• 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

Dialectic

Major Contributor
Forum Donor
Joined
Sep 26, 2017
Messages
1,738
Likes
3,090
Location
a fortified compound
Continuing the recent theme of putting my money where my mouth is, I recently added BACCH technology to my system by investing in the BACCH4Mac software and a late 2012 Core i7 Mac Mini to run it. The early results with the Dutch & Dutch 8Cs are impressive. A report of my observations is forthcoming.

5kDETSD9irtyQDcj-HgrtwvZT19JpJUK71mtroHvyI5sy_JC8IZsIyxg7cBJeekDQKibTg_jZLpuQVRY69QwptodNZmDPBO9WRzGRUHpiE7T1PfyhPbnfSeKGgUThwccSegXGcIB4w=w1046-h784-no


Initial observations
I've now done many hours of listening to familiar recordings.

Pre-BACCH
With the Dutch & Dutch 8Cs, Imaging without BACCH is surprisingly precise, and bass is subjectively deep and well-controlled because of the 8Cs' DSP. I've measured a peak at 31 Hz and a dip at 58 Hz that are troubling, but subjectively, the system sounds impressive and free of obvious coloration without BACCH. The center image is solid, and the image extends between the loudspeakers, which I currently have in a roughly 60-degree equilateral triangle configuration. The speakers are slightly toed in and, as recommended by D&D, close to the front wall, with the distance programmed into the speakers' DSP.

With BACCH
After initial setup, the first test prescribed by Theoretica was to listen to one of the tracks included in iTunes on the Mac Mini, a Chesky binaural recording of Wycliffe Gordon performing New Orleans-style jazz. I was told I'd hear a trombone enter at approximately 9:00 to my ears. I heard it at a position more like 10:00, but the effect was still stunning, as the trombone did not seem to emanate from the speakers. Indeed, no sounds seemed to be emanating from the speakers. If you allow me to use an audiophile cliché, the speakers in front of me seemed to disappear.

The effect is similarly striking on other binaural recordings, but the bulk of my listening involves classical music. By and large, classical music is recorded using either the spaced omni (A-B) technique or the mid-side technique. Some older recordings, particularly those made by Decca Records, use the Decca tree. All of these techniques generate spatial cues that BACCH can enhance, and almost universally, recordings of classical music are enormously enhanced by BACCH.

On these recordings, activating BACCH enhances the sense of space, causing the speakers to seem to disappear. The soundfield tends to span beyond the sides of the speakers and extends far behind the speakers. Hearing the distance to the woodwinds and the tympanis is possible on nearly all orchestral recordings. String sections sound like fully fleshed out sections comprising individual performers, not nebulous agglomerations. The string sections are, moreover, much more easily distinguishable with BACCH than without. The first violin and second violin sections have never been easier to differentiate. I do not hear many sounds around me or behind me, but my experience listening to two other BACCH systems and my conversation with Professor Choueiri suggests that I may hear more such sounds if I listen in a room with fewer early reflections.

Recordings of chamber music are also strikingly different. On conventional systems, the sense of scale on chamber recordings tends to be commensurate with the scale of the system. I recall hearing chamber music on the Focal Grande Utopia EMs in a huge room; it sounded huge (and phony). With BACCH, the scale of chamber music is appropriately small; the soundfield on my preferred chamber recordings (most on the Philips label) is condensed and does not extend beyond the speakers. Nonetheless, the individual performers, especially in string ensembles, are more easily distinguishable than ever before.

I have not noticed any tendency of BACCH to homogenize recordings. The sense of space on different recordings is strikingly different. My non-audiophool spouse has observed that it feels strange sitting in the room while listening to the system with BACCH because, in contrast to a conventional system, the BACCH system generates sound totally at odds with the surrounding space.

Some jazz recordings have produced good results, too. I'll listen to more jazz before expressing any observations.

Because of the way in which popular music is recorded, popular music reproduced with BACCH processing is a mixed bag. The Pink Floyd albums from DSOTM to The Wall sound spectacularly three-dimensional. I also have enjoyed listening to Captain Beefheart's Trout Mask Replica with BACCH. With some other popular music, BACCH makes too obvious how it was recorded; the effect can, perhaps surprisingly, be unpleasant.

Early conclusions
At this point, I am a fan. Though the results I have experienced are not as remarkable as those I heard in Mr. Lam's system, I continue to believe that BACCH represents the way forward in music reproduction. As my exchange with @dallasjustice below may demonstrate, this is still a nascent technology, and integration with existing computer audio systems--particularly the complex systems used by members of this forum--may be difficult. Nevertheless, I've experienced few difficulties in my first few days with BACCH.

I still do not believe my system is fully optimized for BACCH in terms of speaker placement or BACCH calibration. I will adjust both of those variables and report here on the results.

I still haven't tried listening through headphones. I'll report on the headphone experience after I have a little time to do so.


Update 2/27
I received some instruction last night from the Theoretica team on headphone playback. The effect is astonishing and addictive. Head tracking with headphones and the BACCH-HP filters allows the ear-brain system to perceive a sonic image away from the head through headphones.

When headphone playback was first engaged, I was certain my speakers were playing at full blast. Superficially, this function resembles that of the Smyth Realiser, a device that I have never heard, but with the BACCH-HP filters, the additional spatial cues revealed by BACCH processing remain audible. This feature will receive heavy use.

I also received some instruction on the BACCH 3D Mixer, the functionality of which is extraordinary and a tool that I wish I had had when I was making orchestral recordings. The Mixer enables the mixing engineer to locate sound sources in three-dimensional space in creating a multitrack recording. If you have ever mixed a recording using a large number of microphones that supplement a main stereo pair, I think you will understand why this functionality is revolutionary: in conventional DAW software, mixing spot mics with the main stereo microphones is an imprecise affair, involving the application of delay and use of the mixer's balance/pan control to locate the sound in the mix. With the 3D Mixer, the position of the spot microphone can be set exactly in software, limiting the extent to which spot mics will interfere with the spatial cues received by the stereo pair.

The 3D Mixer also allows one to mix a multichannel recording into a two-channel, quasi-binaural mix. I will report on that functionality when I have time.
 
Last edited:

Scott Borduin

Member
Forum Donor
Joined
Jan 16, 2018
Messages
55
Likes
136
Location
Portland OR USA
Hoo boy, I'm glad you beat me to it and I don't have to pioneer it on this forum anyway. Very much looking forward to your observations!
 
OP
D

Dialectic

Major Contributor
Forum Donor
Joined
Sep 26, 2017
Messages
1,738
Likes
3,090
Location
a fortified compound
Hoo boy, I'm glad you beat me to it and I don't have to pioneer it on this forum anyway. Very much looking forward to your observations!

I heard a rumor in 2015 that the BACCH-SP (the $54,000 hardware unit) would not calibrate with insanely expensive digital active speakers made by a certain Swiss manufacturer of luxury audio products. Consequently, I was concerned that BACCH4Mac would fail to calibrate with my digital active Dutch & Dutch 8Cs, requiring me to return the BACCH4Mac kit and pay a substantial restocking fee to Theoretica Applied Physics. (The reported reason for the failure to calibrate with the expensive Swiss-made speakers is that they reportedly resample all signals to 48 kHz.)

I was thus nervous when configuring BACCH4Mac for the first time. My concern was unfounded; it works fine with the 8Cs.

This evening I will add an initial report by editing the original post above.
 
Last edited:

Purité Audio

Master Contributor
Industry Insider
Barrowmaster
Forum Donor
Joined
Feb 29, 2016
Messages
9,050
Likes
12,148
Location
London
Hoo boy, I'm glad you beat me to it and I don't have to pioneer it on this forum anyway. Very much looking forward to your observations!
Jon is performing a public service.
Keith
 

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
I heard a rumor in 2015 that the BACCH-SP (the $54,000 hardware unit) would not calibrate with insanely expensive digital active speakers made by a certain Swiss manufacturer of luxury audio products. Consequently, I was concerned that BACCH4Mac would fail to calibrate with my digital active Dutch & Dutch 8Cs, requiring me to return the BACCH4Mac kit and pay a substantial restocking fee to Theoretica Applied Physics. (The reported reason for the failure to calibrate with the expensive Swiss-made speakers is that they reportedly resample all signals to 48 kHz.)

I was thus nervous when configuring BACCH4Mac for the first time. My concern was unfounded; it works fine with the 8Cs.

My spouse was also eager to hear it because setting it up required a lot of wires and audio and computer devices to be connected in our living room. I think she is pleased.

This evening I will add an initial report by editing the original post above.

Sorry to ask an off-topic question, but maybe you could cover it in your next comment?

There is DSP in Dutch, right? Which means oversampling incl. increased bit depth (32 or 64 bit)? So how come the Dutch are not affected by the external Bach box, while the other CHF system is? I just got curious. Never mind if you find it too way off-topic :)
 

Scott Borduin

Member
Forum Donor
Joined
Jan 16, 2018
Messages
55
Likes
136
Location
Portland OR USA
I heard a rumor in 2015 that the BACCH-SP (the $54,000 hardware unit) would not calibrate with insanely expensive digital active speakers made by a certain Swiss manufacturer of luxury audio products. Consequently, I was concerned that BACCH4Mac would fail to calibrate with my digital active Dutch & Dutch 8Cs, requiring me to return the BACCH4Mac kit and pay a substantial restocking fee to Theoretica Applied Physics. (The reported reason for the failure to calibrate with the expensive Swiss-made speakers is that they reportedly resample all signals to 48 kHz.)

I was thus nervous when configuring BACCH4Mac for the first time. My concern was unfounded; it works fine with the 8Cs.

My spouse was also eager to hear it because setting it up required a lot of wires and audio and computer devices to be connected in our living room. I think she is pleased.

This evening I will add an initial report by editing the original post above.

It would seem that speakers like D&D or Kii would work very well for BACCH, because of the flat phase response and controlled directivity. BACCH apparently implements crosstalk cancellation using phase manipulation (unlike other crosstalk cancellation systems which manipulate frequency response). And of course reflections need to be minimized because both ears will hear them, thereby partly undoing the cancellation effect.
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
3,075
Likes
2,180
Location
UK
BACCH apparently implements crosstalk cancellation using phase manipulation (unlike other crosstalk cancellation systems which manipulate frequency response).
Is that quite it? Generally it works like this: for every impulse you send out from the L speaker, you need to send out a slightly quieter anti-phase cancellation impulse from the R speaker to cancel it as it arrives at the right ear. Then you need to send out a slightly quieter anti-phase impulse from the L speaker to cancel out the previous cancellation pulse as it reaches the L ear. And then you need to send out a slightly quieter impulse from the R speaker to cancel that cancellation impulse as it reaches the R ear. Etc.!

It's a recursive algorithm that possibly sounds a bit unusual if you are not in the sweet spot. I think the refinement of the algorithms is in making them only work on the frequencies the ear is most sensitive to, etc. so as to maximise the effect and minimise the unwanted side effects.
 

Scott Borduin

Member
Forum Donor
Joined
Jan 16, 2018
Messages
55
Likes
136
Location
Portland OR USA
Is that quite it? Generally it works like this: for every impulse you send out from the L speaker, you need to send out a slightly quieter anti-phase cancellation impulse from the R speaker to cancel it as it arrives at the right ear. Then you need to send out a slightly quieter anti-phase impulse from the L speaker to cancel out the previous cancellation pulse as it reaches the L ear. And then you need to send out a slightly quieter impulse from the R speaker to cancel that cancellation impulse as it reaches the R ear. Etc.!

It's a recursive algorithm that possibly sounds a bit unusual if you are not in the sweet spot. I think the refinement of the algorithms is in making them only work on the frequencies the ear is most sensitive to, etc. so as to maximise the effect and minimise the unwanted side effects.

Edgar Choueiri, the professor behind BACCH, lays out the various approaches in chapter 5 of this book: https://tinyurl.com/yde4zbyq . The discussion has a lot of dense math, but the basic outlines are comprehensible. The approach in systems prior to BACCH relied on manipulating the signals heavily in the frequency domain, which colors the sound outside of the sweet spot and creates dynamic range problems due to the large FR boosts required. BACCH manipulates things purely in the time domain (phase is probably not the right word), which leaves FR untouched. The impulse responses are indeed altered through convolution, but not nearly in such a straightforward manner as you describe. BACCH claims the sound outside the sweet spot is like conventional stereo - I'm sure our thread host will enlighten us on that claim :)

Edit: one thing that comes out from the discussion is that the issues with previous approaches are somewhat minimized by narrowing the angle between the speakers. This explains those ambiophonics demo pictures with two huge SoundLab ESLs parked right next to each other :)
 

dallasjustice

Major Contributor
Joined
Feb 28, 2016
Messages
1,270
Likes
907
Location
Dallas, Texas
On the theoretica website it says that BACCH can work with Roon as well as many other players. Where do the BACCH filters reside in the digital playback chain? In terms of playback, does BACCH work inside of Roon?

I know Roon DSP permits customized order of operation for DSP. Would it be possible to apply BACCH filters in Roon DSP first and then apply convolution filters after that? IOW, is it possible to use digital crossovers and EQ after the BACCH filters are applied?

I also want to thank @soundArgument for “hitting the barbed wire first” as we say here in Texas. :)
 

Scott Borduin

Member
Forum Donor
Joined
Jan 16, 2018
Messages
55
Likes
136
Location
Portland OR USA
On the theoretica website it says that BACCH can work with Roon as well as many other players. Where do the BACCH filters reside in the digital playback chain? In terms of playback, does BACCH work inside of Roon?

I know Roon DSP permits customized order of operation for DSP. Would it be possible to apply BACCH filters in Roon DSP first and then apply convolution filters after that? IOW, is it possible to use digital crossovers and EQ after the BACCH filters are applied?

I also want to thank @soundArgument for “hitting the barbed wire first” as we say here in Texas. :)

@dallasjustice, BACCHs hardware recommendations are Quad core I7, 16GB, SSD. It apparently does pretty intensive convolution itself, and there might not be enough CPU left over to do a separate convolution process. You can email [email protected] and ask your questions, they got back to me with detailed responses to a list of questions within the day.
 
OP
D

Dialectic

Major Contributor
Forum Donor
Joined
Sep 26, 2017
Messages
1,738
Likes
3,090
Location
a fortified compound
On the theoretica website it says that BACCH can work with Roon as well as many other players. Where do the BACCH filters reside in the digital playback chain? In terms of playback, does BACCH work inside of Roon?

I know Roon DSP permits customized order of operation for DSP. Would it be possible to apply BACCH filters in Roon DSP first and then apply convolution filters after that? IOW, is it possible to use digital crossovers and EQ after the BACCH filters are applied?

I also want to thank @soundArgument for “hitting the barbed wire first” as we say here in Texas. :)

My pleasure!

BACCH4Mac is a standalone app. Two channels of audio from playback software is routed to the BACCH app through RME's software mixing interface, which is activated whenever the Babyface Pro included with the BACCH4Mac package is connected. (This mixing interface came fully configured on the Mac Mini I bought from Theoretica. If you were to use a Mac you already own, they provide configuration files for the mixing interface, eliminating any need to mess with it.) The BACCH filters thus cannot be applied in Roon itself. This may pose a challenge to those using convolution in Roon, but I don't think the challenge is insuperable.

In a setup like mine, in which there is a USB audio interface separate from the RME Babyface, music playback software sends its output to two of the 16 inputs on a synthetic BACCH4Mac "aggregate device", which is a MacOS fiction consisting of both the Babyface and the separate USB output interface. Receiving two channels of audio as input from playback software through inputs 5 and 6 of the aggregate device, BACCH4Mac processes the audio and routes it to outputs 15 and 16 of the aggregate device, which correspond to the left and right channels on my USB interface.

It may be possible to loop the output from BACCH4Mac back through a software convolution filter and then send the output from that filter to the speakers. I note generally that the BACCH4Mac Pro version has a huge number of features that I have not tested; I do not know for certain that it is impossible to add filters to the BACCH4Mac software itself. Using such filters within BACCH4Mac would obviate the need to do loopback to a separate convolution app within MacOS.

Though using BACCH4Mac with Roon DSP may be tricky to configure, BACCH4Mac is fully compatible with Roon for playback, and Professor Choueiri told me that Roon is the playback software of choice of the vast majority of BACCH4Mac users. However, because I have a huge classical library for which Roon has trouble with metadata, I cannot use Roon effectively. I thus explored the other options on Mac and tested them with BACCH4Mac:
  • iTunes works fine and comes configured on the BACCH4Mac Mac Mini along with a really nice selection of binaural recordings. iTunes does not play FLACs, SACD ISOs, or DSD files, so I cannot use it as my main player. It also, of course, lacks the technical options that more sophisticated players have, and one cannot see what it's sending to the BACCH software.
  • JRiver Media Center for Mac seems very buggy and slow. It crashed when I tried to play SACD ISOs. I managed to route its output to BACCH4Mac when playing PCM files.
  • In my setup, Audirvana+ is not capable of sending its audio output to BACCH4Mac because of a bug within Audirvana+.
  • The free and minimalist Pine Player can send its output to BACCH4Mac but has other shortcomings--principally, that it is extremely slow to load audio files and, contrary to its menu screens, is not capable of gapless playback. It also misidentifies high-res PCM files as MQA files. Until it is upgraded and fixed, it will be unacceptable, but it's promising.
  • Vox can send audio to BACCH4Mac and has the same classical-incompatible library feature of Roon. It seems to lack the sophisticated audio features of Roon.
  • Foobar 2000 for Mac Beta is too minimal at this point to be useful for our purposes.
  • Foobar 2000 for Windows works perfectly with BACCH4Mac under Wine! Configuring audio to go to BACCH4Mac was no problem; configuring certain other elements of Wine was painful. Foobar for Windows has a nice library feature that can maintain the folder structure of one's hard disk. This is how I (and many others with similar libraries) organize classical recordings, so Windows Foobar is my choice for now.
  • The Tidal app for Mac also works perfectly with BACCH4Mac. I'm using this app for streaming along with Foobar for file playback.

@dallasjustice, BACCHs hardware recommendations are Quad core I7, 16GB, SSD. It apparently does pretty intensive convolution itself, and there might not be enough CPU left over to do a separate convolution process. You can email [email protected] and ask your questions, they got back to me with detailed responses to a list of questions within the day.

A moderate amount of CPU horsepower and RAM are required for BACCH4Mac. I bought the upgraded late 2012 Mac Mini that Theoretica offers because, at $1,500, it is more powerful than a brand new $1,400 Mac Mini from the Apple store and comes with BACCH4Mac fully installed and mostly configured. Yes, the core i7 late 2012 Mac Mini is faster than the Mac Minis sold today!

I imagine a new Mac Pro would be better, but it looks like a trashcan, does not fit nicely in an audio rack, and costs at least $3,000. No thank you.

I haven't measured CPU usage while using BACCH4Mac with head tracking, but I have had no hiccups. The Mac Mini gets warm to the touch but stays quiet.
 

dallasjustice

Major Contributor
Joined
Feb 28, 2016
Messages
1,270
Likes
907
Location
Dallas, Texas
BACCH4Mac is a standalone app.
Thanks. I was afraid of that. It’s too bad BACCH wants to remain standalone in all of its iterations.

If Roon had an ASIO line in, then it would be possible to loopback with a Hilo. Of course, ASIO does not exist on a Mac.
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
3,075
Likes
2,180
Location
UK
Edgar Choueiri, the professor behind BACCH, lays out the various approaches in chapter 5 of this book: https://tinyurl.com/yde4zbyq . The discussion has a lot of dense math, but the basic outlines are comprehensible. The approach in systems prior to BACCH relied on manipulating the signals heavily in the frequency domain, which colors the sound outside of the sweet spot and creates dynamic range problems due to the large FR boosts required. BACCH manipulates things purely in the time domain (phase is probably not the right word), which leaves FR untouched. The impulse responses are indeed altered through convolution, but not nearly in such a straightforward manner as you describe. BACCH claims the sound outside the sweet spot is like conventional stereo - I'm sure our thread host will enlighten us on that claim :)

Edit: one thing that comes out from the discussion is that the issues with previous approaches are somewhat minimized by narrowing the angle between the speakers. This explains those ambiophonics demo pictures with two huge SoundLab ESLs parked right next to each other :)
Looking at the various papers, I think I summed it up pretty well! As I described, it is a time domain 'ping-pong' that, if you were to apply it flat across all frequencies would build up large peaks in the frequency response that would only sound anything like normal in the sweet spot in a dead room (and stress the amplifiers and speakers). The refinement of the algorithm is in 'blunting' it (or as they call it 'regularization') so that it maintains the perceived stereo widening effect as much as possible while killing off the worst of the side effects. And they, too, favour a narrow angle for the speakers.
It is relevant to mention in the context of loudspeaker span that keeping Θ small offers advantages that have been recognized since Kirkeby and co-workers presented their analysis[20] of the “stereo dipole” configuration, which has a span of only 10◦ . Objective and subjective evaluations of the effects of loudspeaker span in XTC systems have indicated that such a low-Θ configuration gives a larger sweet spot than that obtained with larger loudspeaker spans[18]....
....Another argument in favor of small loudspeaker spans is particular to the use of analytical filters based on a free-field model, such as those discussed in this paper. Since the free-field model ignores the presence of the listener’s head, it should be expected that filters based on it perform better when the effects of head shadowing are minimized. This can be achieved by decreasing the span angle as can be seen, for instance, in Fig. 3.13
Etc.
 
OP
D

Dialectic

Major Contributor
Forum Donor
Joined
Sep 26, 2017
Messages
1,738
Likes
3,090
Location
a fortified compound
Thanks. I was afraid of that. It’s too bad BACCH wants to remain standalone in all of its iterations.

If Roon had an ASIO line in, then it would be possible to loopback with a Hilo. Of course, ASIO does not exist on a Mac.

As a fan of open systems, I sympathize with your sentiment about the standalone nature of BACCH. I wish it ran on Windows or, better yet, Linux.

I think it might be possible to do a software loopback using the RME software mixing interface. Essentially, the outputs of BACCH4Mac would be configured to send its audio outputs to two loopback channels, which would go into a separate convolution program. That program would send audio to all the channels of your DACs. I'm not sure if Roon will take an input from an external app, but if it did, the convolution and subsequent output to DACs potentially could be done in Roon.

I looked at the possibility of running convolution via a VST plugin within the RME mixing interface, which is called TotalMixFX. RME adamantly refuses to add VST functionality to the app, so that is a no-go.

If you consult the folks at Theoretica, they'll be able to give you a more definitive answer. They're very courteous and responsive.

CORRECTION 2/27: BACCH4Mac includes VST functionality, so convolution can be applied after the BACCH filters in the playback chain.
 
Last edited:

dallasjustice

Major Contributor
Joined
Feb 28, 2016
Messages
1,270
Likes
907
Location
Dallas, Texas
As a fan of open systems, I sympathize with your sentiment about the standalone nature of BACCH. I wish it ran on Windows or, better yet, Linux.

I think it might be possible to do a software loopback using the RME software mixing interface. Essentially, the outputs of BACCH4Mac would be configured to send its audio outputs to two loopback channels, which would go into a separate convolution program. That program would send audio to all the channels of your DACs. I'm not sure if Roon will take an input from an external app, but if it did, the convolution and subsequent output to DACs potentially could be done in Roon.

I looked at the possibility of running convolution via a VST plugin within the RME mixing interface, which is called TotalMixFX. RME adamantly refuses to add VST functionality to the app, so that is a no-go.

If you consult the folks at Theoretica, they'll be able to give you a more definitive answer. They're very courteous and responsive.
Thanks. I think I’ll steer clear for now. I’m not looking to complicate things anymore than they are.

For me to be interested, BACCH would need better integration with Roon. There’s no reason why that wouldn’t be possible. I also don’t see why the RME is needed other than taking the sweeps. It seems like the BACCH app should be to do routing without using totalmix since the output will be simple 2CH.
 
OP
D

Dialectic

Major Contributor
Forum Donor
Joined
Sep 26, 2017
Messages
1,738
Likes
3,090
Location
a fortified compound
Thanks. I think I’ll steer clear for now. I’m not looking to complicate things anymore than they are.

For me to be interested, BACCH would need better integration with Roon. There’s no reason why that wouldn’t be possible. I also don’t see why the RME is needed other than taking the sweeps. It seems like the BACCH app should be to do routing without using totalmix since the output will be simple 2CH.

Yes, the use of different apps in the current BACCH4Mac implementation poses certain challenges. Having seen what you've accomplished using Audiolense and the Roon convolver, I don't blame you.
 

Scott Borduin

Member
Forum Donor
Joined
Jan 16, 2018
Messages
55
Likes
136
Location
Portland OR USA
Looking at the various papers, I think I summed it up pretty well! As I described, it is a time domain 'ping-pong' that, if you were to apply it flat across all frequencies would build up large peaks in the frequency response that would only sound anything like normal in the sweet spot in a dead room (and stress the amplifiers and speakers). The refinement of the algorithm is in 'blunting' it (or as they call it 'regularization') so that it maintains the perceived stereo widening effect as much as possible while killing off the worst of the side effects. And they, too, favour a narrow angle for the speakers.

Etc.

On a brief scan, I think that might be the exact paper reproduced in the book I mention above. Read it carefully - the approach used in BACCH does not use regularization hacks, or require the narrow angle for speakers.

Edit: the paper is not exactly a model of clarity - it bounces back and forth between describing the older idealized approach and the BACCH approach. I was confused myself a couple of times and had to reread to figure out the context of a particular exposition.
 
Last edited:

Gurpreet Jauhal

New Member
Joined
Jun 28, 2017
Messages
1
Likes
3
I use the BACCH system with horn speakers and a tube amp in a very small square room and I can hear sounds behind me and to the left and right. It is an incredible feeling. The best was when I was listening to some Australian aboriginal music and I could hear sounds in a complete circle around me!
I have had no trouble with using Roon or Audirvana.
The customer service is incredible. I am in London, UK and they have called me in the middle of the night (New York time) to solve problems with my system.
 

Purité Audio

Master Contributor
Industry Insider
Barrowmaster
Forum Donor
Joined
Feb 29, 2016
Messages
9,050
Likes
12,148
Location
London
Gurpreet I didn’t realise anyone was using Bacch in London, do you thinkI might be able to hear it in action ?
BW Keith
 
Top Bottom