• Welcome to ASR. 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!

Multi-Speaker Crosstalk Reduction (XTC) – Beta Testing

Joined
Mar 27, 2018
Messages
13
Likes
14
Location
France
Hi everyone,

I’ve just posted a technical dive into Crosstalk Reduction (XTC) using a multi-speaker array rather than the traditional 2-channel setup.

The goal is to move beyond the limitations of standard XTC by using additional speakers push crosstalk rejection to much higher levels, achieving a level of binaural immersion that 2-speaker systems cannot reach . You can find the full write-up and simulations here:

Blog: Crosstalk Reduction with Multiple Speakers

I’m looking for ASR members with multi-channel setups (3.0, 4.0, 5.1, or similar) to test these filters in their own rooms.

The only technical requirement is that your playback system must be capable of applying convolution filters (e.g., Roon, Audrivana, JRiver, etc..)

Best,
Thierry
[email protected]
 
achieving a level of binaural immersion that 2-speaker systems cannot reach
2.1 phased array, also SBA that can cancel even the low frequencies here.

04.jpg
 
Call me skeptical. The idea of cross-talk cancellation has been around for more than half a century. If using more than a pair of speakers solves a real problem, why hasn't this idea been discovered and in use already. Not like the long list of researchers, from Schroeder to Choueiri, are suffering from a lack of ingenuity.
 
Hi everyone,

I’ve just posted a technical dive into Crosstalk Reduction (XTC) using a multi-speaker array rather than the traditional 2-channel setup.

The goal is to move beyond the limitations of standard XTC by using additional speakers push crosstalk rejection to much higher levels, achieving a level of binaural immersion that 2-speaker systems cannot reach . You can find the full write-up and simulations here:

Blog: Crosstalk Reduction with Multiple Speakers

I’m looking for ASR members with multi-channel setups (3.0, 4.0, 5.1, or similar) to test these filters in their own rooms.

The only technical requirement is that your playback system must be capable of applying convolution filters (e.g., Roon, Audrivana, JRiver, etc..)

Best,
Thierry
[email protected]
I would say that Polk SDA is basically XTC with multiple speakers except they are contained in one unit. With a well optimized BACCH system I have over 20db of XTC. 20db is apparently the limit for what actually makes any difference. I am being lazy by asking this but is there a specific filter to be run with multiple channels? It may be interesting to try.
 
Call me skeptical. The idea of cross-talk cancellation has been around for more than half a century. If using more than a pair of speakers solves a real problem, why hasn't this idea been discovered and in use already. Not like the long list of researchers, from Schroeder to Choueiri, are suffering from a lack of ingenuity
Using multiple speakers for XTC is not a new idea, but making them tonal neutral and better compatible with stereo recordings is a step forward

I am being lazy by asking this but is there a specific filter to be run with multiple channels? It may be interesting to try.
Yes you need specific filters for each loudspeaker. I can generate them based on your geometrical layout
 
@STC @Gwreck are our local XTC enthusiasts. Not sure if they have multichannel speakers though.

Keith, I thought I had elevated myself from a mere enthusiast to a developer. Hehehe.

Anyway, the idea is great. I started 3 pairs implemented based on Ralph’s papers before moving to DCH. Hopefully HAF can come up with a one-box solution viable for all audiophiles because audiophiles still don’t know what they are missing with stereo.

ST
 
Using multiple speakers for XTC is not a new idea, but making them tonal neutral and better compatible with stereo recordings is a step forward


Yes you need specific filters for each loudspeaker. I can generate them based on your geometrical layout

I have and AV receiver Denon a1h ,with a 9.3.4 configuration. Which speakers would be used and how would the filter be applied?
I am not sure if this has ever been tried but it seems to me those in cars with the headrest speakers in some OEM systems would be a perfect location for the XTC signal albeit with significant DSP added.
 
I have and AV receiver Denon a1h ,with a 9.3.4 configuration. Which speakers would be used and how would the filter be applied?
I am not sure if this has ever been tried but it seems to me those in cars with the headrest speakers in some OEM systems would be a perfect location for the XTC signal albeit with significant DSP added.
You need to be able to feed your AVR from a player having convolution capabilities (Roon, JRiver, Audirvana, Foobar, or using Hang Loose Convolver)
Each channel of the stereo files will be filtered and sent to the selected speakers. We can try different layouts but starting point is mains + center
 
It reminds me of the Optimal Source Distribution idea.

 
It reminds me of the Optimal Source Distribution idea.

Yes, there are a good number of publications and patents on XTC but to my knowledge the question of compatibility with stereo tracks is not addressed
 
Oh that sounds interesting. Do I understand correctly that it remains based on stereo tracks?

In that case the plug-in should automatically bypass when multichannel tracks are being played. Like I sometimes mix playlists and it would be annoying to have to manually switch the XTC on and off, depending on the next song's architecture in the playlist.

Is the registration later on easily transferable to another system and cross-platform compatible? Or is one registration required for each device? I'm asking because I noticed that @mitchco is planning to bundle some of your plug-ins in demo mode with his upcoming Hang Loose Processor.
 
Oh that sounds interesting. Do I understand correctly that it remains based on stereo tracks?

In that case the plug-in should automatically bypass when multichannel tracks are being played. Like I sometimes mix playlists and it would be annoying to have to manually switch the XTC on and off, depending on the next song's architecture in the playlist.

Is the registration later on easily transferable to another system and cross-platform compatible? Or is one registration required for each device? I'm asking because I noticed that @mitchco is planning to bundle some of your plug-ins in demo mode with his upcoming Hang Loose Processor.
Yes, the objective is to play stereo tracks taking benefit of a multichannel setup to implement more advanced XTC. The testing phase will be with convolution filters and if there is positive feedback and interest, it might turn into an audio plugin (which could include the bypass feature you mention)
 
You need to be able to feed your AVR from a player having convolution capabilities (Roon, JRiver, Audirvana, Foobar, or using Hang Loose Convolver)
Each channel of the stereo files will be filtered and sent to the selected speakers. We can try different layouts but starting point is mains + center
I have roon that can run into my Denon a1h. What information do you need? I have been very busy at work so it may take me a bit to gather.
 
Are the additional speakers purely "support" channels?
Or are you in a multichannel stereo mode?

Do the speakers need to be identical?
 
I am buying additional surround speakers soon so I will have a 4.4 setup with 4 subs and okto dac 8 pro. At the moment I have a stereo plus 6 subs. All independent channels run with Jriver and vst plugins. What info do you need for your project?
 
I have roon that can run into my Denon a1h. What information do you need? I have been very busy at work so it may take me a bit to gather.
I will need a geometrical description of your speakers layout and a confirmation of the channels mapping in Roon (I'll send different config files to check the mapping)
Are the additional speakers purely "support" channels?
Or are you in a multichannel stereo mode?

Do the speakers need to be identical?
The additional speakers are support ones. It is better if they have similar frequency response to limit coloration but it is not a strict requirement as they will not work fullrange
I am buying additional surround speakers soon so I will have a 4.4 setup with 4 subs and okto dac 8 pro. At the moment I have a stereo plus 6 subs. All independent channels run with Jriver and vst plugins. What info do you need for your project?
It's the same answer as to Gwreck's question
 
Back
Top Bottom