• 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 many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

BACCH4Mac Pro Edition: a report

oivavoi

Major Contributor
Forum Donor
Joined
Jan 12, 2017
Messages
1,721
Likes
1,940
Location
Oslo, Norway
I am glad oivavoi brought up the question of speakers directivity.

The 3D3A Lab at Princeton University, headed by Prof. Choueiri, who also heads Theoretica, the maker of BACCH-SP and BACCH4Mac (disclosure: I am a development engineer at Theoretica) has been carrying out an extensive study of speakers directivity, see https://www.princeton.edu/3D3A/Directivity.html and the technical papers referenced on that page. The webpage contains a dynamic chart that displays numerous plots of directivity measurements conducted on, 26 different loudspeakers, and counting, in the anechoic chamber of the 3D3A Lab. The speakers can be ranked (using the dropdown menu on that chart) according to various indices characterizing various aspects of directivity. Some of these indices have been recently proposed to the AES for standard characterization of speakers directivity, which has been badly lacking in the field of audio engineering with many manufacturers describe directivity haphazardly due to the lack of standard indices.

The question of how speaker directivity affects BACCH 3D Sound is answered in some detail in the 3rd Q&A on Theoretica’s FAQ page (https://www.theoretica.us/faq.html).

It has been a common wisdom in audio to favor more omni speakers (low directivity), as alluded to by soundArgument reference to "Harman's conclusions about listener preferences”. Indeed without BACCH, the reverb in recordings, no matter how prevalent, is not enough to give a sense of envelopment and spatial realism as the reverb is perceived to emanate solely form the speakers, therefore exciting the room with late and uncorrelated reflections, as omni speakers do, is desirable to avoid a sense of sterile and dry listening space. With BACCH this wisdom goes out the window. BACCH particularly excels in realistically reproducing the reverb in the recording, which is perceived by the listener to be as enveloping as it was in the recording venue, and not confined artificially to the location of the speakers. (In fact, we often demonstrate this to visitors by playing back reverb-rich acoustic recordings through BACCH in an anechoic chamber, where reverb tails as long as many seconds are reproduced with uncanny realism). Therefore more omni directional speakers are not needed for BACCH but can be used as long as the ratio of direct to reflected sound is not too low.

There is however one advantage of using more omni speakers with BACCH that is not mentioned in our online FAQ: the sweet spot for 3D listening with BACCH extends further in front and behind the listener with decreasing directivity. This advantage is not of critical value to most audiophiles, who typically sit in the sweet spot (where the filter is designed), unless they wish to share the 3D image with more listeners sitting behind them in a row.

William Guiracoche
Development Engineer
Theoretica Applied Physics

Thank you so much, William (if I may), this was incredibly useful! Grateful that you take the time to respond in full.

I read through the FAQ, and now I suddenly feel much wiser concerning what BACCH does. And I became even more curious. Do you know if anybody in Norway has bought the system yet? I would like to audition it!

(and I'm secretly hoping that prices will reach a range I can more easily afford once you really get the business going... :) )

One small comment concerning the statement in the FAQ that early reflecions are the enemy of good imaging. I think the jury is sitll out on this, as there has been very little research done on the topic. The only study I know of where imaging was the exact outcome of interest, was Choisel (2005). In that study, he didn't find any detrimental effects of early reflections on imaging. In his study the reflecion was 9.5 ms delayed though, which is more than what typically occurs in listening rooms. See the attached screenshot from the latest edition of Toole's book, where he summarizes the study (which is not available online).

XTxlOKP.jpg


It may nevertheless be that early reflections are detrimental to what Bacch does, of course!
 
Last edited:
OP
D

Dialectic

Major Contributor
Forum Donor
Joined
Sep 26, 2017
Messages
1,792
Likes
3,242
Location
a fortified compound
One small comment concerning the statement in the FAQ that early reflecions are the enemy of good imaging. I think the jury is sitll out on this, as there has been very little research done on the topic. The only study I know of where imaging was the exact outcome of interest, was Choisel (2005). In that study, he didn't find any detrimental effects of early reflections on imaging.

The problem with reflections for imaging is that, if you hear them, you room is adding to the sound. Yes, with non-BACCH systems, some side wall reflections can make the sound seem more spacious and widen the perceived stereo image. I suspect that the perceived benefit of sidewall reflections has caused wide dispersion to be one of the traditional hallmarks of good speaker design. But that effect is additive: it causes the listener to hear what is in the room, not what is on the recording.

With BACCH engaged, reflections on each channel of the recording are steered to the listener's corresponding ear. Like room reflections, these recorded reflections make the sound spacious, but because these reflections were captured by microphones in the space where the recording was made (or added later by a mixing engineer), the recorded reflections depict with striking realism the space in which the recording was made (or, on lesser recordings, give you an idea of what plugin the mixing engineer was using).

When I switch BACCH on in my system, my ears tell my brain that I'm in the space in which the recording was made. When I disengage BACCH, my ears tell my brain that I'm listening to my speakers in my room.

Admittedly, I'm using Dutch & Dutch 8Cs, not the most directional speakers but not omnis either. I got them because I live in Manhattan and needed controlled bass in a badly shaped listening room. Though they're not ultra-directional Sanders electrostats or Quad 57s, BACCH works brilliantly with them.
 
Last edited:

oivavoi

Major Contributor
Forum Donor
Joined
Jan 12, 2017
Messages
1,721
Likes
1,940
Location
Oslo, Norway
When I switch BACCH on in my system, my ears tell my brain that I'm in the space in which the recording was made. With I disengage BACCH, my ears tell my brain that I'm listening to my speakers in my room.

Thanks! I think I'm getting a clearer and clearer idea about what BACCH can do. Really hope to try it out some day.
 

Scott Borduin

Member
Forum Donor
Joined
Jan 16, 2018
Messages
57
Likes
140
Location
Portland OR USA
The closest thing you might have heard is perhaps multichannel, but the BACCH presentation is different.

With recordings of classical music, BACCH provides the acoustic envelopment that one gets from rear channels on multichannel recordings. One can engage and disengage BACCH with a click of the mouse while music is playing, and the first thing I notice when I switch from non-BACCH to BACCH is that BACCH reveals information about the acoustic space in which the stereo recording was made. One hears that information all around one's head with BACCH--on every single recording of classical music that I've tried. This information is almost completely lost in non-BACCH stereo playback, even with excellent systems. The only place I've ever heard this information before is in the concert hall. (I happen to have commercial recordings of a few concerts that I attended (mainly back when I lived in Chicago and went to CSO concerts all the time), so, while acknowledging that memory is imperfect, I have a "reference.")

There's other information that one hears in a BACCH setup that I've never heard in a non-BACCH system, stereo or multichannel. That information relates to what audiophiles would call "image specificity," but for clarity, let's call it placement of instruments in the soundfield. On recordings of classical music played back through BACCH (or BACCH-dHP, which is BACCH for headphones), this is what you hear coming from the side of the room where the speakers are placed. With orchestral recordings, for instance, reproduction of string sections in particular is dramatically enhanced with BACCH. With BACCH engaged, they sound like entire sections of massed string instruments, and the listener gets a more realistic idea of where the players are sitting with BACCH than without.

On recordings of chamber music, BACCH provides clearer information about where the players are sitting than one gets with conventional steewo. More startlingly, BACCH corrects the distortion to the sense of scale that results when one plays back chamber music recordings on a non-BACCH system. On non-BACCH systems, the scale of the chamber ensemble tends to follow the size of the system. Ever heard chamber music played back on the giant Focal speakers, big MBLs, or big Genesis speakers? I have, and, my goodness, chamber ensembles sound enormous on those systems. When I activate BACCH while listening through my 8Cs, the sense of scale is corrected. Chamber ensembles sound like chamber ensembles. Microphone placement of course makes a big difference, but this affects the listener's perception of distance to the musicians, not the sense of scale.

Hope this is helpful, though I'm not doing it justice.

The scary thing about BACCH is that you get used to it. I no longer have interest in listening to music played back through non-BACCH systems. When I listen to BACCH for a few hours (every listening session now turns into a few hours), I sometimes disengage BACCH for a reminder of what non-BACCH stereo sounds like, and that reminder is always painful. I'm not exaggerating.

I started writing a lengthy report on BACCH, then it got lost in browser-land somehow. Fortunately, you've already summarized a lot of my subjective observations very well! Some further points:

- My big concern with BACCH before I heard it was the possibility of another version of Floyd Toole's "circle of confusion", in which recordings are mastered with one set of standards, expectations, and equipment, and played back with another. To understand the concern, the aim of BACCH is crosstalk cancellation at your ears - to keep sound from the left speaker from being heard at your right ear, and vice versa. Now think of a simple stereo recording with a couple of omni mics in front of a performing ensemble. For any given instrument in that space, the two mics will vary in sound level, time of arrival at the mics, and the reflected and diffused sound picked up by each mic. All of these things are clues to the ear brain system as to the direction of sound origin, as well as the space in which the sound originates - and by reducing crosstalk, BACCH enhances all of those clues, which is what makes it sound so compelling. But, most recordings are not made with just a couple of mics - would recordings mastered for conventional stereo playback sound bizarre over BACCH? Fortunately, this seems not to be the case - I've yet to hear a recording, pop/rock or classical, which was not enhanced by BACCH.

- In addition to the dramatically expanded sense of space and ambience, and the size of the soundstage, I've been struck by the sound of individual instruments or voices within the soundstage. Instruments sound more vivid, three dimensional, real. This is beyond the usual descriptions of clarity or transparency: there is a richness of texture and timbre that is much more reminiscent of live music, and a big part of the addictive nature of BACCH for me. That said, if you are expecting the kind of laser-etched pinpoint imaging on a stereo canvas that is associated with some versions of high-end audio, it might take you a bit to get used to hearing instruments and voices more as they sound in real performance space.

- With head tracking enabled, the image remains quite stable as you move your head side-to-side, unlike conventional stereo. Although I must say, it is hard to break the decades-long habit of staying absolutely still in the center seat :)

- BACCH claims that the sound outside of the main listening axis sounds the same as conventional stereo. In my brief observations, this seems to be true. Previous attempts at crosstalk cancellation involved manipulating frequency response at the speaker, and thus degraded the sound outside the main listening axis - BACCH manipulates things in the time domain only, maintaining flat frequency response at the speaker.

- To the question of directive speakers, William pointed out that the primary requirement is to minimize first reflections, by speaker design or absorption, or both. My first week with BACCH, I was using a quite directional speaker (the JansZen Valentia electrostatic). I then switched to Revel Salon2, which have very wide directivity by design (and are also not time-aligned, which I wondered about as well). The subjective impression with the Salons was nonetheless quite superior - perhaps because they are very good speakers in general - but note that I am aggressively absorbing first reflections from side wall, ceiling, front wall, and back wall. The sound with BACCH is, however, the very opposite of a dry or overdamped room effect, because BACCH is increasing the perception of the natural reflections and ambience in the recording.

- On the caveats side, BACCH did some not very nice things to the bass response in my room. Due to modal effects, I can't run any main speaker full range in my room with tolerable bass response; currently I am running three subs with hand tuned EQ and delay to give even bass response, then crossing over to the mains at 80 Hz. For whatever reason, BACCH seems to exaggerate a modal effect around 80 hz in my room, as well as a dip around 100-130 hz, which are quite audible. Why this is true is not clear to me, even after an email exchange with Dr. Choueiri. BACCH provides a 31 band graphic EQ, but I found it to have insufficient resolution to deal with bass problems (as is usually the case with graphic EQ). I created another DSP preset in Roon, which has nice parametic EQ facilities, to deal with this issue. So, if you're using EQ to deal with the bass, you may need to re-tune. And if you're not using EQ, your bass response is almost certainly very bad anyway and you might not notice :)

Scott
 
OP
D

Dialectic

Major Contributor
Forum Donor
Joined
Sep 26, 2017
Messages
1,792
Likes
3,242
Location
a fortified compound
On the caveats side, BACCH did some not very nice things to the bass response in my room. Due to modal effects, I can't run any main speaker full range in my room with tolerable bass response; currently I am running three subs with hand tuned EQ and delay to give even bass response, then crossing over to the mains at 80 Hz. For whatever reason, BACCH seems to exaggerate a modal effect around 80 hz in my room, as well as a dip around 100-130 hz, which are quite audible. Why this is true is not clear to me, even after an email exchange with Dr. Choueiri. BACCH provides a 31 band graphic EQ, but I found it to have insufficient resolution to deal with bass problems (as is usually the case with graphic EQ). I created another DSP preset in Roon, which has nice parametic EQ facilities, to deal with this issue. So, if you're using EQ to deal with the bass, you may need to re-tune. And if you're not using EQ, your bass response is almost certainly very bad anyway and you might not notice :)

I've also had an issue with bloated bass with BACCH but have had some success in mitigating it. While BACCH improves the spatial presentation of essentially all recordings, I have noticed that it can alter the bass in such a way that changes tonal balance. Many solo piano recordings sound a bit leaner in the bass with BACCH than without; this issue is small and has not bothered me. More problematic is boosted response around 100 Hz that I've noticed on some orchestral and large-ensemble jazz recordings.

Initially, I used BACCH on my Dutch & Dutch 8Cs with only boundary adjustments set through the D&D app. With this setup, the 100 Hz boost was especially problematic. Since then, I have done some REW measurements with a very cheap Behringer mic and have set some adjustments in the 8Cs' EQ filters, which can be set with very precise frequency, Q, and amplification/attenuation values. These adjustments mitigated the boost around 100 Hz, though the 8Cs' limits of five filters for each speaker limits flexibility. (I experimented a bit with the 31-band EQ mentioned by @Scott Borduin above, but, because I worry about doing more harm than good with EQ, I prefer more precision in my adjustments.) Some bloated bass around 100 Hz remains, and I don't think it's a problem that is easily resolved. Balancing this problem with all of the pluses of BACCH, I can happily live with it.

I suspect that some of BACCH's changes to perceived sub-200 Hz response occur because recordings are engineered on conventional, non-BACCHed speakers. Thus, recording engineers are accustomed to hearing certain constructive and destructive interference between the two channels--especially in the bass, where nearly all speakers become omnidirectional. Some of this interference is reduced by BACCH, perhaps especially in the lowest frequencies where BACCH operates. This could account for some of the unpredictable things I've heard in the bass with BACCH.

If I can pull myself away from listening, I might experiment with setting up some BACCH filters that use a higher cutoff frequency. The default is 94 Hz, but the software provides some options for going higher.
 

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
BASS AND SOUND QUALITY

«In assessing the factors contributing to sound quality (...) it was shown that about 30% of the overall rating is contributed by factors related to low-frequency performance».

Toole, latest edition page 215.

It would be nice to hear from BACCH on this.

Have they (BACCH) done blind tests on their own?
 
OP
D

Dialectic

Major Contributor
Forum Donor
Joined
Sep 26, 2017
Messages
1,792
Likes
3,242
Location
a fortified compound
BASS AND SOUND QUALITY

«In assessing the factors contributing to sound quality (...) it was shown that about 30% of the overall rating is contributed by factors related to low-frequency performance».

Toole, latest edition page 215.

It would be nice to hear from BACCH on this.

Have they (BACCH) done blind tests on their own?

All of Toole's research was done without BACCH. These percentages purporting to tell us what's important in the listening experience go out the window with BACCH.
 

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
All of Toole's research was done without BACCH. These percentages purporting to tell us what's important in the listening experience go out the window with BACCH.

Do you suggest BACCH makes previous audio research redundant?

In this case, we talk about something as essential as bass, lower-frequency reproduction which is the center of so many audio discussions (and so much audio research).
 
OP
D

Dialectic

Major Contributor
Forum Donor
Joined
Sep 26, 2017
Messages
1,792
Likes
3,242
Location
a fortified compound
Do you suggest BACCH makes previous audio research redundant?

In this case, we talk about something as essential as bass, lower-frequency reproduction which is the center of so many audio discussions (and so much audio research).

I'm not saying BACCH makes all earlier research redundant. But these relative percentages concerning what is important from earlier research are no longer valid.

Toole et al. did not have BACCH when they were assessing listener preferences. If they had had BACCH, their findings on speaker directivity, relative listener preferences, and some other matters would have been very different.
 

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
I'm not saying BACCH makes all earlier research redundant. But these relative percentages concerning what is important from earlier research are no longer valid.

Toole et al. did not have BACCH when they were assessing listener preferences. If they had had BACCH, their findings on speaker directivity, relative listener preferences, and some other matters would have been very different.

This is an interesting claim, hypothesis. But it may belong in another thread dedicated to this theme?
 

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
The questions of BACCH are interesting because it is about hifi, i.e. being true to the source.

Room compensation DSP means that the original signal is altered to arrive at the listening position in intact, true form or state.

DSP like BACCH means altering the original signal to arrive at the listening position in altered form or state.

Which begs the question: When does devation from hifi occur? When room compensation sets in or when BACCH or similar aims to give the listener an altered signal?

Is hifi a practical term in 2018?
 

Floyd Toole

Senior Member
Audio Luminary
Technical Expert
Industry Insider
Forum Donor
Joined
Mar 12, 2018
Messages
383
Likes
4,147
Location
Ottawa,Canada
Hello all, I just dropped in and I see my name mentioned in a context that brought back strong memories from around 1995, 23 years ago. In spite of the publicity given to the recent commercialization of crosstalk cancelling systems, the idea is not new. Not long after I joined Harman I negotiated a licensing arrangement to employ the Cooper-Bauck "transaural" crosstalk cancelling method into a product called VMAX - Virtual Multi-Axis. At that time the burgeoning market was computer audio, and the product appeared in several incarnations of computers (mainly Compaq) and later in a much simplified form in Harman-Kardon receivers. However, it was my hope that a high quality version could be marketed. Sadly it was not to be. Management of the period could not find a way to incorporate "software" as a product in our offerings - licensing being an obvious possibility. It drifted into nothingness.

However, for several months we maintained a fully implemented, no compromise, demonstration of its capabilities for ourselves and visitors to experiment with. The loudspeakers were at +/- 10 deg., on either side of a then popular TV, listening at about 10 ft, in an acoustically well-damped (but certainly not dead) room, using conventional cone & dome speakers. The goal was to deliver phantom 5 channel home theater from two loudspeakers. This began with cancelling the crosstalk using the Cooper-Bauck methodology, and then creating binaural cues simulating loudspeakers at specific locations. The starting point was +/- 30 deg - conventional stereo. Immediately the benefits, compared to real loudspeakers at those locations, could be heard. The real +/- 10 deg. loudspeakers were not localizable due to the crosstalk cancellation and the sounds that could be localized were perceived at directions and distances determined by information in the recordings. Hard panned L & R images were at their correct locations, but there was no sense of them coming from a loudspeaker, and certainly not at 10 ft. Some recordings were not impressive, but the good ones were simply amazing - the signal processing in the mix was a large factor. Symphony orchestras had realistic width and depth, and opera singers could sometimes be heard at distances that put them outside the walls of the room. I spent a lot of time listening to this system, concluding that at its best it was the most impressive stereo imaging I had ever heard. Perceived distance and ambiance were the distinguishing characteristics. The sound quality was definitely HiFi. There may have been compromises, but they were not at all distracting. There was a sweet spot of course, but leaning out of it simply caused the sound to progressively and gracefully degrade to conventional stereo from the real +/- 10 deg loudspeakers.

Pink noise could be convincingly steered to +/- 30 deg, +/- 90 deg and of course center.

At a trade show we mounted a demonstration of real vs. phantom 5.1 home theater. Long lines of people waited to hear it, and we collected reactions when they left. About half of the listeners said that they had no preference, but of the half that expressed a preference the phantom system won. When asked why, the popular answer could be interpreted as "distance" - the sounds came from much farther away than the real loudspeakers and the images were "softer".

The key difference between what I have described and what I think you are hearing through the BACCH system is that VMAX simulated loudspeakers at the correct stereo locations. Of course we listened to the "naked" crosstalk canceller, and what we heard parallels some of the descriptions I have read in this thread. When there are no phantom loudspeakers to provide directional anchors, almost anything is possible, and much of it is far from what the artists and recording engineers intended - the "soundstage" is very fluid, and envelopment can be profound. In this mode it is really a sound-effects generator and opinions of like or dislike will predictably vary. Please correct me if I am wrong.

I explain the process in the second part of the attached white paper that I wrote in 1996. The first half discusses multichannel audio and it obviously needs updating, but the second half is still valid.

Cheers, Floyd
 

Attachments

  • Direction and Space_ScreenQuality.pdf
    800.3 KB · Views: 254

oivavoi

Major Contributor
Forum Donor
Joined
Jan 12, 2017
Messages
1,721
Likes
1,940
Location
Oslo, Norway
No way. We have been touched by God!!

That's actually how I feel right now. Warm welcome to the forum, dr. Toole.

And the fact that you think that the operating mechanisms in BACCH have a sound psychoacoustic basis makes me want even more to try it out.
(I will read your paper tomorrow, thanks for posting it. Now it's night here in Europe)
 

Scott Borduin

Member
Forum Donor
Joined
Jan 16, 2018
Messages
57
Likes
140
Location
Portland OR USA
Hello all, I just dropped in and I see my name mentioned in a context that brought back strong memories from around 1995, 23 years ago. In spite of the publicity given to the recent commercialization of crosstalk cancelling systems, the idea is not new. Not long after I joined Harman I negotiated a licensing arrangement to employ the Cooper-Bauck "transaural" crosstalk cancelling method into a product called VMAX - Virtual Multi-Axis. At that time the burgeoning market was computer audio, and the product appeared in several incarnations of computers (mainly Compaq) and later in a much simplified form in Harman-Kardon receivers. However, it was my hope that a high quality version could be marketed. Sadly it was not to be. Management of the period could not find a way to incorporate "software" as a product in our offerings - licensing being an obvious possibility. It drifted into nothingness.

However, for several months we maintained a fully implemented, no compromise, demonstration of its capabilities for ourselves and visitors to experiment with. The loudspeakers were at +/- 10 deg., on either side of a then popular TV, listening at about 10 ft, in an acoustically well-damped (but certainly not dead) room, using conventional cone & dome speakers. The goal was to deliver phantom 5 channel home theater from two loudspeakers. This began with cancelling the crosstalk using the Cooper-Bauck methodology, and then creating binaural cues simulating loudspeakers at specific locations. The starting point was +/- 30 deg - conventional stereo. Immediately the benefits, compared to real loudspeakers at those locations, could be heard. The real +/- 10 deg. loudspeakers were not localizable due to the crosstalk cancellation and the sounds that could be localized were perceived at directions and distances determined by information in the recordings. Hard panned L & R images were at their correct locations, but there was no sense of them coming from a loudspeaker, and certainly not at 10 ft. Some recordings were not impressive, but the good ones were simply amazing - the signal processing in the mix was a large factor. Symphony orchestras had realistic width and depth, and opera singers could sometimes be heard at distances that put them outside the walls of the room. I spent a lot of time listening to this system, concluding that at its best it was the most impressive stereo imaging I had ever heard. Perceived distance and ambiance were the distinguishing characteristics. The sound quality was definitely HiFi. There may have been compromises, but they were not at all distracting. There was a sweet spot of course, but leaning out of it simply caused the sound to progressively and gracefully degrade to conventional stereo from the real +/- 10 deg loudspeakers.

Pink noise could be convincingly steered to +/- 30 deg, +/- 90 deg and of course center.

At a trade show we mounted a demonstration of real vs. phantom 5.1 home theater. Long lines of people waited to hear it, and we collected reactions when they left. About half of the listeners said that they had no preference, but of the half that expressed a preference the phantom system won. When asked why, the popular answer could be interpreted as "distance" - the sounds came from much farther away than the real loudspeakers and the images were "softer".

The key difference between what I have described and what I think you are hearing through the BACCH system is that VMAX simulated loudspeakers at the correct stereo locations. Of course we listened to the "naked" crosstalk canceller, and what we heard parallels some of the descriptions I have read in this thread. When there are no phantom loudspeakers to provide directional anchors, almost anything is possible, and much of it is far from what the artists and recording engineers intended - the "soundstage" is very fluid, and envelopment can be profound. In this mode it is really a sound-effects generator and opinions of like or dislike will predictably vary. Please correct me if I am wrong.

I explain the process in the second part of the attached white paper that I wrote in 1996. The first half discusses multichannel audio and it obviously needs updating, but the second half is still valid.

Cheers, Floyd

Hi Dr. Toole, thanks for joining the discussion and I hope we hear more from you here! Your work has been invaluable to me personally and the community at large, even though some of them don't know it :)

For reference, here is a paper from Dr. Edgar Choeuri, the Princeton professor who led the research into the BACCH algorithms and runs the BACCH company that commercializes the technology:

https://www.princeton.edu/3D3A/Publications/BACCHPaperV4d.pdf

Too long/didn't read: BACCH implements pure crosstalk calculation in a different way than Cooper-Bauck, that does not require severe manipulation of frequency response and the attendant limitations on dynamic range and speaker placement.

Anyway, the base BACCH functionality does not implement virtual speaker positioning, to my knowledge anyway (see below for a note on the advanced version). It just does crosstalk cancellation, and lets the effect fall where it might. The BACCH FAQ, and Dr. Choueiri himself, say that this leads to an enhanced effect with a large number of recordings, but can sound odd with a smaller percentage, particularly where aggressive close or spot mic technique is used. In my auditioning, I have noted one or two moments in classical recordings where the spot mics are more obvious than in conventional stereo (BACCH allows to switch between processed and unprocessed on the fly). These moments are by no means sufficiently bothersome to make me want to forgo the processing. More surprisingly, I have also strongly preferred the effect with all pop recordings I've tried, although I'm too old to have much modern pop in my collection :)

Of course, the effect is not what the producer or artist explicitly intended. How could it be, given that their auditioning circumstances did not include this processing? But of course the same argument can be made for listening on well-engineered speakers, equalization, tone controls, or upmixing to multi-channel. We have to deal with the recordings we've got, not the recordings we wish we had :) I used a Meridian 861 for many years, and almost always chose to use the Trifield upmixing algorithm, even though the center image and timbre in particular was fairly different (and to my ears superior) than anything which comes through in conventional 2 channel. BACCH is audibly more dramatic than that, but to my ears in a good way.

Interestingly, given your comments above on the XTC system you experimented with, the "Pro" version of BACCH includes a binaural authoring environment for creating almost unlimited virtual speaker layouts, with or without virtual rooms, and mapping and mixing channels from various sources into those virtual speakers. It is designed for recording and mastering from multi-channel, Higher Order Ambisonics, etc into binaural, but I would think creating a processing environment to map 2 channels into a virtual +-30 degree speaker array would be trivial. And of course mapping multi-channel into virtual speaker layouts. Given your interest in that subject, you might find this interesting. https://www.theoretica.us/bacch-dsp/

Hope this is helpful, and thanks again for your participation here.

Scott
 

Floyd Toole

Senior Member
Audio Luminary
Technical Expert
Industry Insider
Forum Donor
Joined
Mar 12, 2018
Messages
383
Likes
4,147
Location
Ottawa,Canada
I think I have a slight. bias because in the early phases of his work, Dr. Chouieiri promoted it as if he had discovered binaural sound and crosstalk cancellation. There was a "gee whiz" quality to the promotions. That was clearly not the case. Since then, he and his aids have obviously gone further, and I have no problems with this. The concepts are valid, and with today's computing power, many things are possible - for a single listener willing to listen in a sweet spot or through headphones.

On the research side, using real-time binaural head-position tracking, binaural recordings are now used as laboratory apparatus for "transporting" listening experiences. (called Binaural Room Scanning). It is a "you are there" experience enabling listening tests to be done in different places at different times through calibrated headphones. It is powerful for circumstances that do not permit real-time double-blind listening. I discuss it in my book.

In the end, I have long ago decided that multichannel audio is a preferable option for me because it is capable of providing comparably good sound and spatial impressions for multiple listeners - I frequently enjoy company, sharing an experience. To date, the most impressive demo I have experienced was Auro3D playing concert and organ performances created for that format by the inventor. One could walk around the playback room and remain in the recorded venue. But even more modest multichannel music recordings can be impressive. I hope I live long enough to experience more of this :). It is a pity that we must rely on the film industry to provide us with the playback technologies. The music side of the audio business has dragged its feet all the way. I recall when stereo was introduced that "experts" said it was not necessary. When 5.1 came along it was considered "only for movies" by some. And so it seems to continue . . .

Many critics of multichannel music playback have systems in which the money has been spent on the L&R, with less than neutral speakers in other locations. The requirements for timbral accuracy are the same for all channels. Two ears and a brain know the difference. More channels and loudspeakers mean fewer phantom images - all of which are timbrally and spatially compromised.

It is a good thing that human listeners are adaptable, tolerant and with a brain that fills in many gaps :). Meanwhile there are still live concerts.
 

Scott Borduin

Member
Forum Donor
Joined
Jan 16, 2018
Messages
57
Likes
140
Location
Portland OR USA
I think I have a slight. bias because in the early phases of his work, Dr. Chouieiri promoted it as if he had discovered binaural sound and crosstalk cancellation. There was a "gee whiz" quality to the promotions. That was clearly not the case. Since then, he and his aids have obviously gone further, and I have no problems with this. The concepts are valid, and with today's computing power, many things are possible - for a single listener willing to listen in a sweet spot or through headphones.

On the research side, using real-time binaural head-position tracking, binaural recordings are now used as laboratory apparatus for "transporting" listening experiences. (called Binaural Room Scanning). It is a "you are there" experience enabling listening tests to be done in different places at different times through calibrated headphones. It is powerful for circumstances that do not permit real-time double-blind listening. I discuss it in my book.

In the end, I have long ago decided that multichannel audio is a preferable option for me because it is capable of providing comparably good sound and spatial impressions for multiple listeners - I frequently enjoy company, sharing an experience. To date, the most impressive demo I have experienced was Auro3D playing concert and organ performances created for that format by the inventor. One could walk around the playback room and remain in the recorded venue. But even more modest multichannel music recordings can be impressive. I hope I live long enough to experience more of this :). It is a pity that we must rely on the film industry to provide us with the playback technologies. The music side of the audio business has dragged its feet all the way. I recall when stereo was introduced that "experts" said it was not necessary. When 5.1 came along it was considered "only for movies" by some. And so it seems to continue . . .

Many critics of multichannel music playback have systems in which the money has been spent on the L&R, with less than neutral speakers in other locations. The requirements for timbral accuracy are the same for all channels. Two ears and a brain know the difference. More channels and loudspeakers mean fewer phantom images - all of which are timbrally and spatially compromised.

It is a good thing that human listeners are adaptable, tolerant and with a brain that fills in many gaps :). Meanwhile there are still live concerts.

Agreed, BACCH is pretty sophisticated at promoting themselves compared to most academic spinouts I have experience with. On the other hand, very unlike other academic spinouts I've experenced, they are also better at customer support than any other high end company I've dealt with.

In any case, for more than 20 years I've been attracted by the idea of multi-channel, and never quite impressed by the results. It seemed the soundstage was "stretched" to the rear, creating some level of surround ambience, but never quite realistic in terms of what it did to envelopment from the frontal sound stage. I was quite taken with your observations that envelopment channels were more effective at +-60 degrees, and I'm currently in process on putting together a 9.1.4 system, using your 30 - 60 - 105 - 120 layout, based on Revel Salon2/Voice2 LCR, and JBL 705i surround/heights, with a JBL SDP-75 controller using anechoic speaker corrections to match timbre as closely as possible, and decorrelating the 60-105 channels.

One of the interesting things to me about BACCH or similar algorithms, independent of my subjective evaluations, is that there is a disproportionate investment these days into binaural sound as a result of the technology industry focus on virtual or augmented reality (VR/AR). Whether it is AES meeting topics, or commercial announcements at CES and the like, there are a lot of smart people working on this topic right now. I had an interesting conversation with Dr. Choueiri on this topic, and his basic observation, which I agree with, is that there are a lot of smart, sincere people working in the audio industry working on advancing the existing paradigms. Squeezing one more bit of resolution out of a DAC, lowering the noise floor of an amp, perfecting pistonic motion of a driver, etc, etc. Meanwhile, there is a whole new generation of researchers and enthusiasts who are pursuing audio not as an end in itself, but as an element of immersion in virtual reality. They aren't interested in incremental improvements of current stereo or multi-channel representations of reality, they are thinking from the ground up about the problem of simulating or reproducing true auditory reality. They no more expect improvements in stereo to fix the 3D audio reality gap than they would expect improvements in flat panel display resolution to fix the 3D visual reality gap. IMO, these people are driving a revolution in audio reproduction as big or bigger than the revolution started by the telephony researchers who figured out that you could transmit voices with digital encoding instead of frequency modulation.

And yes, thank god, there are still live concerts!

Edit: removed comments not appropriate to thread.

Scott
 
Last edited:

Floyd Toole

Senior Member
Audio Luminary
Technical Expert
Industry Insider
Forum Donor
Joined
Mar 12, 2018
Messages
383
Likes
4,147
Location
Ottawa,Canada
Scott: We seem to be very much on the same page, even to our envisaged immersive sound systems. The only difference in mine is the use of Revel Gem2s as surrounds, including +/- 60 deg locations, and the likely be the use of Revel elevation speakers - not timbrally better or worse than the JBL 705i alternatives. The SDP-75, yet to be delivered, will allow for anechoic response refinements, sound field managed bass, and many interesting experiments. I attach a photo of the present situation. The acoustical "secret ingredient" is the scattering from bronze statues of the female form :)

As explained in my new book, I delight in the challenge of delivering superb sound without cumbersome and ugly acoustical devices - a "stealth" approach is more rewarding to me. Not having a "dedicated" home theater is a deliberate choice for our life style. Those suspecting that wall mounting of the Salon2s is a problem (and it is a popular challenge) should note that that surface is deliberately very irregular - it is really a low/mid-frequency scattering surface, and adjacent boundary issues are minimal. Having enjoyed the interim system for a while, it is my impression, and that of other trusted ears that have heard it, that the inverted speakers might sound better than they do in conventional floor-standing mode. The soundstage is at the right height (determined by the height of the tweeter) and the sound is remarkably open and clean. They also liberate several square feet of expensive real estate :)

Yes, the current VR/AR activities are bound to generate new knowledge and supporting technologies. The reward will be fascinating experiences for gamers, personal movie goers, and video concerts. It is a huge asset to have credible visual cues to help guide our aural perceptions. Recordings of live concerts should be especially persuasive with surround visuals and audio. My traditional 10 ft screen and surround audio system provides superb entertainment from the handful of excellent video concerts (most are mediocre, sadly), and it does it for a room full of people.
 

Attachments

  • The Toole entertainment room (smaller).jpg
    The Toole entertainment room (smaller).jpg
    512.1 KB · Views: 422

Scott Borduin

Member
Forum Donor
Joined
Jan 16, 2018
Messages
57
Likes
140
Location
Portland OR USA
Scott: We seem to be very much on the same page, even to our envisaged immersive sound systems. The only difference in mine is the use of Revel Gem2s as surrounds, including +/- 60 deg locations, and the likely be the use of Revel elevation speakers - not timbrally better or worse than the JBL 705i alternatives. The SDP-75, yet to be delivered, will allow for anechoic response refinements, sound field managed bass, and many interesting experiments. I attach a photo of the present situation. The acoustical "secret ingredient" is the scattering from bronze statues of the female form :)

As explained in my new book, I delight in the challenge of delivering superb sound without cumbersome and ugly acoustical devices - a "stealth" approach is more rewarding to me. Not having a "dedicated" home theater is a deliberate choice for our life style. Those suspecting that wall mounting of the Salon2s is a problem (and it is a popular challenge) should note that that surface is deliberately very irregular - it is really a low/mid-frequency scattering surface, and adjacent boundary issues are minimal. Having enjoyed the interim system for a while, it is my impression, and that of other trusted ears that have heard it, that the inverted speakers might sound better than they do in conventional floor-standing mode. The soundstage is at the right height (determined by the height of the tweeter) and the sound is remarkably open and clean. They also liberate several square feet of expensive real estate :)

Yes, the current VR/AR activities are bound to generate new knowledge and supporting technologies. The reward will be fascinating experiences for gamers, personal movie goers, and video concerts. It is a huge asset to have credible visual cues to help guide our aural perceptions. Recordings of live concerts should be especially persuasive with surround visuals and audio. My traditional 10 ft screen and surround audio system provides superb entertainment from the handful of excellent video concerts (most are mediocre, sadly), and it does it for a room full of people.

I had seen pictures of this room before, and I am envious of both its size and visual aesthetic. And I appreciate your willingness to share details. Our media room is a modest 20 x 14 x 9, hence my preference for the more compact 705i surrounds. And bass trapping is an absolute must for me - with these dimensions, multi-sub cancellation of first order modes only fixes part of the problem. Higher order modes are in fact the more problematic issue in my situation.

Inverting the Salons was a clever idea, one that I will probably not try to replicate, however. Just maneuvering them around to find the most balanced response without causing oneself physical injury is enough for me :) I may try putting the Voice2 on a motorized stand to more nearly match the elevation of the Salons when listening to multi-channel music, but only if the variation is noticeable.

I am looking forward to having a great dual-purpose system, with BACCH for dedicated 2 channel listening, and high-end surround for multi-channel music, movies, and parties.

On the AR/VR side, I think the most interesting developments might be less on the reproduction than the recording side. Recording techniques for immersive audio, like Higher Order Ambisonics, lend themselves well to multiple advanced reproduction schemes, from binaural over headphones to binaural over speakers to 3D multi-channel.

Edit: Keep us posted, here or elsewhere, on your experiments with the SDP-75. Having read the manual over closely, I am excited by the possibilities. Only downside I've noticed: no dedicated tone controls. This will have to be done with extra presets, I guess.

Scott
 
Last edited:
Top Bottom