
What stereo was meant to be
Stereo is a pretty basic scheme: two speakers let us hear musicians performing between them, and maybe a bit beyond. It was first invented and patented in 1931 by Alan Blumlein at EMI. In the first four or five decades since then, great strides were made in performance and fidelity, but around 1980 or so, stereo had pretty well matured as a playback medium.
There's one thing that wasn't solved, however: once you've got your ideal system set up and all the issues ironed out, then if you're being completely candid, the music still does not sound quite the same as live.
You might not care about this, because it can sound perfectly wonderful, but if you do care, bear with me. I'll be explaining the cause and how it was solved by a music loving rocket scientist who is also an audio researcher at Princeton University, Dr. Edgar Choueiri.
In early 2014, as usual, I was happily thinking my understanding of stereo playback was pretty good. That didn't last long after I got a call from John Chen, an audio pro and one of Professor Edgar Choueiri's close friends. John invited me to meet him at Dr. Choueiri's audio lab at Princeton University. Of note is that Dr. Choueiri's adjoining lab studies plasma propulsion.
This was early on in the Professor Choueriri's development of BACCH stereo purification. John felt that my speaker designs might be especially well suited to the BACCH process, and I was invited to Dr. Choueiri's audio lab at Princeton University to see what he was up to.
After a brief tour, Edgar handed me a pocket sized Bluetooth speaker and invited me to pair my phone with it and take a listen while holding it a couple of feet from my face. Here's what it looked like:

It was fascinating. I heard a full, wide soundstage with very decent fidelity, and the vibration of the speaker in my hands gave the impression of bass response -- Jawbone had licensed an early version of Dr. Choueiri's BACCH innovation from Princeton for its Jambox speakers!
My first thought was that high end home audio would be killed off by this technology, which would be bad news for me and my high end audio company, but I only let on that I was impressed. Little did I know yet that Edgar was nearly done developing the first version of BACCH for home audio, and he thought my speakers would work especially well with it for reasons I'll get into below.
Ultimately, in Dr. Choueiri's vast audio lab at Princeton, I was seated at desk before a pair of planar desktop speakers and given the complete BACCH treatment — microphones in ears to measure my own personal head transfer function while tone-sweeps came from the speakers. Music was then played through those speakers.
What I heard proved that the standard stereo playback we all know and love causes spatial confinement of the playback image. I suddenly heard what the microphones heard — musicians far to the sides, stage depth obvious, the sound of the recording venue all around me, even in some cases above and behind me. It was all there, and I knew listening would not be quite the same when I went home.
The world of stereo enhancement has many facets. There are image manipulation schemes that widen the soundstage using phase inversions, extra drivers (mainly in sound bars), or physical barriers. They're not super high in fidelity, and in some cases are inconvenient.
Multichannel playback is another approach, but buying and installing the additional channels is expensive and not easy, and while the result is more immersive, it still doesn't sound live.
There are also stereo signal filtering schemes that widen the image, but don't create a fully holographic sound field, and in the process, they create sonic artifacts and deviations in tonality. This works well for cinema, but not so well for music.
Another approach to making stereo genuinely work better for music is lighting up the room more heavily than is possible with unidirectional, speakers, even ones with wide dispersion This is the use of omnidirectional speakers and dipole speakers, and they do a great job at increasing spaciousness and immersion. By the same token, however, they can have the drawback of making the image more diffuse.
As it turns out, rather than adding sound, the optimal solution goes the other way.
It may seem counterintuitive at first, but BACCH works by subtracting sound -- not a fraction of the whole sound, but a particular part of the sound that occurs as an inevitable side effect of stereo.
This is the sound that leaks around your head from the left speaker and into your right ear and vice versa. It's called crosstalk, and when it's gone, the experience changes. The process of subtracting this sound is thus referred to as crosstalk cancellation (XTC), and the result is called Stereo Purification (SP).
BACCH-SP purifies stereo by filtering out the crosstalk that we never notice, but which is hiding the full musical experience, leaving only the pure music, just as the microphones heard it.
1. Interaural Crosstalk
The least subtle issue is that a substantial amount of the direct sound from your right speaker finds its way around your head and into your left ear, and of course vice versa. This is called crosstalk. This is not caused by sound that bounces off the walls, but sound that actually diffracts around your head. This part of the sound doesn't care that your head is in the way.
Why is that a problem? In real life, each sound source is produced by a single instrument. By contrast, in standard stereo, two speakers are simulating an array of instruments. Crosstalk causes your ears to hear both speakers at once, blurring the spatial cues that let your brain place sounds precisely.
BACCH eliminates that interference, so each ear hears what it’s supposed to, just as in real life. The result isn’t the “inside your head” effect of headphones, but a spacious, three-dimensional soundstage that opens up in front of and indeed all around you, where voices and instruments appear real and detached from the speakers.
Now, it's pretty obvious that a live performance is wider than the space between a pair of speakers, plus there's the sound of the performance hall all around you. When listening at home, however, we accept that a playback image mostly hangs between the speakers, and is augmented by wall splash to add realism. This wall splash is critical to the standard way of hearing stereo. Without it, recordings sound dry, and the sound is completely pinned between the speakers.
In addition, because of this normally critical wall splash, room ambiance covers up the recorded ambiance.
Another effect of standard stereo delivery is that the sense of height and depth is limited to what the speaker design can deliver.
When crosstalk has been markedly reduced, however, you'll hear the actual image fully in all three dimensions. If you've got a room with treated first refection points, and sufficiently directive speakers, this will even include the area behind your seat. This is because, in addition to a fully realized image, reduced crosstalk also lets the sound of the performance space better compete with your local room ambiance, allowing you to experience the original performance space.
When your eyes are closed, especially while a two-microphone or binaural recording is being played back, the sensation is very close to being where the microphones were.
If you're familiar with other systems that purport to serve this function, it's critical to understand that BACCH introduces absolutely no signal processing artifacts. The sound remains pure, unadulterated by resonances, distortions, tonality shifts, swishing sounds, or any other artifact.
To sum up this point, by cancelling crosstalk, the BACCH system unlocks the hidden glory in your stereo recordings, which is thrilling to me. It was mind bending the first time I heard it, and as Dr. Choueiri continues to improve and add features to his system, has gone from mind bending to indispensable. It sounds like your room has no speakers or walls. This transformation is a key element of the term stereo purification, which is what the SP in BACCH-SP stands for.
2. Local Room vs. Recorded Venue Ambiance
This next point has to do with further reducing the effect of your room along with any potential issues with speaker placement, regardless of room treatment.
With conventional stereo, your room alters the sound substantially. Furthermore, standard stereo depends on having a certain amount of local ambiance. Unfortunately, almost never does a room's reverberant and reflected sound match the venue sound on recordings, so in a sense, its presence can be considered a form of distortion.
One of the hidden treasures on your recordings is the original ambiance of the room or hall where the recording was made. BACCH's unique form of ambiance recovery completes the task of making your room disappear, thus transporting you to the original performance space. It's called ORC, for Optimal Room Correction.
The ORC software module was new as of 2024. It almost completely cancels your room's influence on the sound. It does this by equalizing the direct and indirect sound separately, and does it in both the time and frequency domains. This makes it unique in the world of automated or semi-automated room correction, and the result is uniquely satisfying.
You may be hearing for the first time just how amazingly beautiful and convincing recorded music can be, and not just those few very special recordings, but many of them.
3. Microphone vs. Ear Functional Disparity
There's a disconnect between how microphones work vs. how ears work. This accounts for a good bit of how recorded music can usually be easily identified as being played back vs. live.
The functional difference is that the tonal and loudness response of our ears depends on the direction that the sound is coming from. A microphone, however, is equally sensitive to sound coming from any direction, and in some cases, also equally sensitive to the loudness. This causes an inherent brightness or sometimes stridency in two-channel recordings where the engineer hasn't addressed the issue, which is a fairly common situation.
Most microphones take the sound received from the sides and rear, and make a signal that has the same tonal balance and loudness as if it were received from the front, namely where the performers usually are. Your ears hear things differently than that when hearing a live performance, so the sound that reaches the microphones from the side and rear walls of the performance space are reproduced in stereo differently than you'd hear them live. It's important to note that this ambient sound can be a large portion of the total sound heard live in a large hall, depending on how far you are seated from the stage.
As an aside, it's worth mentioning that binaural recordings do not have this problem, because binaural microphones receive sound similarly to our ears, and BACCH works especially well with them. Such recordings are rare in the general marketplace, but one studio that produces them exclusively is Chesky Records.
This issue can't be completely resolved for recordings made with conventional microphones, but is resolved to some extent by a second aspect of ORC, namely equalization to a psychoacoustically idealized target response curve. This goes a long way toward compensating for the ear vs. microphone disconnect. If you decide this compensation isn't for you, however, you can turn it off.
4. System Response Anomalies
Lastly, all systems have anomalies of one sort or another in frequency response and phase coherence, mainly from the speakers and their setup, but also potentially from other items in the signal chain. You may hear this as a tendency for your system to sound better on certain types of music, or to only come alive above a certain loudness, or the image seems recessed or forward, etc.
In addition to optimally implementing room correction, ORC will correct most system response anomalies. Until you hear it for yourself, it can be hard to believe how transformative this enhancement can be. Of course, you may prefer your system's tonality as it is, in which case you can just turn off that part of ORC.
The technically inclined may be interested to know that BACCH stands for Band-Assembled Crosstalk Cancellation Hierarchy, and is the subject of academic papers and a patent. It is the result of groundbreaking scientific work done by Professor Edgar Choueiri at Princeton University.
Here, finally, as you've probably already surmised, is the answer to what makes stereo sound recorded and reproduced: it's crosstalk. Without it, the sound can be so close to live that you will not need your imagination to believe it. Closing your eyes completes the illusion by hiding the obvious visual evidence to the contrary.

There have been other approaches to XTC over the decades, but they have all had one or more drawbacks. These range from unwieldy hardware to audible sonic artifacts to inadequate cancellation.
Not until BACCH has a high degree of crosstalk cancellation with no audible artifacts been possible. It accomplishes this with just two speakers arranged in a conventional manner.
Reviewers and experts in home and professional audio agree that this is a fundamental breakthrough, and they sing its praises. You can read some of the reviews on this page on the Theoretica Applied Physics web site. There's also a patent and an AES paper.
The BACCH-SP comes in three forms:
The BACCH4Mac is the least expensive implementation. It runs on an Apple Macintosh, typically a Mac Mini. It can become the central hub of your audio system, and comes with everything you need, even the computer, if you want. You can also run it on a MacBook.
The BACCH-SP, is a self-contained hardware implementation, available with or without analog inputs and outputs. Its computational hardware is different from a Mac Mini and optimized for the purpose.

The Grand BACCH-SP has all imaginable connectivity and options.

BACCH is based on analyzing the sound that comes directly from the speakers. Additional sound bouncing and scattering around the room will interfere with the analysis.
Because the direct sound is well defined, the way it interacts with your head is measurable and can be mathematically modeled. The measurement is done by placing microphones in your ears and playing a sweeping tone from each speaker.
The crosstalk part of the test sound from each speaker diffracts around your head and goes in the microphone in the opposite ear. It is meticulously measured, characterized, and as much as 99% subtracted by way of BACCH processing (20 dB cancellation). The room sound, however, escapes cancellation, because it is too disordered to characterize.
Wide dispersion speakers, especially in reverberant rooms, cause reduced XTC by increasing the sound that reaches the ears from the walls, floor, and ceiling compared to directly from the speakers. Such room excitation can be an important part of standard stereo presentation, but interferes with XTC.
Directive speakers minimize room surface splash, which leads to higher XTC by maximizing the amount of sound coming directly from the speakers compared to the room. This makes directive speakers preferable for BACCH, especially in untreated rooms.
You might have felt this coming, but truthfully and sincerely, JansZen speakers assist BACCH processing by inherently directing more of the sound straight toward the listener while maintaining tonality across a ±10° horizontal window. BACCH tracks head movement and compensates for timing/phase, loudness, and tonality changes, and the less it has to compensate for tonality differences, the better.
Just as importantly, because they are electrostatic, they also have very low distortion and incomparably fast and accurate transient response. This enhances the realism of the experience. When combined with BACCH, the sense of live performance is impressive.
Any of our models will enhance BACCH. We have, however, joined hands with Theoretica Applied Physics to offer a special version of our Valentina floor-standing speaker. It is optimized for use with BACCH systems that have been equipped with two of the optional modules, namely Crossover and ORC (Optimal Room Correction).
This is the speaker that Dr. Choueiri himself uses at home. It is available in a bundle with BACCH systems.
The Martina4BACCH is bi-amplified and completely analog. It has individual balanced analog inputs, one for the woofers and one for the electrostatic panels. The main difference from the Valentina A8 is that the BACCH does all the optimizing on top of its core XTC function. This eliminates the general purpose processing included in the Valentina A8, and provides the most transparent possible signal chain.
System bundles can be ordered through us or Theoretica Applied Physics, but dedicated web pages have not yet been posted yet as of Sept 1 2026.
JansZen will soon also be launching a version of our floor-stander called the Valentina XC, sold directly by us, where XC stands for eXternal Crossover. It has the same XLR inputs as the Martina4BACCH, but with additional gain that allows it to reach full volume with active external crossovers that are intended for home rather than studio use.
For owners of a BACCH4Mac who use their own DAC rather than the included RME studio DAC, the Valentina XC can be paired with a line level external crossover to provide an optimal BACCH experience. We will be posting a page about it once it is in production later this year (2026), along with a list of recommended crossovers. At some point, we will start producing an active crossover to go with it, for those who prefer a JansZen-only solution.
Conclusion
BACCH can be integrated into practically any stereo playback system, whether digital, analog, or mixed, and whether the speakers are passive or active. The transformation is immediate and satisfying.
Each BACCH system comes with unlimited, live tutorial assistance from a BACCH expert to make sure it is working perfectly and to your satisfaction. This is provided via Internet while your tutor takes remote control of your BACCH system and optimizes it with you, while speaking with you by telephone.
JansZen is pleased to be offering BACCH systems from our location in Columbus, Ohio, where we can provide demonstrations by appointment.
Please have a look at the Theoretica Applied Physics web site and then give us a call to answer your questions.