• 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!

Speakers for BACCH

I have D&D 8C's with subwoofers and Dirac ART. I also have the uBACCH plugin, but I don't really use it. I tried the uBACCH plugin again the other day for the first time since having Dirac ART and it seems to reduce bass. So I wonder if the two technologies effectively "trip" over each other, because they both are designed to cancel sound waves in a room. For a long time I've wished that uBACCH would allow configuration of a frequency range instead of just full range. That way, one could try uBACCH outside of Diract ART's upper frequency range (150Hz).
 
So I've given the u-bacch version a try with a pair on Neumann KH 120ii, and from personal opinion and experience with different spec monitors it appears that it simply widens directivity.
as if they where wide dispersion monitors, at least in stereo.
Then again I could be doing something wrong?
Don't quote me. :)
 
So I've given the u-bacch version a try with a pair on Neumann KH 120ii, and from personal opinion and experience with different spec monitors it appears that it simply widens directivity.
as if they where wide dispersion monitors, at least in stereo.
Then again I could be doing something wrong?
Don't quote me. :)

I am sure that @STC will be along shortly. From what I understand of his writings, the purpose of crosstalk cancellation is NOT to widen the soundstage. It is to enhance the clarity of the recording. And since the recording has audible soundstage cues, enhancing the clarity of the recording also enhances the spatial cues which are naturally present in the recording. Which is why you get a wider soundstage.
 
  • Like
Reactions: STC
I am sure that @STC will be along shortly. From what I understand of his writings, the purpose of crosstalk cancellation is NOT to widen the soundstage. It is to enhance the clarity of the recording. And since the recording has audible soundstage cues, enhancing the clarity of the recording also enhances the spatial cues which are naturally present in the recording. Which is why you get a wider soundstage.
I find it has the effect of making the overall soundstage wider and more panoramic, just like you might find on a wider dispersions speaker. However, it's been my experience that those wider dispersion speakers also smear out the specificity of images with the soundfield. In contrast, something like Ubacch preserves the specificity of the images while also enhancing the space between the images, leading to the wider soundspace. The best of both worlds.
 
I find it has the effect of making the overall soundstage wider and more panoramic, just like you might find on a wider dispersions speaker. However, it's been my experience that those wider dispersion speakers also smear out the specificity of images with the soundfield. In contrast, something like Ubacch preserves the specificity of the images while also enhancing the space between the images, leading to the wider soundspace. The best of both worlds.
Yes you know what ? I agree from the testing, even if its only just through listening, we need proof here on this sight none the less.
I am personally sold on the software and will purchase it when possible. :)
 
But it does!

What I was trying to say in my post that its primary purpose is to enhance the clarity of the recording. Which lets you hear the soundstage which is captured within the recording.
 
What I was trying to say in my post that its primary purpose is to enhance the clarity of the recording. Which lets you hear the soundstage which is captured within the recording.

Keith, I understood you perfectly . I actually included a LOL emoji in my earlier post, but it seems to have been filtered out.

I think you're right that the purpose of XTC is not to artificially widen the soundstage or create a presentation that was never intended in the first place. Its primary purpose is to improve clarity by reducing crosstalk, allowing the spatial cues already present in the recording to be heard more clearly.

Thinking about it now, I may not have sufficiently considered that most people discussing XTC are using conventional stereo loudspeaker setups around 60° apart. Most stereo recordings are monitored and mastered for that geometry, so when I say that XTC is not supposed to enhance soundstage, it should be understood in the context of preserving the intended stereo presentation.

If crosstalk cancellation suddenly causes hard-panned sounds to appear fully at the ears, far beyond the loudspeaker positions, that is more likely the result of over-cancellation, or the removal of all localization cues at the ears, rather than the intended outcome.

Having said that, because of the nature of stereo recording techniques and artificial panning, some sounds may legitimately extend beyond the loudspeakers. Ideally, however, most of the main soundstage should remain around the original 60° presentation, with occasional sounds extending beyond that and ambient information extending much wider, potentially approaching 180°.

So I think we're largely saying the same thing. XTC is not there to manufacture width. It is there to let listeners hear the width, depth, and spatial information already contained in the recording more clearly, while remaining faithful to the presentation intended for conventional stereo loudspeakers.

ST
 
Keith, I understood you perfectly . I actually included a LOL emoji in my earlier post, but it seems to have been filtered out.

I think you're right that the purpose of XTC is not to artificially widen the soundstage or create a presentation that was never intended in the first place. Its primary purpose is to improve clarity by reducing crosstalk, allowing the spatial cues already present in the recording to be heard more clearly.

Thinking about it now, I may not have sufficiently considered that most people discussing XTC are using conventional stereo loudspeaker setups around 60° apart. Most stereo recordings are monitored and mastered for that geometry, so when I say that XTC is not supposed to enhance soundstage, it should be understood in the context of preserving the intended stereo presentation.

If crosstalk cancellation suddenly causes hard-panned sounds to appear fully at the ears, far beyond the loudspeaker positions, that is more likely the result of over-cancellation, or the removal of all localization cues at the ears, rather than the intended outcome.

Having said that, because of the nature of stereo recording techniques and artificial panning, some sounds may legitimately extend beyond the loudspeakers. Ideally, however, most of the main soundstage should remain around the original 60° presentation, with occasional sounds extending beyond that and ambient information extending much wider, potentially approaching 180°.

So I think we're largely saying the same thing. XTC is not there to manufacture width. It is there to let listeners hear the width, depth, and spatial information already contained in the recording more clearly, while remaining faithful to the presentation intended for conventional stereo loudspeakers.

ST
After stumbling on to this XTC concept a couple months ago and, as a completely non-techincal person, I was nonetheless intrigued. I run an LMS/piCorePlayer setup on a couple of RPI4's and was rather astonished to find out that, via the free Camilladsp add-on to piCorePlayer, XTC could run quite nicely on that little RPI4. It's been a real bumbling (and humbling) process up and down the learning curve of filters, pipelines, RACE processors, etc., but I have it more or less working. Just wanted to say that that your description of what XTC does and does not do is, in my limited experience, very accurate.
 
Last edited:
Due to real-world living room limitations, my two-speaker stereo setup is at just under 50 degrees. I still hear a very clear benefit using my xtc implementation.
 
60 degrees is not the correct angle for XTC, so I've been informed also.

I think I may have over-explained my earlier comments regarding XTC extending the soundstage.My comments were mainly in the context of crosstalk cancellation systems designed around narrow loudspeaker angles, where natural pinna cues are largely preserved and contribute to externalisation. BACCH was designed for conventional stereo loudspeaker geometry around 60°, so the psychoacoustic considerations may be different. For that reason, my earlier comments should not necessarily be taken as applying directly to BACCH.

ST
 
The role of the loudspeaker span (i.e. angle) was discussed in the original BAACH paper.

BACCHPaperV4d_p18.png
 
  • Like
Reactions: STC
So 60 degree still hold correct true as correct angle?

Mmm…..I think only the developers can answer that.

With the reservation that I am now wearing the hat of an XTC developer and naturally have my own biases, I will try to give a general answer.

XTC can certainly be implemented with a conventional 60° stereo setup, but only if one can explain how this is achieved without compromising the important role of the pinnae in binaural perception.Most XTC papers focus on recovering the recorded ITD and ILD cues by eliminating contralateral crosstalk. What is much less discussed is the role of monaural localization from the listener's own pinnae.Even with one ear, humans can localize sounds surprisingly well. If you are trying to find a cricket in a room, you naturally turn your head slightly and rely more on one ear. This is why listeners with severe hearing loss in one ear can often still localize sounds.

The earlier Ambiodipole (Stereo Dipole) work used much smaller speaker spans, around 10-20°. This not only helps filter stability and reduces coloration, but also reduces the loudspeaker's own monaural localization influence while preserving the listener's natural pinna processing.Therefore my question has never been whether 60° XTC can work. It clearly can. My question is whether it can achieve the same result as a narrow-angle Ambiodipole without suppressing or compensating for the listener's natural pinna cues. If those cues are being modified, then what mechanism is now providing externalization?That distinction is rarely discussed in the XTC literature.

We can all perform a simple experiment.Take any stereo recording and create a version where one channel contains the original signal and the other channel has been digitally attenuated to silence. The recording remains a valid two-channel stereo signal for the XTC processor.Compare the result using an Ambiodipole/Stereo Dipole style implementation and an XTC system claimed to perform equally well at 60° speaker spacing.Pay attention not only to image placement, but also to externalization and naturalness. The answer should become fairly obvious.

ST
 
Wow, thank you for the thorough response, you seem to know a fair bit of the granular on the topic and you enjoy it most importantly. :) I am barely getting into the subject for the betterment of my audio system.

What would you say is the optimal angle for it to sound closest to natural at least a starting point, if there is even such a thing?
 
  • Like
Reactions: STC
Like Tifune, I have the D&D 8C's with the B4Mac Audiophile edition. The 8C's are regarded has having quite directional output, which is helpful in achieving a high level of XTC. Reducing first room reflections is very important, and high directivity of course helps this. To further increase the sidewall and ceiling reflection damping, I constructed absorbers using frames totaling 2' deep, 4' tall, app. 6' long, loosely stacked with fluffy fiberglass (Knauf Ecobatt R-19) absorbers along the side walls. For the ceiling, I constructed three identical absorbers creating an absorptive cloud 8 feet wide and 6 feet long, with about 1 foot of airspace above it. I made all these absorbers quite thick so they will absorb frequencies throughout most of the audible range, including a good bit of the lower frequencies as well as high frequencies. So my listening space is quite non-reflective, while the rest of the room (the room is 15.5 feet wide and about 21 feet deep) remains largely untreated, but with a thick wool rug, cabinetry, and a sloped ceiling to further break up reflections. So the room itself doesn't sound dead, but there is virtually no echo. The crosswalk cancellation is profound, soundstage is accordingly very wide, and the bass is very tight. The Dutch and Dutch speakers do put out exceptional tight base, down to 20 Hz! By the way, the speakers above the 8C's are not used for stereo listening.View attachment 565773
you are currently single, obviously :).
Edgar may be costing marriages...
 
The role of the loudspeaker span (i.e. angle) was discussed in the original BAACH paper.

View attachment 568155

Thanks for the paper. Long before it, narrow angles were already shown to work best. What bothers me is the sentence “This remains to be verified experimentally.” The lab had the equipment to measure the same filter at wider angles (30°, 40°, 50°, 60° and beyond) and publish the results. That would have answered the question raised therein and helped developers still trying to make wide-angle plugins work.

My opinion is that the further the angle moves away from the optimum, the more difficult it becomes to achieve clean crosstalk cancellation without complication.

ST
 
Back
Top Bottom