Hello ambiophonic lovers!
I am very interested in xtc filters, but I'm not versed in digital filters.
Acoustics and loudspeaker design is my thing.
I have spoken with Ralph Glasgal and Stephan Hotto over email about R.A.C.E.
They are in agreement that it is necessary to cut off filtering around 4khz to avoid irratic dips and spikes in frequency.
I want to have a discussion about the BACCH filter and if it solves this problem.
If it does solve this problem, can we replicate this filter for the DIY community, or alter it in a way to avoid patent infringement?
I have read this article http://www.princeton.edu/3D3A/Publications/BACCHPaperV4d.pdf
It is too involved and in depth for my tastes
Again....I know little about digital filters.
I have made this a new thread topic because there has been little activity on the ambiophonics thread, also I want to concentrate on the BACCH filter, not R.A.C.E.
I am very grateful to Ralph and the ambiophonic insitute for their extensive work and unrelenting pursuit in bringing ambiophonics to the forefront or audio reproduction, but I'm even more grateful that this technology is not patened.
I read that the folks at Princeton want to team up with TV manufacturers to implement their filter! TV manufacturers?
I want these new findings in filter design to be put to good use, not mass marketing crapola.
I am very interested in xtc filters, but I'm not versed in digital filters.
Acoustics and loudspeaker design is my thing.
I have spoken with Ralph Glasgal and Stephan Hotto over email about R.A.C.E.
They are in agreement that it is necessary to cut off filtering around 4khz to avoid irratic dips and spikes in frequency.
I want to have a discussion about the BACCH filter and if it solves this problem.
If it does solve this problem, can we replicate this filter for the DIY community, or alter it in a way to avoid patent infringement?
I have read this article http://www.princeton.edu/3D3A/Publications/BACCHPaperV4d.pdf
It is too involved and in depth for my tastes
Again....I know little about digital filters.
I have made this a new thread topic because there has been little activity on the ambiophonics thread, also I want to concentrate on the BACCH filter, not R.A.C.E.
I am very grateful to Ralph and the ambiophonic insitute for their extensive work and unrelenting pursuit in bringing ambiophonics to the forefront or audio reproduction, but I'm even more grateful that this technology is not patened.
I read that the folks at Princeton want to team up with TV manufacturers to implement their filter! TV manufacturers?
I want these new findings in filter design to be put to good use, not mass marketing crapola.
Hi, the equations in CHoueri´s papers are mighty scary!
I do find R.A.C.E to be not as effective as a physical barrier in the high frequencies but, in my experience, a small, unobstrusive piece of foam extending some inches from between the tweeters minimizes the problem. I did hear some colorations in the past, but after extended trial and error- including an about face rearrangement of speakers and listening position to get rid of the colorations that the room was adding on its own- I no longer hear any.
It is an interesting proposition that Choueri´s algorithms may improve on R.A.C.E. I am still an old-fashioned hobbyist and will need for someone to put out an inexpensive little box as per ambio4you so I can try it one day. No more USB, DACS and computers messing up my listening experience, thanks.
I think this thread could be merged with the ´Try ambiophonics in your room´
one. After all ambiophonics is all about crosstalk cancellation regardless of the specif method to achieve it. I, for one, would love to hear from someone who holds his ground and is still using a mattress today.
I do find R.A.C.E to be not as effective as a physical barrier in the high frequencies but, in my experience, a small, unobstrusive piece of foam extending some inches from between the tweeters minimizes the problem. I did hear some colorations in the past, but after extended trial and error- including an about face rearrangement of speakers and listening position to get rid of the colorations that the room was adding on its own- I no longer hear any.
It is an interesting proposition that Choueri´s algorithms may improve on R.A.C.E. I am still an old-fashioned hobbyist and will need for someone to put out an inexpensive little box as per ambio4you so I can try it one day. No more USB, DACS and computers messing up my listening experience, thanks.
I think this thread could be merged with the ´Try ambiophonics in your room´
one. After all ambiophonics is all about crosstalk cancellation regardless of the specif method to achieve it. I, for one, would love to hear from someone who holds his ground and is still using a mattress today.
Yeah, I read in one of your posts that you toe out the speakers. I'm sure that enhances xtc at higher frequencies because of the narrowing polar response.
I'm trying to start a small business designing and building loudspeaker systems. So a physical barrier is a hard sell
I have a new design for an open baffle sound bar, I know.....how the heck is that possible?
Anyway, after years of designing loudspeakers for stereophonics, I discovered ambiophonics. ( I know you can use any speakers for ambiophonics ) but, best to stick with open baffles and horns to reduce room reflections.
One of the reasons for starting this thread is to see if someone had the know how to decode BACCH for use of the DIY community or some how make it an open source for everyone, because unfortunately it is patented. That's why you won't see BACCH used in something like the miniambio for a very long time.
Also, I'm hoping to get the word out on ambiophonics, I have given a great review for the iPad app, and I leave little comments on the web here and there. I also tell anybody who will listen. To me, stereo is dead. Especially now that audio is becoming more personal and less of a social event.
I love the RACE filter, but if there is something better, than I think we should pursue it. I just want that pure binaural sound
I'm trying to start a small business designing and building loudspeaker systems. So a physical barrier is a hard sell
I have a new design for an open baffle sound bar, I know.....how the heck is that possible?
Anyway, after years of designing loudspeakers for stereophonics, I discovered ambiophonics. ( I know you can use any speakers for ambiophonics ) but, best to stick with open baffles and horns to reduce room reflections.
One of the reasons for starting this thread is to see if someone had the know how to decode BACCH for use of the DIY community or some how make it an open source for everyone, because unfortunately it is patented. That's why you won't see BACCH used in something like the miniambio for a very long time.
Also, I'm hoping to get the word out on ambiophonics, I have given a great review for the iPad app, and I leave little comments on the web here and there. I also tell anybody who will listen. To me, stereo is dead. Especially now that audio is becoming more personal and less of a social event.
I love the RACE filter, but if there is something better, than I think we should pursue it. I just want that pure binaural sound
Never too late to attend a thread
No, as far as I can see, BACCH do not solve the problem of high freqs.
It is termed as "Optimal Crosstalk Cancellation" and the point of optimisation was freq response irregularities of the filter to avoid dynamic range problems.
Take a look of Figure 7, which is, as I see, the best Choueiri filter can do. No cross talk cancellation at the bass, and at the problematic high freqs cancellation is not performed.
It is a classical trade off - freq response of the filter (loudspeaker signals) vs. cross talk cancellation performance.
There is no 'magic'
- Elias
It sure took Choueiri a lot of math to come up with the wrong answer
As I analyze it (I'm not a fan of ambiophonics myself) there are at least two problems that need to be addressed. In order for cancellation to occur, exactly the same number of wavelengths between the speaker and the ear must occur for both speakers and they must arrive precisely out of phase at the same amplitude all frequencies. That's the performance criteria based on the concept.
One problem is that the number of wavelengths is directly dependent on the frequency (Lambda = v/f where Lambda is the wavelength, v is the speed of sound and f is the frequency.) Before even beginning it is assumed that the speed of sound is the same at different frequencies. If that's not true then an additional time delay correction must be applied. So the time delay must be a function of frequency or the same number of wavelengths reaching the ear will only occur at one frequency. Even where arrivals are out of phase at another frequency, say a multiple it will be for a different number of wavelengths where amplitude at that frequency will have changed.
The other is that amplitude of sound reaching your ear is a function of angle of arrival from the speaker and frequency especially for the tweeter due to its amplitude versus angle characteristics. And that varies with frequency as can be seen from a polar response plot. So if the amplitude is correctly cancelled completely at one frequency it won't be for another especially in the top couple of octaves. This is why the 4 khz cutoff. As frequency gets higher and wavelengths get shorter the precise relationship becomes more difficult to achieve and occurs even approximatly over a smaller region of space. One way around that is to aim the speakers at a point midway between your ears since the falloff of output reaching each ear will be the same because it's at the same relative angle (on the opposite side). So being close on axis will reduce this falloff and make them symetrical.
If these criteria are not met then you will not only get cancellations at some frequencies but reinforcements at others, hence the spectral distortion. I don't see how a frequency correction filter solves the geometry problem. I suppose the argument can be made that phase shift and hence variable time delay as a function of frequency is inherent in the filter itself especially if it is analog but its FR would look like a roller coaster.
What I found remarkable about ambiophonic sound is not merely that it can achieve its goal to the degree that it does but that it works at all even at a single point. That being said I don't see that BAL in any form addresses the issue of accurate sound reproduction of music. It also makes listening to recorded music with whatever advantages XTC offers a solo experience where thousands can listen to live music and enjoy the benefits of acoustics at the same time and place.
As I analyze it (I'm not a fan of ambiophonics myself) there are at least two problems that need to be addressed. In order for cancellation to occur, exactly the same number of wavelengths between the speaker and the ear must occur for both speakers and they must arrive precisely out of phase at the same amplitude all frequencies. That's the performance criteria based on the concept.
One problem is that the number of wavelengths is directly dependent on the frequency (Lambda = v/f where Lambda is the wavelength, v is the speed of sound and f is the frequency.) Before even beginning it is assumed that the speed of sound is the same at different frequencies. If that's not true then an additional time delay correction must be applied. So the time delay must be a function of frequency or the same number of wavelengths reaching the ear will only occur at one frequency. Even where arrivals are out of phase at another frequency, say a multiple it will be for a different number of wavelengths where amplitude at that frequency will have changed.
The other is that amplitude of sound reaching your ear is a function of angle of arrival from the speaker and frequency especially for the tweeter due to its amplitude versus angle characteristics. And that varies with frequency as can be seen from a polar response plot. So if the amplitude is correctly cancelled completely at one frequency it won't be for another especially in the top couple of octaves. This is why the 4 khz cutoff. As frequency gets higher and wavelengths get shorter the precise relationship becomes more difficult to achieve and occurs even approximatly over a smaller region of space. One way around that is to aim the speakers at a point midway between your ears since the falloff of output reaching each ear will be the same because it's at the same relative angle (on the opposite side). So being close on axis will reduce this falloff and make them symetrical.
If these criteria are not met then you will not only get cancellations at some frequencies but reinforcements at others, hence the spectral distortion. I don't see how a frequency correction filter solves the geometry problem. I suppose the argument can be made that phase shift and hence variable time delay as a function of frequency is inherent in the filter itself especially if it is analog but its FR would look like a roller coaster.
What I found remarkable about ambiophonic sound is not merely that it can achieve its goal to the degree that it does but that it works at all even at a single point. That being said I don't see that BAL in any form addresses the issue of accurate sound reproduction of music. It also makes listening to recorded music with whatever advantages XTC offers a solo experience where thousands can listen to live music and enjoy the benefits of acoustics at the same time and place.
Last edited:
Hello ambiophonic lovers!
I am very interested in xtc filters, but I'm not versed in digital filters.
Acoustics and loudspeaker design is my thing.
I have spoken with Ralph Glasgal and Stephan Hotto over email about R.A.C.E.
They are in agreement that it is necessary to cut off filtering around 4khz to avoid irratic dips and spikes in frequency.
I want to have a discussion about the BACCH filter and if it solves this problem.
If it does solve this problem, can we replicate this filter for the DIY community, or alter it in a way to avoid patent infringement?
I have read this article http://www.princeton.edu/3D3A/Publications/BACCHPaperV4d.pdf
It is too involved and in depth for my tastes
Again....I know little about digital filters.
I have made this a new thread topic because there has been little activity on the ambiophonics thread, also I want to concentrate on the BACCH filter, not R.A.C.E.
I am very grateful to Ralph and the ambiophonic insitute for their extensive work and unrelenting pursuit in bringing ambiophonics to the forefront or audio reproduction, but I'm even more grateful that this technology is not patened.
I read that the folks at Princeton want to team up with TV manufacturers to implement their filter! TV manufacturers?
I want these new findings in filter design to be put to good use, not mass marketing crapola.
Perhaps this might help, if you have $$
http://www.pppl.gov/video/WC21SEP2011_EChoueiri_2-H.264/WC21SEP2011_EChoueiri_2-H.mov
wow......not once did he mention the ambiophonics institute, or anybody involved with it.
He's kind of an apple fanboy too. It figures he would go with a company who's ecosystem is on lockdown and extremely proprietary.
Look, I understand having patents on intellectual property......but why limit this property to big companies? Why do we have to by a license for use? He could just make a piece of hardware like the miniambio. That way any speaker designer could take advantage of the filter, with much better results than the jambox or MacBook air. That way he's selling an actual physical piece rather than just licensing fees. I'm sure someone in his position could find a hardware producer to calibrate with, instead of just selling intellectual rights to companies with already existing products.
Sorry, I'm very passionate about discovery for the good of humankind rather than soley monitary reasons.
He's kind of an apple fanboy too. It figures he would go with a company who's ecosystem is on lockdown and extremely proprietary.
Look, I understand having patents on intellectual property......but why limit this property to big companies? Why do we have to by a license for use? He could just make a piece of hardware like the miniambio. That way any speaker designer could take advantage of the filter, with much better results than the jambox or MacBook air. That way he's selling an actual physical piece rather than just licensing fees. I'm sure someone in his position could find a hardware producer to calibrate with, instead of just selling intellectual rights to companies with already existing products.
Sorry, I'm very passionate about discovery for the good of humankind rather than soley monitary reasons.
"I'm copying and pasting what I posted on the ambiophonic thread"
Ok, I listen to the demo's on this site.
I compared the pink floyd's "money" demo, to my own set up with RACE.
My set up with the speakers at 5 degrees spread with attenuation at 6db, delay set to 22.7 microseconds, sounded much better. More separation, more stable imaging, and equal in timbre to BACCH.
I know that the BACCH demo wasn't optimized.......
It doesn't say at what angle to place the speakers on the site, so I placed them from 5 degrees to 60 degrees in 5 degree increments.
The most important comparison was that RACE equaled the BACCH filter in timbre
Ok, I listen to the demo's on this site.
I compared the pink floyd's "money" demo, to my own set up with RACE.
My set up with the speakers at 5 degrees spread with attenuation at 6db, delay set to 22.7 microseconds, sounded much better. More separation, more stable imaging, and equal in timbre to BACCH.
I know that the BACCH demo wasn't optimized.......
It doesn't say at what angle to place the speakers on the site, so I placed them from 5 degrees to 60 degrees in 5 degree increments.
The most important comparison was that RACE equaled the BACCH filter in timbre
Choueiri said. “When you create an invention, it is your responsibility to help the University during the process.”
U. licenses professor?s new speaker technology - The Daily Princetonian
A year since the last reply but due to a link to Professor Edgar Choueiri's work I stumbled over this post. I thought this might be interesting:
IR Library Page
IR files from Prof. Edgar Choueiri dating back to 2006 uploaded by Ralph Glasgal.
I haven't ever used IR files but I'm sure we can figure out if it works?
IR Library Page
IR files from Prof. Edgar Choueiri dating back to 2006 uploaded by Ralph Glasgal.
I haven't ever used IR files but I'm sure we can figure out if it works?
Should be interesting to try, here's the Professor on the subject in 2011: (source: http://www.mail-archive.com/sursound@music.vt.edu/msg00546.html)
So it looks like we need X-Volver or one of the other convolver plugins to try it.
Any help from an experienced IR user?
Edgar Choueiri said:
So it looks like we need X-Volver or one of the other convolver plugins to try it.
Any help from an experienced IR user?
Those are quite old filters. I don't know if they count as BACCH.
I have tried them with Convolver — a convolution plug-in
The convolver works very well with windows media player. If you use the VST you need to read the VST notes carefully. If you get things wrong it can be difficult to recover sometimes, even by deleting the files. You need to create a text config file and select .txt in the drop down box to see it, and not try to load a .wav file directly. Once it works, it works well.
The config file I found for those filters looked like this:
44100 2 2 0
0 0
0 0
C:\Impulses\EYCv2L-44.wav
0
0.0
0.0
C:\Impulses\EYCv2L-44.wav
1
0.0
1.0
C:\Impulses\EYCv2R-44.wav
0
1.0
0.0
C:\Impulses\EYCv2R-44.wav
1
1.0
1.0
There are a couple of mine here:
Index of /Filters
I have also used VoxengoPristineSpace (free trial) VST with these filters.
Let me know here, if you need help.
Thanks for that, David.
Sadly they were the only files I found from Professor Edgar Choueiri. I do get it isn't the actual Bacch filters.
I have bought and tried the Race VST plugin from Mosc and played with that as well. That has a Wareing preset based on your work right?
The few Bacch demo's floating around seemed promising enough to want to try more. They sure gave a sense of envelopment, even in a more traditional stereo angle setup. Too bad no newer files are around, can't find any on ambiophonic.net but lot's of clues they were once there.
Can you tell a bit more about your own filters?
Edit; got it working in JRiver Media Center, seems these are indeed very old filters that suffer from a hole in the middle like mentioned on the electro-music.com forums: http://electro-music.com/forum/post-194966.html (see the post from durwood)
So there have been later versions floating around at one time.
Last edited:
To have a knob or setting named after one is an honour, I believe, a bit like a Blue Peter badge. ( BBC - CBBC - About Blue Peter Badges )
Those filters are my Hybrid (HYbrid Bi-Rectangular Interaural Delay ) XTC filters. They were the best I could do at the time. The particular ones there were crossed over to mono at 200Hz to save the public's small speakers being driven too hard by off centre bass. Aimed at Jambox, really.. I have some with lower crossovers, or none at all, if it sounds promising.
My filters should give a flat response for mono and stereo at the ears and for mono input at speakers (apart from one peak )
The BACCH filters seem more aimed towards a flat speaker response but with compromises for XTC and the direct sound. It probably depends on where you are compared to the reverberation radius and whether you like bloated bass as to which sounds flatter. He seems, ahem, to have some problems with crossovers..
Those filters are 3 sample delay filters, so the same as the EYC . See around http://www.diyaudio.com/forums/mult...ns-beaten-behringer-what-219.html#post3423206 - the same basic XTC filter was used with some extra added controls for https://www.dropbox.com/s/k0dzhjli29...re%20ambio.mp3
Actually it is only the 'both ears mono' that is flat. You cannot have it flat everywhere and cancel.
http://www.wareing77.plus.com/Filters/DW4-frequency.png
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
You really should not listen in stereo or headphones. These filters can sound awful unless you get the angles and distance right.
Ralph has largely given up support for convolution methods in favour of his own RACE method. Can't blame him, particularly with the slagging off of Ambiophonics to sell BACCH.. 3D3A Lab at Princeton University
Well I did try the ambiophonics setup and it is remarkable with the vst plugin.
I played for a while (with my ancient big Yamo D-365 speakers) in various setups and got quite pleasing results with the Race plugin.
Not so much with the old files from Prof. Choueiri.
His demo that you linked in the Linkwitz thread still sounds good though, in a stereo 60 deg setup. I could not get that same envelopment with the race plugin no matter what I tried.
I thought I had something playing "Fell in love with a boy" from Josh Stone with the backing vocals coming from the sides but it turned out this was even true without the ambio plugin engaged! Must be in the mix already
.I noticed the points made on Ambiophonics on that site! It does seem the main focus for the BACCH is to make money.
A shame really, would have been cool to be able to try it on some music of our own choice. The intro from Pink Floyd sounds nice but off somehow. I guess you are right about the crossover, it does not sound coherent from low to high.
I would tell Sony the all secrets of binaural for $3M too:Well I did try the ambiophonics setup and it is remarkable with the vst plugin.
I played for a while (with my ancient big Yamo D-365 speakers) in various setups and got quite pleasing results with the Race plugin.
Not so much with the old files from Prof. Choueiri.
His demo that you linked in the Linkwitz thread still sounds good though, in a stereo 60 deg setup. I could not get that same envelopment with the race plugin no matter what I tried.
I thought I had something playing "Fell in love with a boy" from Josh Stone with the backing vocals coming from the sides but it turned out this was even true without the ambio plugin engaged! Must be in the mix already
I noticed the points made on Ambiophonics on that site! It does seem the main focus for the BACCH is to make money.
A shame really, would have been cool to be able to try it on some music of our own choice. The intro from Pink Floyd sounds nice but off somehow. I guess you are right about the crossover, it does not sound coherent from low to high.
"Sony has given Choueiri $3 million to fund his lab for the next three years"
Princeton Alumni Weekly: BREAKING GROUND: Unlocking the key to 3D sound
You are losing me a bit on what you have tried. What is running on JRiver? Convolver? ConvolverVST? Robin's VST? other?
It is worth trying small (computer) speakers close up to see what the imaging is like, before trying to scale up.
Which Prof. Choueiri demo did I link to?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.