BACCH and other XTC

[Gwreck]

Trying to understand what bacch does… i asked chatgpt..



Is the above broadly how it works?

I think at the most basic level all electronic XTC methods do what you are saying. Just because it sounds easy doesn’t mean it is(doing it well at least). If you go to the Princeton 3d3a labs website the base equations for BACCH are available to see. There are many more factors though that are part of the process. I don’t even pretend to understand all of the complications. The descriptions of the basic idea of XTC you use tend to also be the descriptions that BACCH naysayers use to discredit it.

 

[STC]

I think at the most basic level all electronic XTC methods do what you are saying. Just because it sounds easy doesn’t mean it is(doing it well at least). If you go to the Princeton 3d3a labs website the base equations for BACCH are available to see. There are many more factors though that are part of the process. I don’t even pretend to understand all of the complications. The descriptions of the basic idea of XTC you use tend to also be the descriptions that BACCH naysayers use to discredit it.

XTC is a rather a simple process and not too complicated. The principle involves sending and inverted sound of the opposite channel to the other hear to attenuate the level at the opposite ear. This process continues until the ears no longer hear the unintended sound ( crosstalk ).

The simple process gets complicated because the sound heard by the intended ear and the delayed crosstalk heard by the other ear are not of the same frequency response as HRTF modifies the response. Other XTC, like BACCH does not make the correction why requires EQ to account for the changes as such getting high attenuation with non EQ XTC is not possible.

The recursive nature of cancellation can be done with effects like ping pong and or even a convolution engine using IR. The problem of using EQ with such cancellation is that EQ introduces phase changes. Alternatively, to response is broken into different frequency range so that the cancellation level can be adjusted for other ear as higher frequencies gets attenuated far more than the lower frequencies due to head shadow.

But all this while looks nice on paper, in practice it is not so simple because our head is constantly moving which makes such high precision cancellations redundant.

OTOH, we are reading about the effectiveness of XTC over another. I have witnessed or rather heard a non RACE XTC somehow projected the binaural recording well behind my head which I don’t even get such even with headphones. Does this qualify to be superior to other XTC? Unfortunately, listening to my fav “Norah Jones” I rather stick to stereo if I had to use such XTC which seemed to doing extraordinary things that XTC not supposed to do as this is stepping into artificial placements. No wonder some XTC emphasis the word that their plugin is intended for binaural recording. In my opinion, any XTC kills the enjoyment of song like Norah Jones or my favorite Hotel California is definitely not worth its place in audiophile world.

Lastly, can XTC cause coloration? I believe most XTC I tried sounded almost like stereo( when correctly setup) and I am always wondering about this claim.

 

[Gwreck]

XTC is a rather a simple process and not too complicated. The principle involves sending and inverted sound of the opposite channel to the other hear to attenuate the level at the opposite ear. This process continues until the ears no longer hear the unintended sound ( crosstalk ).

The simple process gets complicated because the sound heard by the intended ear and the delayed crosstalk heard by the other ear are not of the same frequency response as HRTF modifies the response. Other XTC, like BACCH does not make the correction why requires EQ to account for the changes as such getting high attenuation with non EQ XTC is not possible.

The recursive nature of cancellation can be done with effects like ping pong and or even a convolution engine using IR. The problem of using EQ with such cancellation is that EQ introduces phase changes. Alternatively, to response is broken into different frequency range so that the cancellation level can be adjusted for other ear as higher frequencies gets attenuated far more than the lower frequencies due to head shadow.

But all this while looks nice on paper, in practice it is not so simple because our head is constantly moving which makes such high precision cancellations redundant.

OTOH, we are reading about the effectiveness of XTC over another. I have witnessed or rather heard a non RACE XTC somehow projected the binaural recording well behind my head which I don’t even get such even with headphones. Does this qualify to be superior to other XTC? Unfortunately, listening to my fav “Norah Jones” I rather stick to stereo if I had to use such XTC which seemed to doing extraordinary things that XTC not supposed to do as this is stepping into artificial placements. No wonder some XTC emphasis the word that their plugin is intended for binaural recording. In my opinion, any XTC kills the enjoyment of song like Norah Jones or my favorite Hotel California is definitely not worth its place in audiophile world.

Lastly, can XTC cause coloration? I believe most XTC I tried sounded almost like stereo( when correctly setup) and I am always wondering about this claim.

Do you even BACCH bro?
In all seriousness though what’s your experience with XTC. If you have any experience with it at all you should know that BACCH far better than other forms of XTC. Using a Nora Jones song as an argument against XTC is like using 3 day old pizza as an argument against refrigeration. I see you have read something about the most basic and primitive forms of XTC. Most were effect at the XTC part but (AFAIK) BACCH is the only that doesn’t cause huge coloration to the sound.

Also the head tracking and ORC with BACCH is nothing to sneeze at.

 

[STC]

Do you even BACCH bro?
In all seriousness though what’s your experience with XTC. If you have any experience with it at all you should know that BACCH far better than other forms of XTC. Using a Nora Jones song as an argument against XTC is like using 3 day old pizza as an argument against refrigeration. I see you have read something about the most basic and primitive forms of XTC. Most were effect at the XTC part but (AFAIK) BACCH is the only that doesn’t cause huge coloration to the sound.

Also the head tracking and ORC with BACCH is nothing to sneeze at.

i am not attacking BACCh or other XTC. i am explaining what XTC should do and deliver. more can be enticing but not necessarily correct.

i tried uBACCH but didn’t work. Choueiri went to Ralph’s place to demo BACCH but also didn’t work in Ralph’s main system. He managed to get it work in his other system.

XTC is not invented by Choueiri. if anyone should take the credit for practical implementation then it should be Carver using one shot analogue cancellation. It had several limitations.

Yes head tracking is a plus but do you walk around in your room while listening to the music? And Norah Jones is important to me. Hotel California is important to me. Thousands of songs similar to them are important to me. i almost given up on XTC because Hotel California and other usual favorite tracks sounded awful DUE to my poor implementation in the beginning and too bad the instruction is not clear either with the DSP plugin manual.

Maybe if you take a pink noise impulse response and do the cancellation and capture the digital output before amplifier ( with DAW you do it easily) then compare all the spectrum of different XTC and then do the listening test to see if it is really colored as BACCH claimed.

Howard Keller heard the best of BACCH and Ralph’s main system and he is probably the best person to give an unbiased opinion that’s is not tainted with other motives.

 

[NTK]

XTC is not invented by Choueiri. if anyone should take the credit for practical implementation then it should be Carver using one shot analogue cancellation. It had several limitations.

XTC was first (publicly) described in the patent (US 3,236,949) by Atal and Schroeder (the same Schroeder that came up with the "Schroeder" frequency) of Bell Labs, filed in Nov 1962, granted in Feb 1966. In the patent they also described the recursive XTC technique. The concept of XTC is almost as old as 2 channel stereo itself.

 

[Gwreck]

XTC was first (publicly) described in the patent (US 3,236,949) by Atal and Schroeder (the same Schroeder that came up with the "Schroeder" frequency) of Bell Labs, filed in Nov 1962, granted in Feb 1966. In the patent they also described the recursive XTC technique. The concept of XTC is almost as old as 2 channel stereo itself.

View attachment 382316

Was there someone claiming the he did invent XTC?

 

[Gwreck]

i am not attacking BACCh or other XTC. i am explaining what XTC should do and deliver. more can be enticing but not necessarily correct.

i tried uBACCH but didn’t work. Choueiri went to Ralph’s place to demo BACCH but also didn’t work in Ralph’s main system. He managed to get it work in his other system.

XTC is not invented by Choueiri. if anyone should take the credit for practical implementation then it should be Carver using one shot analogue cancellation. It had several limitations.

Yes head tracking is a plus but do you walk around in your room while listening to the music? And Norah Jones is important to me. Hotel California is important to me. Thousands of songs similar to them are important to me. i almost given up on XTC because Hotel California and other usual favorite tracks sounded awful DUE to my poor implementation in the beginning and too bad the instruction is not clear either with the DSP plugin manual.

Maybe if you take a pink noise impulse response and do the cancellation and capture the digital output before amplifier ( with DAW you do it easily) then compare all the spectrum of different XTC and then do the listening test to see if it is really colored as BACCH claimed.

Howard Keller heard the best of BACCH and Ralph’s main system and he is probably the best person to give an unbiased opinion that’s is not tainted with other motives.

Certainly there are situations where BACCH makes no sense to use. A 3d sound field that would be consistent from many locations would likely require wave field synthesis which is beyond complicated for home use. I have heard many different types of XTC and even different form of the same type had varying effect and coloration. The MiniDSP “Miniambio” has fairly low coloration for RACE but other versions I have heard were terrible much like having an Vox Wah pedal in the signal chain.

 

[STC]

Certainly there are situations where BACCH makes no sense to use. A 3d sound field that would be consistent from many locations would likely require wave field synthesis which is beyond complicated for home use. I have heard many different types of XTC and even different form of the same type had varying effect and coloration. The MiniDSP “Miniambio” has fairly low coloration for RACE but other versions I have heard were terrible much like having an Vox Wah pedal in the signal chain.

As I said before, coloration of XTC is no longer an issue. It was when you use just the attenuation and delays to do the XTC because successive signals changes the original signal where the sound can be rather bright and extremely coloured. Can you hear the colouration in this video? if you analyzed the frequency spectrum it would be very different but can you hear them? Use good headphones or AirPod ( turn off spatial). The clicking sound of the mouse is where I alternate between the conventional stereo playback and XTC ( NOT BACCH) . The playback was capture with DPA binaural microphone and you are hearing the actual sound waves reaching the ears.

 

[LIΟN]

As I said before, coloration of XTC is no longer an issue. It was when you use just the attenuation and delays to do the XTC because successive signals changes the original signal where the sound can be rather bright and extremely coloured. Can you hear the colouration in this video? if you analyzed the frequency spectrum it would be very different but can you hear them? Use good headphones or AirPod ( turn off spatial). The clicking sound of the mouse is where I alternate between the conventional stereo playback and XTC ( NOT BACCH) . The playback was capture with DPA binaural microphone and you are hearing the actual sound waves reaching the ears.

nice. btw I'm curious about your opinion making cancellations with 3 or 5 speaker arrays. positive or negative?

 

[STC]

nice. btw I'm curious about your opinion making cancellations with 3 or 5 speaker arrays. positive or negative?

Frankly, I think the best way to XTC is having 4 identical speakers. The first pair play the stereo unaltered and the cancellation is done by the other speakers. I think Polk SDA crosstalk also having similar concept with additional drivers.

I cannot comment about your approach with 2 or 5 speakers because I am more trying to get the best from the existing stereo setup with two speakers without any additional expenses not withstanding DCH the concept. yes… I am contradicting myself. )

 

[NTK]

Frankly, I think the best way to XTC is having 4 identical speakers. The first pair play the stereo unaltered and the cancellation is done by the other speakers. I think Polk SDA crosstalk also having similar concept with additional drivers.

But that will mean:

• The XTC correction signals (supplied by the correction speakers) to cancel the main sound (supplied by the main speakers) will need to compensate for the different HRTFs of the main speakers due to different angles of arrival (since main speakers and correction speakers aren't co-located).
• The subsequent XTC correction signals to correct the earlier correction signals will need to compensate for another set of HRTFs, since this time the correction signals will be compensating the earlier correction signals from the correction speakers.

 

[STC]

But that will mean:

• The XTC correction signals (supplied by the correction speakers) to cancel the main sound (supplied by the main speakers) will need to compensate for the different HRTFs of the main speakers due to different angles of arrival (since main speakers and correction speakers aren't co-located).

Your cancellation signal is coming from the other side. Just like you cancel with one pair of speakers now you use the second pair. When You do in ear measurements you can send the inverted signal from the other speakers. With less than one degree difference I doubt it would really matter as our hearing does not rely on precise mathematically value that need to be exact.

In one the plugin that I was testing, the delay value was based on exact calculation but attenuation was decided based on actual hearing. The sound was good. In fact, that was the best version. Later, I found out the actual delayed value entered in μs changes to a default value despite in the plugin box it still showed the value I entered.

When we brought it to the developer, he confirmed as follows “
I have reproduced the issue where the UI displays 0.1 for
the SDelay delay time after reloading the document. I haven't determined
why this is, or whether it actually affects the delay time.”

So despite all wrong values the cancellation worked perfectly. So all the concerns by observers without actual experience of doing crosstalk only appears valid on papers but in practice they are not.

• The subsequent XTC correction signals to correct the earlier correction signals will need to compensate for another set of HRTFs, since this time the correction signals will be compensating the earlier correction signals from the correction speakers.

As above.

 

[NTK]

So all the concerns by observers without actual experience of doing crosstalk only appears valid on papers but in practice they are not.

Big talk. We'll see if you can do better than Cooper, Bauck, Møller, Farina, Nelson, Aarts, Choueiri, etc.

 

[LIΟN]

Frankly, I think the best way to XTC is having 4 identical speakers. The first pair play the stereo unaltered and the cancellation is done by the other speakers. I think Polk SDA crosstalk also having similar concept with additional drivers.

I cannot comment about your approach with 2 or 5 speakers because I am more trying to get the best from the existing stereo setup with two speakers without any additional expenses not withstanding DCH the concept. yes… I am contradicting myself

I haven't tested the array, but I think it's going to be pretty tricky to set up the delay as well.
I'll have to try it later... Thanks info

 

[STC]

Big talk. We'll see if you can do better than Cooper, Bauck, Møller, Farina, Nelson, Aarts, Choueiri, etc.

Now I am remember why I used the ignore button. Farina? He himself was skeptical at first.

 

[STC]

I haven't tested the array, but I think it's going to be pretty tricky to set up the delay as well.
I'll have to try it later... Thanks info

The Polk SDA tech.

 

[LIΟN]

The Polk SDA tech.

thanks. i'll take a look for later

 

[STC]

Thanks to @Lion who asked for actual measurements of attenuation of crosstalk cancellation of RACE measurements, here are the graphs. The complete graphs will be attached in a video as they are too many.

The best measurement is 26dB difference around a single point of 1277kHz.

 

[Keith_W]

Thanks to @Lion who asked for actual measurements of attenuation of crosstalk cancellation of RACE measurements, here are the graphs. The complete graphs will be attached in a video as they are too many.

How did you obtain those measurements, STC?

 

[STC]

How did you obtain those measurements, STC?

I used the IR of 100 to 5000Hz and taken the measurements with DPA binaural microphones. Then use Audacity to convert the data and use Excel to generate the graph. These are actual measurements and not simulated. Only certain portion you get above 10dB attenuation. Above 5000Hz it is difficult to measure because of rather short wavelength making good cancellations impossible.