Jump to content

Recommended Posts

  • Volunteer
Posted

For a while now I've been intrigued by the notion of crosstalk cancellation for speakers. The idea being that when listening to live music, spatial cues are given by volume and time differences between the sound arriving at your left and you right ears. With conventional stereo, this information is muddied because you hear both speakers in both ears - there is 'crosstaIk' between them, so the idea is to cancel this crosstalk and present a more natural set of spatial cues.

 

I was aware that a company called Theoretica had developed a system called BACCH which had some very favourable reviews but was very very expensive (In fact searching for it on SNA revealed that I had actually started a topic about it three years ago: here)

 

The company was set up by Prof Edgar Choueiri who as well as being an extremely helpful and lovely guy, is also a plasma physicist (yes you could call him a rocket scientist). Whilst I don't pretend to understand how it works, I think what sets the BACCH crosstalk cancellation apart from other forms is that it is achieved through phase adjustments rather than EQ.

 

So, fast forward to a few weeks ago and I was investigating it again and it turns out that there is a much more affordable software version called BACCH4Mac which, as the name suggests, it is only available for Macs :)https://www.theoretica.us/bacch4mac/

 

 

I pestered the folk at Theoretica with weeks of questions. How would it work in my system, how would I connect it, what are the differences between the different levels of software etc. They responded promptly and helpfully every time and never treated me like the tyre kicker I may have been, I must say their pre sales service was brilliant.

 

So I decided to bite the bullet and take a trial of the Intro version.

 

[Quick diversion here. There are a few different levels of the software. The cheapest and most basic level approximates the crosstalk cancellation via tape measurements that you input manually (or you can do it by ear). Essentially it makes some assumptions about your head shape and your room and uses the angle between your head and the speakers to determine the filters. So there is quite a small sweet spot - not quite head in a vice but smaller than most 'conventional' speakers

 

The next level up is the "audiophile version". This actually measures your room and your head transfer function via a set of binaural in-ear microphones and determines the filters that way. It can also implement a head tracking camera which means that the sweet spot follows you if you move your head. This seems like a very desirable feature but unfortunately the jump in price is very large - I guess the tech involved is significantly more "techy"

 

There are higher levels that implement filters for headphones and also allow you to do your own mixing - neither of which are of interest to me)

 

End diversion]

 

Prof Choueiri himself hopped on a Skype session with me and remoted in to my Mac. He did all the setup - as I said, the service from these guys is really something else.

 

He then got me to play a test track. 

Artist: Wycliffe Gordon

Album: Dreams of New Orleans 

Song: Down by the Riverside

 

This is a Chesky binaural recording at which Prof Choueiri was present so he knows exactly the location of all the instruments. At about 25 seconds or so the trumpet comes in and I nearly fell off my chair. It was waaay out to the left probably at 9 o'clock. Now binaural recordings should give a good sense of space but this was insane.

Switching the filters off (you can do it on the fly with zero delay), still had a great instrument separation but the trumpet felt like it has moved forward and was now 'stuck' to the left speaker.

The rest of the soundstage also collapsed. I hadn't realised because the  trumpet position had stunned me so much but there was also a tremendous sense of depth to the rest of the band.

If you listen to that recording without BACCH you'll be impressed at what a binaural recording can do, but once you turn BACCH on you realise how much better it can be.

 

Spending the next couple of weeks of the trail period going through my music I would summarise BACCH as follows:

On some recordings, the improvement is simply astounding 

On most recordings the improvement is subtle but it is significantly larger than any improvement I've heard from swapping DACS or cables (which sometimes cost many times more).

There is never any suggestion that this is an 'effect' or artificial in any way. In fact in every case it makes the recording sound more natural.

 

I'm told that BACCH is more effective with speakers that have good directivity and/or in a well damped room - you want to minimise reflections. I've heard it with panels and with KIIs and the impact is brilliant. 

Anyone with panels, especially Sanders panels, needs to listen to this

 

Needless to say, I have bought the intro version. It is not going anywhere :)

 

The next question is, what can I flog so that I can upgrade to the audiophile version!

 

 

 

 

 

 

 

  • Like 7
  • Thanks 1

Posted

Thanks for sharing. So what was the final cost of the software/hardware? Presumably every source needs to be routed through the software?

 

A few years ago I tried  a different solution for cross talk cancellation via convolution filters which were plugged into Roon or HQPlayer from HomeAudioFidelity. Had similar reactions to you when I first tried them. They were also better filters for correcting room issues than I could generate myself with Acourate or REW. A no brainer price wise if using computer audio as a source.

https://www.homeaudiofidelity.com/

  • Like 1
  • Volunteer
Posted
36 minutes ago, AudioGeek said:

Thanks for sharing. So what was the final cost of the software/hardware? Presumably every source needs to be routed through the software?

 

A few years ago I tried  a different solution for cross talk cancellation via convolution filters which were plugged into Roon or HQPlayer from HomeAudioFidelity. Had similar reactions to you when I first tried them. They were also better filters for correcting room issues than I could generate myself with Acourate or REW. A no brainer price wise if using computer audio as a source.

https://www.homeaudiofidelity.com/

 

The pricing is as per the BACCh4Mac page https://www.theoretica.us/bacch4mac/ and depends on which level you go for (I bought the intro version).

 

If you just have digital files or streaming there is no need for any additional hardware (assuming you have already have a Mac that can run the software). The software runs on your Mac and you create an aggregate device in Audio Midi which outputs to your DAC

If you want analog sources then you need an A to D converter

 

 

 

Posted

Ok, thats pretty expensive. Surely accoustic measurements are going to be significantly better than tape measurements . Still need room correction, and to account for that with cross talk cancellation. Then there is room shaper as well.

 

What do you do for bass modes/nulls?

  • Volunteer
Posted
9 minutes ago, AudioGeek said:

Ok, thats pretty expensive. Surely accoustic measurements are going to be significantly better than tape measurements . Still need room correction, and to account for that with cross talk cancellation. Then there is room shaper as well.

 

What do you do for bass modes/nulls?


The tape measurements have nothing to do with room correction. They are purely to determine the angle between your head and the speakers for crosstalk cancellation. 

Posted

Yeah, I realise that. Still think that accoustic measurement for cross talk cancellation is going to be significantly better than tape. And room/bass correction is going to have much more impact than cross talk reduction. In fact  just correcting timing and freq response will create better "imaging".

  • Volunteer
Posted (edited)

Of course acoustic measurement is better, no one disputes that (although the results I’m getting with tape measurements are such a massive improvement already it’s hard to imagine acoustic measurements would provide more than just an incremental improvement. In fact BACCH  themselves reckon you get a very high proportion of the improvement without needing the in-ear measurements ). 
 

Regarding room/bass correction you are conflating things that might be usually related but I don’t think BACCH works by correcting FR. 


I’ve tried Dirac and Acourate and I currently correct bass frequencies using PEQ. None of their improvements to imaging come anywhere near what BACCH does.

 

You shouldn’t think of BACCH as room correction - that’s not what it’s about. 
 
 

 

Edited by sir sanders zingmore
Posted

I’m running the same version of BACCH as @sir sanders zingmore, but also FIR filters produced by Acourate (running in Roon).

 

Sounds great - dynamic and with controlled bass, with the same benefits that is mentioned above. 
 

I plan to run FIR filters in BACCH once Theoretica add support to BACCH-dsp to access VST plugins in the output stage.

 

I do have to say that BACCH is a better improvement that swapping cables, DACs or other components (other than loudspeakers themselves).

 

BACCH seems to run well on the Apple M1 Mac mini.

 

I would like to compare with HAF filters - something I hope to try over the holidays!

  • Like 2
Posted

Have been using cross talk reduction filters (plus room correction) for over 3 years now. Hard to listen without them, whatever strategy you use.

 

There are some nice video/headphone demos on the HAF site on how it works and sample tracks.

  • Like 1
  • Volunteer
Posted
1 hour ago, AudioGeek said:

Have been using cross talk reduction filters (plus room correction) for over 3 years now. Hard to listen without them, whatever strategy you use.

 

There are some nice video/headphone demos on the HAF site on how it works and sample tracks.

Just looking at the site now. Did you do the in-ear mic acoustic measurements for the HRTF version?

  • Like 1

Posted

Indeed I did, they provided filters from both in-ear and external mic measurements I took to compare. The funny thing is the in-ear was better at x-talk reduction, but subjectively I thought the external mic measurements smoothed out bass issues better. They offered to do some more measurements and work on it more but I never got around to it.

 

You guys really need to try Room Shaper as well.

https://audiophilestyle.com/ca/reviews/home-audio-fidelity-room-shaper-review-r855/

 

I thought I was pretty good with using Acourate, but HAF produced much better filters based on REW measurements. I think its their different strategy of also targeting reverberant field which they also describe on the site.

 

 

  • Like 2
  • Volunteer
Posted (edited)
13 minutes ago, AudioGeek said:

Indeed I did, they provided filters from both in-ear and external mic measurements I took to compare. The funny thing is the in-ear was better at x-talk reduction, but subjectively I thought the external mic measurements smoothed out bass issues better. They offered to do some more measurements and work on it more but I never got around to it.

 

 

 

 

One of the things that interest me about the audiophile version of BACCH is that they have head tracking (via a webcam) for their crosstalk cancellation. The only complaint I have about my setup (independent of BACCH) is that the sweet spot is quite small because my speakers are highly directive. Having head tracking would eliminate this (but at a rather big jump in price :( )

Out of interest, which in-ear mics did you use? It's not all that clear which ones are recommended by HAF.

Also, what speakers do you have? BACCH works best with more directive speakers like horns or panels

 

The other thing that makes me think that the approach to crosstalk cancellation is different is that BACCH recommends having speakers wider apart whereas HAF seems to take the more traditional crosstalk approach of having quite a narrow angle.

 

 

Edited by sir sanders zingmore
Posted

I have the Sound Professionals in ear microphones. Not sure if that is what is recommended, but had them for another application.

 

I used the HAF x-talk filters with my PSE-144 horn based system. I have those speakers basically in the corners, so as wide as they will fit in my room.

Reminds me, I should get some filters for my Lx521 system as well. They are over a meter away from all boundaries as a requirement.

Posted

BACCH v9.6 added support for running plugins as the last stage before output to the chosen device.


So I trialled and purchased the SIR3 plug-in although I think Reverberate would also have been a good choice. 
 

I’m using this setup to run convolution for room correction produced by Acourate. 


When I compare how music sounds between (a) running convolution in Roon (ie upstream) vs (b) in BACCH (downstream) I find that I prefer the latter as music sounds tighter, more coherent and generally more open and natural. 
 

Purely subjective, I could not even try to work out why. YMMV, go figure. 
 

Aris

 

  • Like 1
  • 2 months later...
Posted

So, I’ve come here from that other thread, because I’m intrigued. However, looking at the Theoretica website, the price of admission is not trivial. What is the cheapest way of trialling this technology? Is external hardware required in every scenario?

 

  • Like 1

  • Volunteer
Posted
Just now, Steffen said:

So, I’ve come here from that other thread, because I’m intrigued. However, looking at the Theoretica website, the price of admission is not trivial. What is the cheapest way of trialling this technology? Is external hardware required in every scenario?

 

 

It depends what you mean by not trivial. The "intro' version which I have, is relatively cheap. Certainly in my opinion, the improvement is easily worth the entry price and is much greater than improvements I've got from say, DACs that cost much more….

 

The intro version can be trialled for two weeks I think, with a full refund if you don't like it.

As far as hardware goes, the minimum requirement is that you need a Mac to run it. This was no issue for me as I already used a mac in my system for my Roon server. If you have analog sources then it can get messy as you need to digitise.

 

There is also the home audio fidelity method which @AudioGeek mentioned above

https://audiophilestyle.com/ca/reviews/home-audio-fidelity-room-shaper-review-r855/

 

This is quite a lot cheaper and runs as a simple convolution file that you import into Roon (or any other player that can run convolution). In many ways it's much simpler than Bacch4Mac and apparently produces some very good results.

So it requires no external hardware (apart from a measurement mic I guess)

Posted
28 minutes ago, sir sanders zingmore said:

 

 

The intro version can be trialled for two weeks I think, with a full refund if you don't like it.

As far as hardware goes, the minimum requirement is that you need a Mac to run it. This was no issue for me as I already used a mac in my system for my Roon server. If you have analog sources then it can get messy as you need to digitise.

 

 

That’s good to hear. I’ve got a Mac running Roon server, too. I might give the trial version a go.

  • Like 2
  • 1 year later...
Posted
1 minute ago, TheBlackDisc said:

Invite for Hi-Fi Show in October?

 

It's being discussed ;)

  • Like 1
  • Thanks 1

  • 3 months later...
Posted

Sadly, no. And kicking myself for not realising it was there!

  • Like 1
  • 1 month later...
  • 8 months later...
Posted

I've got the software only version but don't know between what two component it fits. I have a HiFi Rose RS 130 streamer, a D2D Gustard converter from USB to I2S and a Gustad r26 DAC feeding an Anthem AV70 8K and then 11 Golden Ear speakers fed by 121 class D amps.

So how is the Mac Mini running Bacch4Mac software physically connected?

  • Recently Browsing   0 members

    • No registered users viewing this page.




×
×
  • Create New...
To Top