What is audio meant to do?

[Fitzcaraldo215]

I find it interesting that no one has a theoretical basis upon which to base 5 channel surround recording. Various tests have been done on the multitude of surround recording methods with results not being consistent as to how one gets best fidelity in multi-channel. Stereo recording initially was based upon how human hearing works to create the illusion. Multi-channel not so much. Just an interesting effect from the video world.

Finding Blumlein recordings is difficult as they are rarely done. For everyone of those are several that employ a Blumlein pair flanked by spaced omnis. And even those are extremely uncommon. Until a few years ago Chesky recordings were done with just a Blumlein pair. They are done quasi-binaurally now.

Reskimming the chapter on Mch in Toole's latest book, it seems that after the poorly conceived disaster of '70's Quad, it was Dolby who led the way to the basic 5.0/.1 standard. Their focus was on improving cinema sound and their ideas became the basis for strikingly similar standards by the ITU, now widely used for music, and DTS, also for cinema. All have evolved somewhat over the years. But, they remain mutually congruent for 5.1.

But, I think that the reason you do not see the theoretical papers on Mch standards is the proprietary tendencies of Dolby and also of DTS. They do not share technology. I do not know what may have been published by the ITU, but they did not appear to just rubber stamp Dolby or DTS. I suspect they may have done considerable testing, experimenting and consensus building on their own.

But, really, 5.1 is an extraordinarily good and workable standard. The center channel is huge for cinema dialog, but it also pays major dividends for music. And, Toole also shows that even much higher channel counts have little benefit, although he is keenly interested in Immersive 3D.

Incidentally, Toole says that Blumlein's experimental work on stereo in the '30's was also motivated by a desire to improve cinema sound. So, music listening and all audio as we know it owes a great deal to movies. They lead the way.

Also, per Toole, 2 channels was a compromise known even in the '50's to be merely "good enough for the home". It was adopted not for techical reasons of superior sound, but only because that was all anybody knew how to put on the LP. So, there is also no compelling technical basis for 2-channel stereo sound as we know it, other than Blumlein's proofs that it was superior to the mono sound that then existed.

 

[Guermantes]

Finding Blumlein recordings is difficult as they are rarely done. For everyone of those are several that employ a Blumlein pair flanked by spaced omnis. And even those are extremely uncommon.

Perhaps this is not an issue for those with big budgets and dedicated recording sessions, but the major drawback of the Blumlein pair of figure 8s for me was the practicalities of microphone placement. In order to arrive at a satisfactory balance of direct and reflected sound, it required careful placement yet you are limited in choice if an audience is present. Spaced omnis allow flexibility in placement of the main pair: you can move them closer (e.g. on stage or in front of the audience) and then mix ambience in at the console.

This brings me to another point. We have been discussing the goal as reproducing the experience of being in the original acoustic, however microphone placement almost never reflects the perspective of an audience member. Usually (in classical recording at least) they are placed well above the average audience member's head and this is considered optimal due to the sound radiation from the acoustic instruments. The engineers/producers are ultimately more interested in how it sounds through their own monitoring equipment though aesthetic consideration is given to whether that reflects the sound of the acoustic (sorry for the pun).

So I agree with @Fitzcaraldo215, it is a soundfield that is being captured but one that is through a frame that is primarily optimised with stereo playback as its focus, not as a recreation of the audience experience.

 

[Blumlein 88]

Reskimming the chapter on Mch in Toole's latest book, it seems that after the poorly conceived disaster of '70's Quad, it was Dolby who led the way to the basic 5.0/.1 standard. Their focus was on improving cinema sound and their ideas became the basis for strikingly similar standards by the ITU, now widely used for music, and DTS, also for cinema. All have evolved somewhat over the years. But, they remain mutually congruent for 5.1.

But, I think that the reason you do not see the theoretical papers on Mch standards is the proprietary tendencies of Dolby and also of DTS. They do not share technology. I do not know what may have been published by the ITU, but they did not appear to just rubber stamp Dolby or DTS. I suspect they may have done considerable testing, experimenting and consensus building on their own.

But, really, 5.1 is an extraordinarily good and workable standard. The center channel is huge for cinema dialog, but it also pays major dividends for music. And, Toole also shows that even much higher channel counts have little benefit, although he is keenly interested in Immersive 3D.

Incidentally, Toole says that Blumlein's experimental work on stereo in the '30's was also motivated by a desire to improve cinema sound. So, music listening and all audio as we know it owes a great deal to movies. They lead the way.

Also, per Toole, 2 channels was a compromise known even in the '50's to be merely "good enough for the home". It was adopted not for techical reasons of superior sound, but only because that was all anybody knew how to put on the LP. So, there is also no compelling technical basis for 2-channel stereo sound as we know it, other than Blumlein's proofs that it was superior to the mono sound that then existed.

Oh yes, 3 channels up front would be much better than two. That was known, but no one came up with a 3 channel LP that was practical.

What I resent about Dolby is the whole standards thing. 5 channel could have been done like stereo. You get 5 speakers and a sub. You get 5 channel recordings. You probably need to have the ability to do delays on your end. Otherwise whomever records does all processing and it is baked into the recording. As processing improves you get the results in the 5 channel recording which plays over any 5 channel setup. Instead Dolby and DTS have you doing proprietary decoding on your end. Which is a way for them to come up with new versions every few years requiring you to replace a perfectly good processor with a new one or you can't get the new processing.

 

[Cosmik]

Oh yes, 3 channels up front would be much better than two. That was known, but no one came up with a 3 channel LP that was practical.

That R.E. Greene article I linked to suggested that two channels are just fine with the Blumlein pair, and you only need the centre channel when using other arbitrary arrangements that lead to 'hole-in-the-middle' stereo.

Can you simply scale the Blumlein arrangement to N channels or is there is something special about two? Does the Bumlein arrangement give superior accuracy for the image when the listener turns their head or moves around, for example?

I know that anecdotally a centre channel gives supposedly more stability, but as R.E. Greene says, that might be for less-than-perfect mic configurations where you have to use the blunt instrument of blasting some stuff out of the centre to mask the problems.

It also seems apparent that what should work best with speakers shouldn't work best with headphones and vice versa.

 

[Blumlein 88]

That R.E. Greene article I linked to suggested that two channels are just fine with the Blumlein pair, and you only need the centre channel when using other arbitrary arrangements that lead to 'hole-in-the-middle' stereo.

Can you simply scale the Blumlein arrangement to N channels or is there is something special about two? Does the Bumlein arrangement give superior accuracy for the image when the listener turns their head or moves around, for example?

I know that anecdotally a centre channel gives supposedly more stability, but as R.E. Greene says, that might be for less-than-perfect mic configurations where you have to use the blunt instrument of blasting some stuff out of the centre to mask the problems.

It also seems apparent that what should work best with speakers shouldn't work best with headphones and vice versa.

Well the primary method that leaves a hole in the middle are widely spaced omnis or widely spaced A/B miking (you take cardioids and face them straight ahead and space them apart). I don't think there is a large stability improvement for a centered listener with 3 channels. For anyone off the sweet spot however there definitely is. Even for well done recordings.

I would say the blumlein gives superior image accuracy with the listener near or in the centered position. NOS and ORTF with small spacing can give a little more leeway for being off-center giving a wider area before the image is effected.

As for more than two channels there are a some special conditions. One is that they pick up sound from the rear as well as toward the front. But rear imaging in reversed. When you play this back in stereo, right rear images are heard as front left. And vice versa. I've used this on occasion letting a few musicians stand all the way around both sides of a blumlein pair when recording in a small space. You need to listen over phones and make sure no one is directly imaged where someone on the other side is. So in essence there is no flipping the direction of such a pair in reverse. It sounds the same. You can flip polarity of one of the pair and direction is oriented 90 degrees to the side.Now if you mix a pair of bidirectional mikes, they act as one in between bidirectional mic. The front and rear sides would be halfway between them.

You can use blumlein in angles other than 90 degrees. Viewed from the direction in which you wish to record, narrowing the angle widens the stereo field captured. Increasing the angle narrows the stereo field captured.

As for other numbers of channels I'm not sure what you mean. Could blumlein pairs be used for 4 channel recordings? Or 3 channel recordings
? There are ways to do such things, but they aren't blumlein pairs anymore. In fact you can combine a crossed pair of bidirectional mikes and a coincident omni. This would allow you to create any microphone pattern and pick the direction of aim for any arbitrary number of channels you choose using only 3 recorded channels and appropriate post processing. You can also do the same thing more or less with a front and rear cardioid combined with a side facing figure 8. It is used for some surround recordings and is called double M/S.

As for listening over headphones, blumlein recordings having some imaging though narrower. Mid/side and coincident x/y can sound like an amorphous blob of imaging. I'm not quite sure why the blumlein sounds better. It doesn't sound as nice as near coincident ORTF or NOS in some instances. And I might not be the best judge as I rarely hear good imaging on phones even from binaural recording. Though I can hear enough to tell roughly whether or not two images are in the same location in the sound field. Also one could space the bidirectional mikes apart 7 inches and narrow the angle to 65 degrees which would provide a timing difference for phones while covering the same stereo angle. I've had that in mind to try, but never have gotten around to it.

Sometimes things work which seem they couldn't. Tony Faulkner has used spaced figure 8s. Spaced 7-12 inches apart and both pointed straight ahead. That works wonderfully for large orchestras. I don't know that it is accurate, but it sounds convincing.

 

[Cosmik]

As for other numbers of channels I'm not sure what you mean.

What I'm getting at is that the Blumlein arrangement is very specific - and maybe gives the 'true' image if, and only if, precise conditions are met. Sure we can play around with any mic configurations we like, and they always sound OK from a pair of speakers or from 27 surround sound speakers - most listeners are not listening for the imaging. But is there something very special about the Blumlein arrangement?

If I wanted to scale it to three channels using a coincident 'Blumlein triple' and still retain that true image, would I have to have a different (possibly as yet nonexistent) type of microphone? Or would it work with the same microphones and a wider angle between the L/R speakers, plus a centre speaker? And if it works with N=3, shouldn't it also work with N=9, or whatever, and give 'true' 360 degree surround sound? Surely it's either going to be valid for N=2 and only N=2, or it's going to be valid for any value of N as long as the microphone directionality and speaker angles are suitable..?

It's probably all academic because recordings are just not made that way - and they sound pretty good, anyway. I am, however, intrigued by the idea that true Blumlein stereo might actually be a real 'sonic hologram' in one sense through an elegant mechanism that relies on geometry, the inevitable crossfeed from the speakers, and the brain's response to composite, in some sense complementary, signals at each ear.

One motivation for ambiophonics et al. is that unwanted crossfeed gives a distorted frequency response. But as we have seen, Blumlein stereo gives a perceived directional cue that is counterintuitive. What else about it is counterintuitive? Does the Blumlein arrangement also lead to a perceived suppression of the frequency response distortion? Is it actually, at some level, precisely 'perfect' - by accident or extremely perspicacious design?

 

[sergeauckland]

Regarding microphone placement, I have found this http://www.sengpielaudio.com/Visualization-AB60-E.htm to be helpful.

As to what 'audio' is meant to do, I think we need to distinguish between a recording of an acoustic event, which can be thought of as recording an actual performance (edits notwithstanding - these make it an idealised performance) and a rock/pop recording which is made up of lots of different tracks, many of them recorded individually, with the performers never performing together at the same time, or even in the same studio. It may be tracking in one place, mixing and overdubbing in another, mastering in a third, often on different continents. There is no performance to be reproduced, the final recording on the CD is the performance.

With acoustic recordings, one can think in terms of Peter Walker's 'closest approach to the original sound', with the latter, there is no original sound to approach, so whatever you have is what it is.

My view is that I want my reproducing system to add or detract as little as possible, as evidenced by my measurements, such that what I hear is what comes off the CD, and as modified only by my room. If I like the result, I consider the recording a god one, if I don't, then I don't. I've been listening to the BBC Proms as I usually do over the summer, and find audience noise (why do people with raucous coughs go to concerts....) a distraction although the performances (and hence the recording) have been very good.

S.

ps. Americans have many very civilised qualities, so why are we adopting their unpleasant habit of applauding between movements?

 

[Sal1950]

So much of what is being discussed here has to do with soundstaging, whither 2,3,or 5 channels.
But back to basics I still yet to here a system play back a recording that if you stood outside with the windows open, would fool you that someone was playing a live instrument. Like when you walk outside and some guy down the street is playing his guitar, your ears quickly turn to the sound and you know whats happening is a live performance.
What is being missed at the transducer ends that we still haven't got there?
Or have I just not yet heard a fully transparent system?

 

[Cosmik]

So much of what is being discussed here has to do with soundstaging, whither 2,3,or 5 channels.
But back to basics I still yet to here a system play back a recording that if you stood outside with the windows open, would fool you that someone was playing a live instrument. Like when you walk outside and some guy down the street is playing his guitar, your ears quickly turn to the sound and you know whats happening is a live performance.
What is being missed at the transducer ends that we still haven't got there?
Or have I just not yet heard a fully transparent system?

According to some people around here, few commercial recordings are uncompressed. So the best system in the world wouldn't sound like a $200 guitar amp.

My theory is that room resonances create acoustic feedback which you wouldn't get playing the recording back in another room. And also 'psychological feedback' so the player plays so as to get the best sound in that room.

 

[Fitzcaraldo215]

Oh yes, 3 channels up front would be much better than two. That was known, but no one came up with a 3 channel LP that was practical.

What I resent about Dolby is the whole standards thing. 5 channel could have been done like stereo. You get 5 speakers and a sub. You get 5 channel recordings. You probably need to have the ability to do delays on your end. Otherwise whomever records does all processing and it is baked into the recording. As processing improves you get the results in the 5 channel recording which plays over any 5 channel setup. Instead Dolby and DTS have you doing proprietary decoding on your end. Which is a way for them to come up with new versions every few years requiring you to replace a perfectly good processor with a new one or you can't get the new processing.

Well, yes, Dolby and DTS are proprietary, and they went through many iterations. All their older standards, however, appear to be bundled into HT processors and players, if you want to use them, including for upmixing from stereo.

But, things changed for the better with Bluray and the emergence of lossless codecs. They are basically just compression schemes, and I am able to use them legally on my PC embedded in JRiver. Possibly, JRiver is paying royalties for this, but I doubt it. So, compressed DTS HDMA and Dolby True HD deliver just 5.1 or 7.1 after decompression to LPCM and require no other special processing, unless your speakers need distance correction via DSP delays. Other processing, like room correction, bass management, etc. is optional.

For Mch music, which is primarily on SACD, as long as you have an SACD player, you are good to go in 5.1. Except, DSD does not by its nature support DSP on most all processors and players.

So, with discrete Mch recordings, I have no need for the confusing list of inferior, old legacy Dolby or DTS schemes, just the lossless codecs for BD. The only major enhancements to those I see are Immersive 3D, if one wishes to go there. I don't for now.

And, by the way, the only processor in my setup is JRiver on my PC. I use no prepro or AVR.

 

[cjfrbw]

I have used surround sound in every speaker music system since the original Advent Soundspace Control in the late 70's. With the Soundspace, I used four large Advent speakers.

Went from that the Yamaha DSP1 in the mid 80's, and on to the various Yamaha AVRs.

I have never had anybody who heard these systems even comment on the surrounds during playback. This tech is on just about every AVR and particularly Yamaha, and is seldom used except for home theater. It is strange over the years having wave after wave of audiophiles "re-discovering" surrounds and ambience. I got tired of explaining it's benefits, which are numerous.

You don't need MC recordings, and even MC recordings often aren't properly recorded for the ambience except to produce obvious effects. My goal was always to have the benefits of ambience in the system without it ever being specifically identifiable as a separate part of the playback.

Otherwise, there is nothing wrong with my system that a few million dollars couldn't cure.

 

[Blumlein 88]

What I'm getting at is that the Blumlein arrangement is very specific - and maybe gives the 'true' image if, and only if, precise conditions are met. Sure we can play around with any mic configurations we like, and they always sound OK from a pair of speakers or from 27 surround sound speakers - most listeners are not listening for the imaging. But is there something very special about the Blumlein arrangement?

If I wanted to scale it to three channels using a coincident 'Blumlein triple' and still retain that true image, would I have to have a different (possibly as yet nonexistent) type of microphone? Or would it work with the same microphones and a wider angle between the L/R speakers, plus a centre speaker? And if it works with N=3, shouldn't it also work with N=9, or whatever, and give 'true' 360 degree surround sound? Surely it's either going to be valid for N=2 and only N=2, or it's going to be valid for any value of N as long as the microphone directionality and speaker angles are suitable..?

It's probably all academic because recordings are just not made that way - and they sound pretty good, anyway. I am, however, intrigued by the idea that true Blumlein stereo might actually be a real 'sonic hologram' in one sense through an elegant mechanism that relies on geometry, the inevitable crossfeed from the speakers, and the brain's response to composite, in some sense complementary, signals at each ear.

One motivation for ambiophonics et al. is that unwanted crossfeed gives a distorted frequency response. But as we have seen, Blumlein stereo gives a perceived directional cue that is counterintuitive. What else about it is counterintuitive? Does the Blumlein arrangement also lead to a perceived suppression of the frequency response distortion? Is it actually, at some level, precisely 'perfect' - by accident or extremely perspicacious design?

Well the sengpiel visualizer sergeauckland and I have linked to would offer some insight. You get to change angles and see what recorded angles are. Blumlein is only special in that the geometry and directionality works best at those angles for mike and speaker. And works more accurately for placing objects in the virtual soundfield than other types of miking or other directional patterns. Alan Blumlein arrived that those ideas mathematically. And of course had the idea to let the crossfeed create an effective delay because he knew how hearing worked. The magic of the crossed 8s is that intensity picked up by each mike varies in a cosine function with the angle off axis.

Now you could add more channels. For instance your third 8 could be oriented vertically, and mixed into the other two. This would allow some control over vertical aiming.

 

[Blumlein 88]

I had seen this illustration, and finally remembered where. It is a JA article on miking in Stereophile.
https://www.stereophile.com/content/stereo-image

The top ideal stereo stage shows even segments and how they would be perceived in a left right sense. The other illustrations show how the result upon reproduction differs from that idea stereo portrayal.

 

[Fitzcaraldo215]

Most interesting. Thanks. I have heard Blumlein recordings, mainly from WaterLily, engineered by Kavi Alexander assisted by his acolyte Robt. E. Greene. I used the correct +- 45 degree speaker placement. One - In Nature's Realm, a recording of Dvorak by the Philadelphia Orchestra under Wolfgang Sawallisch - was featured on the front cover of TAS decades ago to high praise by HP himself.

Except, interesting though it was in some ways, it was a dud in terms of capturing the sound of that orchestra in that hall, Philly's Academy of Music. So say I and countless other orchestra subscriber audiophiles I know who are quite familiar with that sound live. Now, we know that hall is a bitch because it is much too dry. It is an opera house modeled on La Scala, not a symphony hall. Major labels - Columbia, RCA, EMI, etc. - went through hoops to avoid it and find other recording venues for their commercial releases. Kavi's release only makes it sound even much drier than the real thing, whatever slightly improved frontal sounstaging qualities it might have had. I have been similarly underwhelmed by Kavi's other releases, including those by the then Leningrad Philharmonic.

So, these are not necessarily a full spectrum proof or disproof of Blumlein technique. But, it never ceases to amaze me how the myth of Blumlein simply does not die, in spite of the almost total lack of recordings that give any credence to the mythology. Yes, Robt. E. Greene, a former math professor, waxes poetic about its theoretical perfection while he greatly overstates the sonic quality of recordings he himself helped make with Kavi. I venture to say that just about every classical recording engineer on the planet has studied and experimented with Blumlein. But, it survives as pure legend with no compelling recorded evidence of its alleged superiority to other less theoretically perfect recording techniques. To me, it fails to deliver.

So, for me, even if I could find more than a handful of Blumlein recordings, I say, no thanks. Discrete hirez Mch using the ITU standard, on the other hand, is quite voluminous and much, much more satisfying in delivering a satisfactory replica of the real thing live.

 

[Cosmik]

...interesting though it was in some ways, it was a dud in terms of capturing the sound of that orchestra in that hall, Philly's Academy of Music.
...it never ceases to amaze me how the myth of Blumlein simply does not die, in spite of the almost total lack of recordings that give any credence to the mythology.

If it only worked in an anechoic chamber I would still be impressed.

In practice, mic'ing instruments close up and then placing them in the correct L/R position using a common or garden pan pot is still 'Blumlein' and still impressive. And then of course, arbitrary amounts of de-correlated ambient wash can be piped into the mix or from surround speakers and it will sound wonderful.

Any old arbitrary, creative playing around with mics will sound OK because people are listening for the performance, not the imaging, but it doesn't detract from the amazing idea of 'correct' stereo being featured in the very first patent for the whole system.

 

[Guermantes]

I had seen this illustration, and finally remembered where. It is a JA article on miking in Stereophile.
https://www.stereophile.com/content/stereo-image

The top ideal stereo stage shows even segments and how they would be perceived in a left right sense. The other illustrations show how the result upon reproduction differs from that idea stereo portrayal.

Interesting. For orchestral recording, DPA seem to recommend A-B microphone setups with spaced omnis approximately 40-60 cm apart:
https://www.dpamicrophones.com/mic-university/classical-orchestra-a-b-stereo

Their recommended position is some metres above or behind the conductor:

The position of the main stereo pair will also be the perspective of the future listener; therefore the sound engineer's goal is to create the illusion of natural perspective in placing the main stereo pair. In doing so, there is not necessarily any correlation between the actual placement of the microphones and the best seats in the hall, and consequently, placing the stereo pair correctly takes much effort. The distance to the orchestra and the height of the microphones above floor level should be adjusted for equal coverage of the different orchestra sections and for the amount of ambience required in the recording. Usually the optimum position is above or right behind the conductor's podium at a height of between three and four meters, so that no musicians obscure the instrument behind them. A good rule of thumb is: if you can see the sound source you can also hear it.
. . .
The use of A-B Stereo techniques without support microphones can create an extremely convincing depth in the stereo image and capture a realistic room impression. The sound sources, ie musical instruments and room reflections, are picked up with the correct time alignment relative to the placement of the main stereo pair, which explains why this method is often regarded as the purist's choice.

 

[Fitzcaraldo215]

If it only worked in an anechoic chamber I would still be impressed.

In practice, mic'ing instruments close up and then placing them in the correct L/R position using a common or garden pan pot is still 'Blumlein' and still impressive. And then of course, arbitrary amounts of de-correlated ambient wash can be piped into the mix or from surround speakers and it will sound wonderful.

Any old arbitrary, creative playing around with mics will sound OK because people are listening for the performance, not the imaging, but it doesn't detract from the amazing idea of 'correct' stereo being featured in the very first patent for the whole system.

Not sure what you are trying to say. Recordings that only work in anechoic listening are doomed to obscurity and commercial failure, except perhaps for the niche of binaural. And, neither you or I know how these particular Blumlein recordings would sound anechoically. From all accounts from scientists and engineers who work with anechoic chambers, it does not seem they go home and try to make their listening rooms anechoic. For example, Floyd Toole sure does not.

And, if people really don't care about the imaging, let's go back to mono. Hipsters are into it because of the Beatles' mono rereleases. So, maybe, that's all we need.

Personally, I think moving toward the goal of ever higher fidelity means not only improving frequency response, distortion, noise, dynamics, etc., but also improving imaging on the recording side as well as the playback side. Hi fi exploded in popularity only thanks to the promise of stereo. Before that, in mono days, it was obscure and probably less significant than, say, ham radio among electronics hobbyists.

 

[Blumlein 88]

Not sure what you are trying to say. Recordings that only work in anechoic listening are doomed to obscurity and commercial failure, except perhaps for the niche of binaural. And, neither you or I know how these particular Blumlein recordings would sound anechoically. From all accounts from scientists and engineers who work with anechoic chambers, it does not seem they go home and try to make their listening rooms anechoic. For example, Floyd Toole sure does not.

And, if people really don't care about the imaging, let's go back to mono. Hipsters are into it because of the Beatles' mono rereleases. So, maybe, that's all we need.

Personally, I think moving toward the goal of ever higher fidelity means not only improving frequency response, distortion, noise, dynamics, etc., but also improving imaging on the recording side as well as the playback side. Hi fi exploded in popularity only thanks to the promise of stereo. Before that, in mono days, it was obscure and probably less significant than, say, ham radio among electronics hobbyists.

I think Cosmik meant recorded in an anechoic chamber.

While stereo may have exploded hifi popularity. You can't say the same for multi-channel. I understand your position, and agree. But multi-channel has been met with an overwhelming "who cares" in the mainstream hifi market. It has been that way for a generation. It may grow some, but I think that will be due to VR use rather than music or hifi.

Trying to interest some of the amateur musicians I've recorded in surround recordings has proven futile. Will it matter over headphones? Will my car/suv do surround? I've got a soundbar for the TV, but its not much good for music. I don't have any way to play surround recordings. If we ever do a video we'll do that. So on and so forth. They care very much that you record them in stereo. Multi-channel even when they've heard it and liked it just doesn't gain any traction with them.

 

[Guermantes]

Multichannel recording for the masses: https://www.zoom.co.jp/products/handy-recorder/h2n-handy-recorder

 

[Blumlein 88]

Multichannel recording for the masses: https://www.zoom.co.jp/products/handy-recorder/h2n-handy-recorder

They are doing a variation on something I mentioned earlier. They have basically 3 cardioids. They record each separately into the 3 channels. After the fact or during it as well, you can create most of the possible microphone types and directionalities. Back to back cards are an omni if in phase and a figure 8 if out of phase. In fact you can take the output mix it both ways and recombine later for other patterns.