BACCH4MAC Short Review / first impressions

[sweetsounds]

This is a BACCH feature?

No, BACCH is stereo-only and achieves the L to R separation by a sort of noise cancellation for the other ear.
A thick divider wall from your nose to the front wall has the same effect, but listening with a 5 inch thick divider in the middle of the room with the nose against it is for real hard-core audiophiles, except when listening to The Wall album.

 

[Dialectic]

No, BACCH is stereo-only and achieves the L to R separation by a sort of noise cancellation for the other ear.
A thick divider wall from your nose to the front wall has the same effect, but listening with a 5 inch thick divider in the middle of the room with the nose against it is for real hard-core audiophiles, except when listening to The Wall album.

Respectfully, I don't think the effect is the same, although it is conceptually similar.

 

[Tim Link]

This is a BACCH feature?

No, I'm pretty sure it's not. My matrix is vastly simpler than BACCH's advanced processing. I'm just just channel mixing. My setup does require 3 speakers though. BACCH can work with just two I believe. There may be options to add more - I don't know. I get a decent amount of crosstalk elimination this way, and also any sounds that are panned center get summed into the center channel only and canceled in the side channels, so center vocalists and instruments are crosstalk free and very pure sounding.

 

[Tim Link]

Respectfully, I don't think the effect is the same, although it is conceptually similar.

I agree. It's not the same but it is similar in that crosstalk is being minimized. I've used a divider wall, which is absolutely primo sounding in my opinion. MY three speaker setup is a compromise between regular stereo and full crosstalk elimination. It has it's advantages. If I decide to spend the bucks on BACCH at some point I think I'll leave my speakers setup as they are and just turn off the center channel when BACCH is in use. It'd be really interesting to compare. I've used other digital crosstalk elimination software that worked really well in terms of creating a huge, immersive soundfield, but the processing was not transparent enough for me. There was a Google Chrome plug-in that worked well, and some software from Soundpimp that was even better. I still have Soundpimp on my laptop but I forgot how to get it started. It's kind of complicated to install.

 

[tifune]

No, I'm pretty sure it's not. My matrix is vastly simpler than BACCH's advanced processing. I'm just just channel mixing. My setup does require 3 speakers though. BACCH can work with just two I believe. There may be options to add more - I don't know. I get a decent amount of crosstalk elimination this way, and also any sounds that are panned center get summed into the center channel only and canceled in the side channels, so center vocalists and instruments are crosstalk free and very pure sounding.

In that regard, then, is BACCH better than modern upmixers if a proper center channel is available? For simplicity sake, we'll say proper = same model as L/R, mounted at same height/distance as L/R

 

[Dialectic]

In that regard, then, is BACCH better than modern upmixers if a proper center channel is available? For simplicity sake, we'll say proper = same model as L/R, mounted at same height/distance as L/R

Almost certainly yes.

 

[Tim Link]

Almost certainly yes.

I would agree that BACCH is ultimately capable of portraying a soundscape that even a highly optimized multichannel setup would not be able to do. And because it has HRTF and head tracking functionality it can even surpass a really good physical barrier. That being said, there are some advantages to a multi-channel system. You get a lot of headroom. Dividing the sound to more speakers and amps is going to ease the load on each. When multiple people are listening in various places in the room the soundstage stays put and coherent for all of them. You can move around and get a different perspective on the soundstage without it collapsing. It's really easy to live with. I've read the advanced versions of BACCH can track your movements and I think they may be able to optimize for more than one person at a time. I thought I read they could do that, or were working on that. Some people's reviews are mixed. They are impressed but they don't like the sound "tugging at their ears." I thought that was an interesting portrayal of the effect it can have sometimes, and it's one that I can totally relate to. Whenever you are in a listening environment and one ear is hearing something much louder than the other, that comes across as rather strange. Crosstalk elimination can do that, and if it's strong enough it can make a hard panned sound seem to be right up next to your ear. But if the HRTF and the nature of the sound itself doesn't match that scenario well enough your brain can struggle with interpretation and it can almost seems like something is being done to your ear - like it's being tugged or something.

 

[Keith_W]

For headphones: does anybody know the difference between what BACCH does and the Smyth Realizer? Both of them utilize head trackers and claim to be able to reproduce a convincing simulation of a listening room on headphones.

 

[phoenixdogfan]

For headphones: does anybody know the difference between what BACCH does and the Smyth Realizer? Both of them utilize head trackers and claim to be able to reproduce a convincing simulation of a listening room on headphones.

While both may be able to do that in two channel audio, only the Smyth will do that in multichannel in all current formats including Atmos, DTS-X, and Auro 3D. I have owned an A16 for three years and it gives realistic "Out of Your Head" renderings of all those layouts. It is just the ticket if you want an Atmos Home Theater without dedicating a room in your home in The Hamptons to it.

 

[Keith_W]

I had a demonstration of BACCH today at a friend's house. In a nutshell, it was rather underwhelming. I struggled to hear a difference between BACCH and no BACCH on my music, so I asked him to play music where the effect was more pronounced. Using the same demo track that Edgar Choueri uses (some song about "ain't no study no more"), I was able to hear that the trumpet solo had shifted to a more lateral position by maybe 5 degrees and the soundstage was a bit wider. We then experimented with changing BACCH settings and repeating the setup process but nothing we did today was able to replicate the dramatic differences reported elsewhere, including on this thread.

My friend said that he has tried BACCH with 3 different speakers and they seem to work best with highly directional speakers, like the Sanders electrostats that he had sitting in another room (but disconnected). They also worked better with the Avantgarde horns he had previously. He is currently using another horn speaker - I have forgotten the manufacturer. On the Sanders speakers, I was told that the trumpet was moved all the way to the left, so it sounded like it was coming directly to the left of the listening position.

He is using the basic version of BACCH4Mac, which does not include the head tracker. He also has some limitations on how he can place the speakers in his room, which meant that the speakers were pushed against the wall and into the corners of the room, and a sofa was placed in front of the speaker. However, he told me that he ran frequency sweeps with and without the sofa and the sofa did not change the sound by much.

I suppose the take-away message from today's demonstration is that the effect of BACCH is speaker and room dependent and I look forward to hearing a more convincing demonstration.

Yes I know that a free trial for 1-2 weeks is available, but for me it would also involve purchasing a Mac to run the software.

 

[fredoamigo]

I had a demonstration of BACCH today at a friend's house. In a nutshell, it was rather underwhelming. I struggled to hear a difference between BACCH and no BACCH on my music, so I asked him to play music where the effect was more pronounced. Using the same demo track that Edgar Choueri uses (some song about "ain't no study no more"), I was able to hear that the trumpet solo had shifted to a more lateral position by maybe 5 degrees and the soundstage was a bit wider. We then experimented with changing BACCH settings and repeating the setup process but nothing we did today was able to replicate the dramatic differences reported elsewhere, including on this thread.

My friend said that he has tried BACCH with 3 different speakers and they seem to work best with highly directional speakers, like the Sanders electrostats that he had sitting in another room (but disconnected). They also worked better with the Avantgarde horns he had previously. He is currently using another horn speaker - I have forgotten the manufacturer. On the Sanders speakers, I was told that the trumpet was moved all the way to the left, so it sounded like it was coming directly to the left of the listening position.

He is using the basic version of BACCH4Mac, which does not include the head tracker. He also has some limitations on how he can place the speakers in his room, which meant that the speakers were pushed against the wall and into the corners of the room, and a sofa was placed in front of the speaker. However, he told me that he ran frequency sweeps with and without the sofa and the sofa did not change the sound by much.

I suppose the take-away message from today's demonstration is that the effect of BACCH is speaker and room dependent and I look forward to hearing a more convincing demonstration.

Yes I know that a free trial for 1-2 weeks is available, but for me it would also involve purchasing a Mac to run the software.

Very interesting ... although you give some clues on the configuration, could you give more details? on the acoustics of the room and the positioning of the speakers in it or the distance of listening ... in a general way we always learn more from what does not work well to understand the why of the how than from what already works well here's why it is interesting

 

[Keith_W]

I did a bit more googling. The speakers are Tune Audio Anima horn speakers and a single SVS subwoofer. The hardware pathway looks like this:

2012 Intel Mac Mini --> RME Babyface (for A/D and D/A conversion) --> MiniDSP (for DIRAC, perform crossover for mains and subs at 40Hz) --> power amp --> speakers (with internal passive crossover).

The room was narrow and long, and opens out to a dining room and kitchen area. The listening area is part of this long room. The speakers are pushed against the wall and into the corners. Both speakers are toed in to the listener. This is a poor drawing I quickly whipped up on Paint, it is by no means dimensionally correct:

(EDIT) at request of my friend I have deleted the image that was originally posted and removed some of the more detailed descriptions I made of the room. I would like to publicly apologize for any offence caused.

 

[Bjorn]

Just as a side note. When removing cross talk with a barrier board, the speaker are suppose to be close to each other.
Last year I tried placing boards between wide spaced speakers, and it did almost nothing. FIY: The board was placed closer to the ears than what picture shows. Righ up to the nose.

Correctly done with Don Keele in the listening chair:

 

[Bjorn]

The Avantgardes showed a nice improvement too. I've not seen a polar plot of them but just anecdotally my experience is that they have a smallish sweet spot. Again possibly this is better for BACCH's algorithms

Here's the horizontal polar of Avantgard Uno XD.

If BACCH works best without early reflections, you can achieve that with applying absorption (effective to Schroeder) on side walls.

 

[sweetsounds]

Just to add a bit of info, I'm the owner of the above system. I largely agree with Keith's assessment of the impact of BACCH with these speakers.
A few points that might add a bit more context.
Firstly, the speaker positioning (in the corners) is a combination of necessity - they are rather imposing beasts, and recommendation. The manufacturer recommends just this placement.

Thanks both of you for sharing. The room looks like a tough challenge (asymmetry, corner positions, long distance ...).

We have to manage expectations, where DSP has its limits, even with more calculation power and algorithms:
First, DSP can equalize a speaker quite well TONALLY to fix problems AT the SOURCE. These effects on the sound generation can be heard immediately. You can also shape the wave-front by DSP'ing multiple drivers (cardioid bass or line-arrays).

Room correction and crosstalk elimination try to treat problems AFTER the sound generation by employing pre-distortion.
There are limitations to this, you can't realistically get rid of cabinet vibrations, reflections, standing waves and echoes. You can REDUCE room modes and cross-talk in SINGLE locations at SOME moments in time.
Example: In one room the PSI AVAA active bass trap clearly elimated annoying room modes at the listening, in my room it did nothing noticable and introduced more boom in another location. A normal bass trap for less than a quarter of the price worked much better (but is bigger of course).

What the pro-DSP fraction is simply saying: before spending another 10k on new electronics and speakers, first optimize your room/speaker position, choose a system you enjoy and then try DSP for touch-ups (Dirac, Acourate, BACCH etc.).

Different speakers introduce different problems. DSP keeps the sonic character of your system and improves the result.

 

[Gwreck]

This is a BACCH feature?

No. The point (I assume)of BACCH is to avoid having to do all of these complications to get crosstalk cancellation.

 

[Gwreck]

Thanks both of you for sharing. The room looks like a tough challenge (asymmetry, corner positions, long distance ...).

We have to manage expectations, where DSP has its limits, even with more calculation power and algorithms:
First, DSP can equalize a speaker quite well TONALLY to fix problems AT the SOURCE. These effects on the sound generation can be heard immediately. You can also shape the wave-front by DSP'ing multiple drivers (cardioid bass or line-arrays).

Room correction and crosstalk elimination try to treat problems AFTER the sound generation by employing pre-distortion.
There are limitations to this, you can't realistically get rid of cabinet vibrations, reflections, standing waves and echoes. You can REDUCE room modes and cross-talk in SINGLE locations at SOME moments in time.
Example: In one room the PSI AVAA active bass trap clearly elimated annoying room modes at the listening, in my room it did nothing noticable and introduced more boom in another location. A normal bass trap for less than a quarter of the price worked much better (but is bigger of course).

What the pro-DSP fraction is simply saying: before spending another 10k on new electronics and speakers, first optimize your room/speaker position, choose a system you enjoy and then try DSP for touch-ups (Dirac, Acourate, BACCH etc.).

Different speakers introduce different problems. DSP keeps the sonic character of your system and improves the result.

I agree that optimizing your room and speakers is a good idea before relying on DSP touch-ups to the sound. “Room correction” IMHO significantly overstates what these programs actually do. One additional thing, Dr. Choueri will strongly (but kindly assert that BACCH is not room correction if one accidentally refers to it as such.

 

[Gwreck]

I am pretty proud of myself as I’m now getting 23db of crosstalk cancellation for each ear. What is anyone else getting with their set up?

 

[Keith_W]

May I ask what version of BACCH you are using, Gwreck? When I saw my friend's setup, it was only a control panel on his Mac.

 

[sweetsounds]

I am pretty proud of myself as I’m now getting 23db of crosstalk cancellation for each ear. What is anyone else getting with their set up?

This is impressive!

According to stereophonic analysis Link, the location of a sound is determined by a sound window of the first 1.8ms and up to 18dB level difference.

So an attenuation of more than 18dB should be eliminating any crosstalk.

it also suggests, that the timing accuracy should be better than 0.1ms to keep the image location. This means that any delta/reflection of 4cm makes a difference.