3D audio report

[Patrick Bateman]

I stumbled across ambiophonics about a year ago, while working on a car project. I tried it with my Summas at home, but found that the effect was more pleasant in the car. I've started a new thread on my forum to discuss it here:

http://forum.audiopsychosis.com/viewtopic.php?f=3&t=7

It's pretty shocking how wide a soundstage can be achieved with the speakers literally inches apart. I tried audiomulch and some other tricks with my Summas, but found the effect distracting and phasey. Basically the soundstage got wider, but the artifacts weren't worth the trouble. The *physical* barrier is intriguing though.

In defense of ambiophonics, I didn't invest the time to tweak the settings like I should have. I have a feeling that one could get very impressive results with a pair of Summas and a day's worth of software tweaking.

The thing that really piques my curiosity about crosstalk cancellation is the fact that waveguides could be a partial substitute for a physical barrier. I'll dig into the hows and whys of that on my forum, as time permits.

 

[Gedlee]

Hey John - welcome. I just joined your forum. I like a lot of your writtings if for just the humor.

As Prof Cheri says, directivity determines limit of crosstalk cancellation. The algorithims can only go so far and after that every7thing comes down to directivity. He wants me to investigate an even narrower directivity, but then when I told hiom what it would take, he kind of choked. Profs with grants have no concept of business with margins.

 

[noah katz]

The thing that really piques my curiosity about crosstalk cancellation is the fact that waveguides could be a partial substitute for a physical barrier.

How so?

 

[Ralph Glasgal]

I am glad you have tried Ambiophonics. Clearly if it works with a set of speakers in a car or elsewhere then there can be no fundamental problem. Generally if you hear anything phasey, the amount of cancellation is excessive due to wrong settings or the speakers are too far apart. If you cancel crosstalk that is not really present then you are listening to two speakers that are a bit out of phase (polarity) and that is often described as phasey. Ambiophonics is a bit harder to adjust than a simple stereo system, but most listeners find it well worth it. There are lots of hints on adjustments at www.ambiophonics.org.

Ralph Glasgal

 

[metronimo]

As Prof Cheri says, directivity determines limit of crosstalk cancellation. The algorithims can only go so far and after that every7thing comes down to directivity.

That's why I might order two Harper kits from you. I built small enclosures with 4" FR drivers especially to use the BACCH filter, but since your speakers were found to be the most directive by Professor Choueiri, I should get them instead of trying to build anything else (like econowave speakers)...

He wants me to investigate an even narrower directivity, but then when I told hiom what it would take, he kind of choked. Profs with grants have no concept of business with margins.

This grant is a good thing, and I expect that Professor Choueiri will release updated free filters for general usage. I wonder if he has tried ultrasonic directive speakers like the "Audio Spotlight"... Dr. Geddes, what do you think about those unusual ultrasonic speakers?

 

[Gedlee]

Dr. Geddes, what do you think about those unusual ultrasonic speakers?

They have an Achiles Heal that you never hear about and that is efficiency. Their efficiency drops with frequency-squared and isn't all that good at 1 kHz. So lets say it takes 1 watt at 1 kHz to get a reasonable level, then at 100 Hz it takes 100 watts (for just a "resonable level", nothing loud) and at 10 Hz thats 10,000 watts. The transducers can only handle about 100 watts max thermally. Clearly they are never going to get the LFs required for music and they can't be used with traditional sources ala subs, because they don't blend with these sources very well. They work well for voice as long as the bandwidth is not taken too low.

 

[metronimo]

They have an Achiles Heal that you never hear about and that is efficiency. Their efficiency drops with frequency-squared and isn't all that good at 1 kHz. So lets say it takes 1 watt at 1 kHz to get a reasonable level, then at 100 Hz it takes 100 watts (for just a "resonable level", nothing loud) and at 10 Hz thats 10,000 watts. The transducers can only handle about 100 watts max thermally. Clearly they are never going to get the LFs required for music and they can't be used with traditional sources ala subs, because they don't blend with these sources very well. They work well for voice as long as the bandwidth is not taken too low.

In Montréal there will be a public architectural installation using similar ultrasonic speakers, so I will have the opportunity to listen to them. I wonder how they would sound in a stereo configuration (even with their limited SPL and reproduction bandwidth).

 

[metronimo]

Clearly they are never going to get the LFs required for music and they can't be used with traditional sources ala subs, because they don't blend with these sources very well. They work well for voice as long as the bandwidth is not taken too low.

I emailed Holosonics and quickly received a reply; the frequency response of the smallest panel is about 300Hz-16Khz. Why wouldn't they blend well with traditional speakers (and subs)?

 

[Gedlee]

I emailed Holosonics and quickly received a reply; the frequency response of the smallest panel is about 300Hz-16Khz. Why wouldn't they blend well with traditional speakers (and subs)?

An abrupt change in directivity - not smooth at all.

 

[metronimo]

An abrupt change in directivity - not smooth at all.

I don't think it would matter for ambiophonics, because a sweet spot is required, and because:

As Prof Cheri says, directivity determines limit of crosstalk cancellation. The algorithims can only go so far and after that every7thing comes down to directivity.

Ultrasonic speakers are probably not well suited for conventional stereo because interaction with the room is important. But I'm just thinking out loud; I have no listening experience with either CD and ultrasonic speakers.

 

[Gedlee]

I don't think it would matter for ambiophonics, ...

I was not restricking my thoughts to ambiosonics. I would always design for standard stereo since I don't think that ambiosonics can support a custom speaker design.

 

[durwood]

I was not restricking my thoughts to ambiosonics. I would always design for standard stereo since I don't think that ambiosonics can support a custom speaker design.

I think you should not confuse ambioSONICS or AMBIsonics was the term pioneered by Michael Gerzon with ambioPHONICS. They are too completely different ideas. Ambisonics was an interesting surround sound encoder/decoder method that was compatible with many forms of playback reproduction. Unfortunately it lost out to Dolby in the commercial arena due to funding issues.

The closely spaced speakers then processed with xtc goes by many names: stereo dipole, transaural, ambiophonics, optimal source distribution, all have slightly different ideas on what the "ideal" setup is supposed to be. One of the most in depth XTC papers I have seen was the early work done by Dr. Keele in the 80's.
http://www.xlrtechs.com/dbkeele.com/papers.htm

I don't see it as a gimic at all and it has usefulness like all the other reproduction techniques, Dolby, Logic7,VBAP, etc. The thing with all reproduction techniques is that they aim to over-ride the listening space by attempting to recreate some type of spaciousness that was captured initially. If you are too close to the boundaries of the room, that reproduction technique is probably not suited for the room size unless you use more speakers, more channels. This is where these closely spaced payback setups will excel over typical "Stereo", you can delay the side wall reflections adjusting the direct/reflect ratio.

The one thing about XTC closely spaced pairs is that in theory, it is less sensitive to off-center seating creating less errors over stereo setups. What makes the sweet spot smaller are higher levels of XTC and what gives that un-natural phasey sound some people have trouble with.

So to completely write it off claiming stereo is the reference is incorrect. Let's be honest here, monophonic is the natural standard since that is what occurs daily in our lives. The problem is that monophonic relies heavily on the environment to give it the spacial feeling. Last I checked, most people don't have rooms equal to what was setup during the recording.

 

[durwood]

Oh yes, ambiophonics only works at a single seating location. I think the better choice for playback of this type will be insert earphones. I am doing a large scale study of these for a client and I'll tell you right now, no loudspeaker even comes close to them for "perfection" of the impulse response. This makes them ideal for ambiophonics.

No actually ambiophonics is not suited for headphone use, it's actually horrible for it, the complete opposite of what you want for that playback reproduction system. Headphones already remove the crosstalk AND the azimuth angle is about 80deg off for each side. The problem with listening to stereo recordings with headphones is that stereo was optimized for speakrs placed at +/-30deg azimuth angles and headphones sit at +/-90 azimuth angles. This is why headphone freaks use crossfeed circuits. Now if you merely mean using convolution and HRTF inpulses to match the azimuth angle of headphones, that is a whole other concept, but let's not confuse them.

 

[metronimo]

One of the most in depth XTC papers I have seen was the early work done by Dr. Keele in the 80's.
http://www.xlrtechs.com/dbkeele.com/papers.htm

The 2 parts paper is titled:
"The Effects of Interaural Crosstalk on Stereo Reproduction and Minimizing Interaural Crosstalk in Nearfield Monitoring by the Use of a Physical Barrier"

After reading it last year I made a temporary physical barrier using two 4'x8' polystyrene foam isolation panels (it took me 10 minutes to setup). I was in heaven for a month then I had to remove them because they were too impractical (and dangerous). I doubt than electronic cross talk cancellation techniques can offer the same listening experience, but they are getting closer to the "real thing".

Let's be honest here, monophonic is the natural standard since that is what occurs daily in our lives.

With proper cross-talk cancellation, mono is restored; centered audio (left=right) really is centered, since the speakers are in front of the listener. With conventional stereo, mono is just smeared between two distant speakers.

Last I checked, most people don't have rooms equal to what was setup during the recording.

Listening rooms are not "audio holodecks" and stereo mixes are made accordingly. The characteristics of recording rooms (like concert halls) are either captured live or faked with reverbs, but then the listening room and the cross talk are adding even more audio transformations. Constant directivity helps reducing the interaction with the room, but the most important is to get rid of the cross-talk.

I suspect that conventional stereo is one of the cause for the decline of so called "classical music", because of it's poor ability to restore the sound field normally heard in a good concert hall. The most difficult and dense orchestral music is a joy to listen when there's no cross-talk; without it, the brain works too much.

 

[metronimo]

The problem with listening to stereo recordings with headphones is that stereo was optimized for speakrs placed at +/-30deg azimuth angles and headphones sit at +/-90 azimuth angles. This is why headphone freaks use crossfeed circuits.

Crossfeed circuits (or filters) can remove some of the disconfort, but they can't move the stereo imaging in front of the listener.

 

[Gedlee]

Could we move the discussion of "techniques" to the appropriate thread. I design for stereo - it still seems to me to be the standard.

 

[Kal Rubinson]

I suspect that conventional stereo is one of the cause for the decline of so called "classical music", because of it's poor ability to restore the sound field normally heard in a good concert hall. The most difficult and dense orchestral music is a joy to listen when there's no cross-talk; without it, the brain works too much.

I am not so sure. Classical music has been a niche for my entire lifetime and it is unlikely that any technical advances in media will change that. OTOH, one should note that on HD/MCH media, like SACD and BR, classical music has flourished greatly out of proportion to its representation on conventional media (CD, FM radio, etc.). This supports your argument that the demands of classical music are greater and that seems to be recognized, implicitely, by the classical listener.

Kal

 

[amirm]

Could we move the discussion of "techniques" to the appropriate thread. I design for stereo - it still seems to me to be the standard.

Yes. Given the interest in the topic we will likely create a new subforum dedicated to it. Please stay tuned.

 

[steve williams]

thread has been moved to new forum

 

[metronimo]

I am not so sure. Classical music has been a niche for my entire lifetime and it is unlikely that any technical advances in media will change that.

Kal,

To enjoy many types of recorded music, including classical music, people must have access to decent reproduction technologies, ideally in their childhood. I would'nt say that classical music is a niche in the western world, but it is certainly difficult to appreciate in bad listening conditions. Classical music must be heard live at least a few times in order to leave strong impressions in the listener brain; but then if he can't recall those impressions while listening to recorded music, then he might just stick to whatever "sounds good". That's why most pop music are acoustically simple and/or spectacular enough on anything that can spit out some noise from recordings. Also, there's a lot of bad classical music recordings, so that doesn't help.

With better reproduction and listening techniques, it's much easier to expand our musical culture and stay in touch with the real acoustical world. For me, listening with conventional stereo is a limit to this expansion.

OTOH, one should note that on HD/MCH media, like SACD and BR, classical music has flourished greatly out of proportion to its representation on conventional media (CD, FM radio, etc.).

Hi Definition (like 24bits/96Khz) doesn't make much difference; well produced 16bits/44.1Khz Compact Discs have enough definition. Classical music enthousiasts are just being channeled to niche markets by media companies.