.
However, I've never listened an audio system able to play a rock concert recording, or a symphonic concert, or any other musical event that has only remotely approximate the "reality".
I think that for that, one has to add his imagination, and perhaps only then the whole "thing" really become an illusion.
Last edited: