Best Bets for Binaural?

[Mark A. Jay]

Originally Posted by rimrock

I think Motion Tracked Binaural is trying to do this (Duda at UC Davis). The idea is to record with multiple ears - each pair of ears being set at a different rotational angle, and then, during playback, to interpolate between the recordings as the listener turns their head.

That's interesting...my take on tracking is here Best Bets for Binaural? but the idea of recording with multiple ears is another way. Still, it would seem that the presence of multiple ears would cause diffraction and reflection - or maybe I am misinterpreting you.

Anyway, check my musings in the link above.

[rimrock]

Originally Posted by Mark A. Jay

a) place the HRTF-capable transducer in the space and location where it will be used to record.
b) obtain the BRIRs there taking care to rotate / angle etc the HRTF-capable device so that an array of BRIRs can be computed; tag these with a capable coordinate system for later reference. *
c) record performance
d) play performance over headphones equipped with sensors to determine head position
e) convolve (in real time?) the HRTF-capable transducer position-dependent BRIR with the playback based on the coordinates produced by the head tracking wares; match the position to the corresponding position's BSIR and convolve

Mark,
Isn't the problem here that the recording can only be made at one transducer orientation? and that the recording will then already have convolved the room with the directional qualities of said transducer.

[Mark A. Jay]

Originally Posted by rimrock

Mark,
Isn't the problem here that the recording can only be made at one transducer orientation? and that the recording will then already have convolved the room with the directional qualities of said transducer.

I'm not sure I'm fully getting your point, but let me put this another way.

What I am getting at is that if we record in a space whose BRIRs (or BSIRs as I prefer to call them as pointed out in another related post) are highly dissimilar, then many of them must be acquired if a 'convincing' simulation is possible. That is, minute changes in head rotation and angle result in large changes in the BSIRs or alternately, the complex FRFs. When the differences in BSIRs is small (negligible in the case of fully diffuse fields) then there's no need for head tracking software.

Assuming that you have a high-resolution look-up table of BSIRs, and that the head tracking apparatus can allow the proper HRTF to be matched, then the playback should match. This is true because the same transducer was used to acquire all of the complex FRFs (and thus the BSIRs). If you were to acquire some BSIRs with 'transducer x' and other BSIRs with 'transducer Y' and still others with 'transducer z' then a normalization would have to be applied - that is, any two of the three transducers normalized to the remaining one (the reference) before those BSIRs could be used in a real-time convolution scheme. In theory (just thinking out loud nere...) you could un-do the nesting effect of the transducer by knowing its impulse response as well. Granted, the acoustic impulse response is many orders of magnitude longer than is the impulse response of the transducer, but it's definitely 'in there'.

Anyway, by taking all of the BSIRs with a given type of transducer any normalization that needed to be done would be the same normalization (i.e. to impose a different or preferred frequency response for example) across the library of BSIRs.

Again, how much head tracking 'matters' (in my opinion) is wholly governed by the amount of similarity between BSIRs, and this is goverened by the acoustics of the boundary conditions; in a highly diffuse environment the BSIRs will all be (pretty cloe to) copies of one another (or very, very close). However, as the conditions morph from diffuse to full free-field (at the opposite end of the decay time spectrum) then the differences observed in the BSIRs will become much, much more disparate (and of ever-increasing import in terms of an accurate synthesis based on head motions). As a consequence, having a high-resolution array of BSIRs (as a function of angle and elevation) would be critical for the simulation to be as convincing as it could be whereas in a highly reverberant (diffuse) space, the differences in BSIRs are bound to be much, much smaller.

Sorry - I started to re-hash this topic (I have posted other thoughts on this in this and other threads...so I'll get off my soap box now).

[rimrock]

Mark,
I can see your idea working in binaural synthesis but not for a live binaural recording. Once a binaural recording has been made it doesn't appear possible to unravel all the directional information embedded in the two signals - or is this the idea?
Using your technique wouldn't you need to record a large set of audio signals which had a similar directionality to your measured BRIRs - and then convolve the two? Could you get these signals from a spherical array microphone? It could well be that I've only got a simplistic understanding of what you're suggesting.

[Mark A. Jay]

Originally Posted by rimrock

Mark,
I can see your idea working in binaural synthesis but not for a live binaural recording. Once a binaural recording has been made it doesn't appear possible to unravel all the directional information embedded in the two signals - or is this the idea?
Using your technique wouldn't you need to record a large set of audio signals which had a similar directionality to your measured BRIRs - and then convolve the two? Could you get these signals from a spherical array microphone? It could well be that I've only got a simplistic understanding of what you're suggesting.

Firstly, I wouldn't exactly call this 'my' technique - these are musings / observations based on the published works of others - but thanks for the 'nod'.

No, you would not need to record the performance at multiple head angles / elevations. In fact, this is why the BSIRs and convolution must be done - because you cannot record the sound at a given location for all those angles; the mannequin head (or similar - more on that later) can only be in one place / orientation during recording. Thus, you do need to have a large array of BSIRs that can be used, in conjunction with head-worn tracking software to perform the convolution of the recording in real time (during playback).

As an aside, my contention is that binaural (as-is) renders the most life-like playback of the systems out there. While one can easily argue that it may not deliver the desired aesthetic for a particular use, I think it's a fair contention that it's the most life-like playback when using good headphones; a great deal of 'truth' about the space and the performance is revealed when binaural recording / playback are employed.

Now, back to the convolution...

I have to state once more that, in my opinion, the perceived necessity for head tracking (and it's a "necessity" only if one wants the playback to mimic what a listener would hear in-situ, freely turning their head this way and that) comes down to the acoustical boundary conditions; the more diffuse (reverberant) the less critical the need for head tracking. Moreover, the number of BSIRs needed would (I should think) drop dramatically as compared to the number required for free-field and near-field conditions (I am kind of lumping the two conditions together because in both settings, the differences in BSIRs will be greatest whereas in diffuse fileds differences will be minor to arguably non-existant).

So, there are two approaches that one could take. First, there could be a sort of 'boiler-plate' approach in which the angle of rotation and elevation (increments) for which BSIRs are to be acquired, regardless of boundary conditions, is set. Think of this as a protocl of sorts. That is, 'Method A" could specify that the BSIRs be acquired every 1 degree of rotation and elevation (for example). This would result in a huge library of BSIRs which would lessen the need for a convolved solution to iterate.

In other words, suppose the protocol was to acquire BSIRs every 90 degrees (admittedly very, very coarse spatial resolution) then when the peron turned their head to any angle between the 90-degree raster, there would be no reliable BSIR and the playback system would have to 'guess' (that is, interpolate a sort of estimated BSIR based on the two known BSIRs) as to what the BSIR would be at that angle of rotation and elevation. Again, this lack of spatial resolution would not matter in wholly diffuse conditions (reverberant chamber etc), but in reality, boundary conditions, apart form an acoustics lab, are seldom wholly diffuse.

A second, and more realistic approach would be to constrain the angles (for which BSIRs must be measured) to typical angles of rotation and elevation that the human head can achieve, and this would lessen the size of the array of 'valid' BSIRs. Now, when the listener, who must be wearing the head tracking hardware for head tracking to work properly, turns their head, the correct BSIR would be called from the (much-smaller) library and used, in real time, for that particular head position, and the need to interpolate would be pretty much eliminated because of the higher spatail resolution in the library of BSIRs.

As for a spherical array, it's not the same thing (mathematically) as a binaural mannequin, but in many cases it can achieve similar results - and I think much of the similarity between binaural mannequins and similar approaches boils down to the acoustic boundary coinditions. So, based on that observation, my gut tells me that a spherical array (aka binaural sphere) could be used to generate the SRIRs (spherical room impulse response) in certain instances.

If you want to know more of these techniques (far more than I know), then I strongly suggest you look at the text "Communication Acoustics", edited by Jens Blauert as well as the AES Monograph on binaural sound written by Mlle Rozen Nicol. Those texts have been very useful for me, and I highly recommend them for anyone with an interest in the subject matter.

Lastly, please remember that these are my opinions, and I am certain there are aspects that I may have over-simplified or even misunderstood, but these are my opinions based on my comprehension to date of the subject matter.

[AlexKall]

Hi everyone,

First, a quick introduction. My name is Alex Kall, and I operate a binaural recording service named Kall Binaural Audio in Halifax, Nova Scotia, Canada. We've only been open since this past December, but are making good progress educating Nova Scotian musicians about the possibilities of binaural audio. You can check us out at www.kallbinauralaudio.com, or hear some binaural samples here.

A potential client recently asked for my opinion about a standard identifier for binaural recordings. That is, once a binaural recording is released, how does the public know they should listen to it on headphones, other than writing "LISTEN WITH HEADPHONES!" on the album cover.

Is anyone aware of any kind of icon (like the little "CD" icon that tell you a disc is a CD and not a DVD) that exists for binaural audio? How do you all make sure your listeners know that a recording is binaural? Metadata works well for MP3s, but what about other file formats, and physical distribution?

I'd really appreciate your thoughts.

Cheers,
Alex

[Doc Vigilanti]

Originally Posted by AlexKall

Hi everyone,

First, a quick introduction. My name is Alex Kall, and I operate a binaural recording service named Kall Binaural Audio in Halifax, Nova Scotia, Canada. We've only been open since this past December, but are making good progress educating Nova Scotian musicians about the possibilities of binaural audio. You can check us out at www.kallbinauralaudio.com, or hear some binaural samples here.

A potential client recently asked for my opinion about a standard identifier for binaural recordings. That is, once a binaural recording is released, how does the public know they should listen to it on headphones, other than writing "LISTEN WITH HEADPHONES!" on the album cover.

Is anyone aware of any kind of icon (like the little "CD" icon that tell you a disc is a CD and not a DVD) that exists for binaural audio? How do you all make sure your listeners know that a recording is binaural? Metadata works well for MP3s, but what about other file formats, and physical distribution?

I'd really appreciate your thoughts.

Cheers,
Alex

What I would do is go ahead and make a logo of your own design that designates it as "binaural" and work it into your already existing marketing scheme. Then I would get into contact with record companies and the likes and attempt to get them to go along with the logo. I'd trademark the logo first though.

Pretty much it's all up to you on how you want it to look and then up to you to get it standardized.

[Mark A. Jay]

Originally Posted by AlexKall

Hi everyone,

A potential client recently asked for my opinion about a standard identifier for binaural recordings. That is, once a binaural recording is released, how does the public know they should listen to it on headphones, other than writing "LISTEN WITH HEADPHONES!" on the album cover.

Is anyone aware of any kind of icon (like the little "CD" icon that tell you a disc is a CD and not a DVD) that exists for binaural audio? How do you all make sure your listeners know that a recording is binaural? Metadata works well for MP3s, but what about other file formats, and physical distribution?

I'd really appreciate your thoughts.

Cheers,
Alex

Alex:

A couple of thoughts...

As someone who is very 'big' on binaural, I've run into a lot of that which you have described, and I wish that I had a simple answer for you (or answers). To wit, one battle that we as purveyors of binaural audio face is that binaural can be seen as a sort of 'new format' when in fact, it's anything but. That's great for technically inclined folks such as people who work in recording, production, etc. who understand the large and small differences between binaural and conventional stereo / multi-channel, but your question has some very real teeth to it; how do you get the general public to be aware of the merits of binaural without going into a lesson on physics and perceptual audio first?

All that I can say is this: keep pushing the exposure however you have to. It's been a very common experience for me that when someone who has not heard binaural before first hears it, they are taken aback / blown away, because its realism is quite unlike anything else.

However, you should really peruse this thread for some of my comments as well as those of Dallas Masters. I say that not as a slam, but instead, to call your attention to the marketing vehicle that has typically served conventional audio (2-channel stereo) well, which is to say "radio" (well, at least historically). You have to remember that to reach a wide audience with binaural, you have to forego things like broadcast radio simply because most listeners will be using speakers, and without the benefit of crosstalk-cancelling hardware, a true binaural recording (depending upon the acoustics of the space in which it was made) just won't sound "right' to a casual listener.

In a way, you have to almost let the product speak for itself, and to that end, I think the recent push towards music on mobile devices (specifically, phones) may help further expose the technique. Indeed, just as it was / is well-suited to the mp3 player-based consumer, the mobile device user (for music) are almost certainly going to use headphones (regrettably, inexpensive ear buds more often than not, but at least they are headphones).

In essence, you have to target markets in which headphones are not an after-thought or something that the listener has to order on-line, or buy from a store, but instead are the assumed default playback device. So to this end, the mobile device / mp3 market is the natural 'outlet' for the technology.

As far as the metadata go, yes, I do that in all of the binaural work that I release. That is, all of my work bears (as either part of the song title, the overall title of the recording, or elsewhere) the tag line "Engineered For Headphone Use". I thought about saying something like "binaural" as part of the meta data, but the fact is, 99.9% of the people who read that word will not have a clue what binaural (intrinsically) means. I find that by stating the recording was "Engineered For Headphone Use" there's no expectation that the end user has an interest in / understanding of physics and or perceptual audio, and they "get" that they are supposed to use headphones. Yes, I agree, on a CD cover or download or whatever, there should always be a sort of 'about this recording" type thing, so that the consumer has to option to read about the technique and thus, become better educated about it as a natural consequence.

An aside: Incidentally, if you or others on this thread are interested, check out the Group "Perceptual Audio" on Linkedin (and yes, I created a heaphone-based logo that I use for my recordingg company (Immersifi) and then modified it for the Group logo as well). Anyway, I moderate the Group, and it's essentially a Technical meeting / exchange forum, with almost no commercial aspects to it (by design) - it's a haven for geeks who are into such things and who wish to focus on / debate the various technical aspects. If you (Alex...and the general Public) are on Linkedin and interested to join that Group, feel free to send a request - you can get to my linkedin page via my signature and from there, find the Perceptual Audio Group.

Anyway, those are my thoughts. Again, the best to you, and indeed, keep pushing, but always remember - look for people for whom headphone use is not a chore, task, or option, but instead, their default and preferred listening mechanism. After all, binaural really does 'sell' itself when headphones are used, so in a way, it can self-promote via the proof that is indeed in the pudding that is binaural.

[RScott]

I'm a big fan of binaural and surround techniques and most anything that leads to a great sense of envelopment in sound or really any virtual space.

As mentioned before, head tracking is very important to envelopment, especially envelopment meant to simulate a space we can move around in. In the simulation of sound, it all changes as we sway our heads, look around, etc. Having binaural not recreate that can really limit options. Even with 5.1 surround sound, an actual sound space is created and our brain verifies this as we move our heads around.

The BEST thing about binaural recordings is the recreation of a continuous sound space. Even 5.1 in a great setup only localizes really well at each speaker and between the front pairs in a phantom image. Localizing within the speaker circle is also hard as things remain just as distant but louder.

The worst part of binaural recordings is IMO the lack of a convincing front sensation. Almost all binaural recordings are very successful behind the viewer, close to the head, and sometimes above the head. But I don't think I've heard a great in-front illusion yet.

It is difficult or impossible to create a convincing "phantom" center between two binaural impulses by fading between them, from my experience. It's certainly much easier with loudspeakers. The image seems to break down. IMO, further work is needed to create an engine that can synthesize a continuously variable impulse for each sound event. This is bad news for games, probably the best bet for binaural sound, as this will require a limited number of sound events and/or lots of computing power.

In Pro Tools, I used a 7.1 fader with each "speaker" routed to a pair of binaural impulses for that location. The results were unimpressive. It may be possible to use a similar setup in order to see how "close" two binaural impulses need to be in order to get acceptable fades by using the surround panner with only a portion of a circle, say 20 or 30 degrees split 7 ways, and panning between them.

World's first video game with no video: binaural iPhone gaming - Ewan McIntosh | Digital Media & Learning
Papa Sangre is a sound-only iPhone binaural game.

Interested and hopeful of how this technology will be used. I'm hugely hopeful for a head-tracking surround sound or binaural system in the future. I think it's the holy grail for immersion.

[John Willett]

Originally Posted by RScott

I'm a big fan of binaural and surround techniques and most anything that leads to a great sense of envelopment in sound or really any virtual space.

As mentioned before, head tracking is very important to envelopment, especially envelopment meant to simulate a space we can move around in. In the simulation of sound, it all changes as we sway our heads, look around, etc. Having binaural not recreate that can really limit options. Even with 5.1 surround sound, an actual sound space is created and our brain verifies this as we move our heads around.

The BEST thing about binaural recordings is the recreation of a continuous sound space. Even 5.1 in a great setup only localizes really well at each speaker and between the front pairs in a phantom image. Localizing within the speaker circle is also hard as things remain just as distant but louder.

The worst part of binaural recordings is IMO the lack of a convincing front sensation. Almost all binaural recordings are very successful behind the viewer, close to the head, and sometimes above the head. But I don't think I've heard a great in-front illusion yet.

It is difficult or impossible to create a convincing "phantom" center between two binaural impulses by fading between them, from my experience. It's certainly much easier with loudspeakers. The image seems to break down. IMO, further work is needed to create an engine that can synthesize a continuously variable impulse for each sound event. This is bad news for games, probably the best bet for binaural sound, as this will require a limited number of sound events and/or lots of computing power.

In Pro Tools, I used a 7.1 fader with each "speaker" routed to a pair of binaural impulses for that location. The results were unimpressive. It may be possible to use a similar setup in order to see how "close" two binaural impulses need to be in order to get acceptable fades by using the surround panner with only a portion of a circle, say 20 or 30 degrees split 7 ways, and panning between them.

World's first video game with no video: binaural iPhone gaming - Ewan McIntosh | Digital Media & Learning
Papa Sangre is a sound-only iPhone binaural game.

Interested and hopeful of how this technology will be used. I'm hugely hopeful for a head-tracking surround sound or binaural system in the future. I think it's the holy grail for immersion.

Interesting - and I agree with a lot of this.

I have listened to really immersive sound only twice:-

The first was a binaural dummy-head recording where I was sitting in the same spot that the microphone was when the original recording was made. So I had all the visual clues and I got excellent magry all round, even in front (which I don't normally get).

The second time was listening to a 9.1 system (5.1 + 4 extra monitors for height information) - that was superb.

[RScott]

Also, in regards to the OP's question of "What does binaural need in order to break into mainstream use?"

It needs to ditch the niche! What I mean is than when you put "you must listen with headphones" on the front of your album or whatever, you instantly remove 95% of your audience or more. You're left with people who are either adventurous or interested specifically in the artist or music or technology.

Instead, make a liner note explaining that binaural technique was used and that a special bonus is provided when using headphones. Binaural audio will still sound pretty good over speakers. If you make people listen to your effect, the album instantly becomes about the effect, not the music. Let the binaural experience be a bonus for those willing to give it a go. 5.1 is similar. Some people can't be bothered. Those who take the time to set up their systems correctly get a great treat with good 5.1 mixes. But the general audience is not alienated either. They get to watch what they came for -- the movie. Only a select few will be listening because of binaural technique. The rest are there for one thing -- the music. Give it to them in whatever way they want.

There is a need for more general purpose recordings made with binaural technique with less emphasis on showing off the effect. Similar to 3D films constantly throwing things at the audience and jumping out of the screen. If this is all binaural is good for, it will be discovered, enjoyed for 5 minutes and then discarded as a cool little trick.

[John Willett]

Originally Posted by RScott

Also, in regards to the OP's question of "What does binaural need in order to break into mainstream use?"

It needs to ditch the niche! What I mean is than when you put "you must listen with headphones" on the front of your album or whatever, you instantly remove 95% of your audience or more.

But do you?

Seeing how much music is downloaded and listened to on an iPod or the like.

Maybe - "recorded for the iPod" may be better than "you must listen on headphones" ;-)



Posted via the Gearslutz iPhone app

[Mark A. Jay]

Originally Posted by RScott

...The BEST thing about binaural recordings is the recreation of a continuous sound space. Even 5.1 in a great setup only localizes really well at each speaker and between the front pairs in a phantom image. Localizing within the speaker circle is also hard as things remain just as distant but louder.

The worst part of binaural recordings is IMO the lack of a convincing front sensation. Almost all binaural recordings are very successful behind the viewer, close to the head, and sometimes above the head. But I don't think I've heard a great in-front illusion yet.

It is difficult or impossible to create a convincing "phantom" center between two binaural impulses by fading between them, from my experience. It's certainly much easier with loudspeakers. The image seems to break down. IMO, further work is needed to create an engine that can synthesize a continuously variable impulse for each sound event. This is bad news for games, probably the best bet for binaural sound, as this will require a limited number of sound events and/or lots of computing power.

Interested and hopeful of how this technology will be used. I'm hugely hopeful for a head-tracking surround sound or binaural system in the future. I think it's the holy grail for immersion.

All valid points.

Some time ago, I created the Perceptual Audio Group on Linkedin (as mentioned), and indeed, this front-to-back (or similarly, poor front localization) is a subject that is getting a lot of attention in signal processing circles - and I posted a question to the Group Members on the topic there. However, my take on it (vis a vis` the question I posted) was along the lines of quantifying that effect in the presence / absence of corresponding visual stimuli.

That is, I was asking if anyone had done structured and controlled juried listening tests wherein the jurors were asked to rate the realism of the image under two conditions:

1) With audio-only, and no accompanying video, and

2) The same audio, only this time accompanied by video shot from the same location and perspective as the audio content was acquired.

3) and possibly, the experiment could be expanded to a third tier in which the visual stimuli are now presented in 3-D, or alternately, if the jurors were to observe an orchestra / musical act 'faking it' in sync with the music while listening. That is, the jurors would have no knowledge that the band / orchestra was not playing.

I'm not the only one who has written about how the visual and auditory cortexes interact when it comes to localization, so I'm curious about this - seems to me a perfect research project for a Graduate Student.

I also think this structured approach would be interesting to undertake with a second set of audio, namely, an ORTF / Jeklin Disc, or spaced-pair substituted for the binaural audio. After such a study (with proper statistical controls to weed out unreliable jurors) I think it would be VERY interesting to look at the localization results depending upon the type of audio and presence / absence of visual stimuli, and further, what type of visual stimuli (of the types I mentioned before).

I'm part of a fair number of juried listening studies in my 'day job', and it's the only real way for us to ascertain differences between product sounds (i.e. when comparing the sounds of powered seat adjusters, though this technique is used extensively for all manner of product sounds - everything from vacuum cleaners, to drills, to blenders, to washing machines to...audio system playback, to...well...you get the point). Indeed, in the Perceptual Research community, calibrated and controlled listening studies have become the de facto method for determining either specific or holistic differences in sounds. It seems to me that the technique would be easily adaptable to the experiment that I proposed in the Perceptual Audio forum.

As far as head tracking, I see its validity, but I disagree that it's required for any and all binaural recordings. I say this because one must think about the uniqueness of the HRTFs as a function of azimuth, elevation, and so on - and all of these variables are a function of the acoustical boundary conditions in which the recordings (and thus corresponding HRTFs) would be made. In other words, as the field become more and more diffuse, the HRTFs (as a function of azimuth and elevation) become more and more alike until finally, in a wholly diffuse filed, the HRTFs are independent of azimuth and elevation. In such a space, sound cannot be localized, so the need for head tracking simply does not exist. However, a fully diffuse field is not good for music reporduction (not in the least).

On the other hand, as the boundary conditions migrate more and more towards free-field, then yes, this is where I see the greatest benefit for head tracking methodologies, but even then, this begs the question of how quickly on-the-fly convolution and interpolation between all of the availabe HRTFs can be accomplished. There is (and has been) a lot written about this in journals and I suspect that it will be some time before real time convolution, or more appropriately, interpolation between known HRTFs can be done without hiccups as the interpolation is being done.

On the other hand, going back to the boundary conditions, the need to interepolate between the HRTFs (in real time) dininishes as the field becomes increasingly diffuse. So in essence, more resolution (a greater number of HRTFs) is required for head tracking in free field conditions than would be in diffuse fields. So, at present, my guess is that the interpolation process appears more 'seamless' for sounds recorded in semi-diffuse fields, again, because the differences between the HRTFs get smaller and smaller as the field gets more diffuse.

Anyway, I do urge those who peruse these threads to have a look-see at the Percpetual Audio Group on Linkedin. There is a very large cross section of disciplines and industried represented by its Members, and that (IMO) promotes some lively and interesting discussions and debates on a variety of subjects related to such matters.

[Plush]

Why is a poster once again talking about ditching headphones for binaural listening?

Binaural recordings require the listener to hear the recording on headphones.
There is a principle of physics involved.

Speaker listening can present a credible stereo image but you're not hearing any binaural attributes.

[NorseHorse]

What I got from the poster was that binaural would become more accepted if it wasn't promoted as "for headphones only" - as if implying that it sounds bad on speakers. Instead, when releasing a binaural recording, just say that it is sounds extra good, special, whatever, on headphones.

"If you make people listen to your effect, the album instantly becomes about the effect."

I think any mention of binaural should just be a little footnote. If it sounds good, I don't really care whether it's binaural or not. It's been around long enough, and it doesn't need to be "coddled" as a recording technique.

[Plush]

No I don't think that washes. Binaural is not an "effect." It is a type of recording for headphones.

Binaural lives or dies with embracing or abandoning headphones. It is a technique based solely on headphone listening.

One can use the Neumann dummy head as a good stereo mic too and have it stand on its own or blend it with your other mics But don't look for binaural cues when listening on loudspeakers.

So those who posit that it is OK to play true binaural material on loudspeakers have just advocated destroying the purpose of the recording.

[RScott]

Originally Posted by Plush

No I don't think that washes. Binaural is not an "effect." It is a type of recording for headphones.

Binaural lives or dies with embracing or abandoning headphones. It is a technique based solely on headphone listening.

One can use the Neumann dummy head as a good stereo mic too and have it stand on its own or blend it with your other mics But don't look for binaural cues when listening on loudspeakers.

So those who posit that it is OK to play true binaural material on loudspeakers have just advocated destroying the purpose of the recording.

I understand the binaural cues won't translate. I also know that headphones only recordings will never repay their costs. So, it's back to the OP's predicament.... what shape will the format have to assume to be successful? It will almost certainly require some dual purpose as a value-added feature such as a special headphones only mix.

Binaural recording is no less an effect than using dual cameras to make 3d films. Loudspeaker development has been attempting to find ways to isolate audio streams to only one ear. It may not be the purist form of imaging you imagine, but I think that the core ideas behind using hrtfs to localize sound may well have a place on loudspeakers at some point.

If binaural recording exists as a "take it or leave it" headphones only technology, most people will just leave it. So will any for profit binaural productions.

Sent from my SPH-D700 using Gearslutz.com App

[Plush]

Originally Posted by RScott

I understand the binaural cues won't translate. I also know that headphones only recordings will never repay their costs. So, it's back to the OP's predicament.... what shape will the format have to assume to be successful? It will almost certainly require some dual purpose as a value-added feature such as a special headphones only mix.

Binaural recording is no less an effect than using dual cameras to make 3d films. Loudspeaker development has been attempting to find ways to isolate audio streams to only one ear. It may not be the purist form of imaging you imagine, but I think that the core ideas behind using hrtfs to localize sound may well have a place on loudspeakers at some point.

If binaural recording exists as a "take it or leave it" headphones only technology, most people will just leave it. So will any for profit binaural productions.

Sent from my SPH-D700 using Gearslutz.com App

Binaural was first proposed in the 1880's and it is a specialized audience format. It's an old format and tacking head tracking on to it is a not so great idea. I like immersive sound too but the Germans who are the leaders in immersive sound and planar sound environments are best left to another discussion. Binaural is what it is and one either experiments with it and can embrace it or is less enthused by it. It's success is hardly judged with a criterion of making money in the record bidniz.

It's not really that expensive to record or release binaural recordings. After all, one can buy the best binaural device, the Neumann dummy head, for $8000 and then record every concert with it along side the regular recording set-up. Then when a good result is obtained with the binaural set-up, one releases it along side the stereo or surround version. It can piggy back on the regular stereo release.

This is how I did it at the Milwaukee Symphony when I was there and as far as I know they are still recording every show with the Neumann head as well as with their regular recording set-up. We were the first to do it and the first to release a series of binaural only recordings. They are still available from iTunes or directly on the MSO web site.

I clearly disagree that it is an effect. An effect requires creation of the effect. Binaural playback virtues range from subtle to dramatic. In one recording for example, the listener may have an increased sense of place ( a certain seat) in the concert hall. More dramatic renderings might present the picture of the orchestra with a chorus on raised platforms as having the chorus be clearly heard as being higher up than at stage level.

Analogy is the weakest form of argument:
3D photography is a manipulation of images whereas binaural recording is a microphone placement that does its thing on its own. It does it automatically because of the design of the system.

I am a practitioner not a theorist.

[RScott]

Originally Posted by Plush

Binaural was first proposed in the 1880's and it is a specialized audience format. It's an old format and tacking head tracking on to it is a not so great idea.

Depends on the application's goals. Maybe not for all productions.

Originally Posted by Plush

I like immersive sound too but the Germans who are the leaders in immersive sound and planar sound environments are best left to another discussion.

I would argue that sound is an immersive experience, and all forms of surround sound are even more so. Maybe you're right that discussing the German approach is a bit narrow, but I think the binaural approach lends itself greatly to immersion applications. Most demonstrations of the technique currently showboat binaural's ability to create a space. So, I think that discussing ways to develop its strengths is a very constructive avenue!

Originally Posted by Plush

Binaural is what it is and one either experiments with it and can embrace it or is less enthused by it. It's success is hardly judged with a criterion of making money in the record bidniz.

Sure it is! Who are you kidding? Commercial success and audience appreciation are two great indicators of a format being universally appreciated and useful for a variety of projects. No one needs to cater to binaural purists. They are already enjoying it! Finding ways to make it relevant and useful as a product for producers and consumers will mean more binaural for everyone.

Originally Posted by Plush

It's not really that expensive to record or release binaural recordings. After all, one can buy the best binaural device, the Neumann dummy head, for $8000 and then record every concert with it along side the regular recording set-up. Then when a good result is obtained with the binaural set-up, one releases it along side the stereo or surround version. It can piggy back on the regular stereo release.

The cost to the recordist may be small. When the binaural recording service is pitched to the client and they don't know what it is, they may be unlikely to pay for it, no matter how small the cost is. With recordings making hardly any money at all these days, it is likely a tough sell.

Originally Posted by Plush

I clearly disagree that it is an effect. An effect requires creation of the effect. Binaural playback virtues range from subtle to dramatic. In one recording for example, the listener may have an increased sense of place ( a certain seat) in the concert hall. More dramatic renderings might present the picture of the orchestra with a chorus on raised platforms as having the chorus be clearly heard as being higher up than at stage level.

Natural physical effects exist. Yes, you can record them, emulate them to a degree (via recorded binaural impulse convolution) and reproduce them (via headphones.) I don't have any problem calling it an effect.

Originally Posted by Plush

Analogy is the weakest form of argument:
3D photography is a manipulation of images whereas binaural recording is a microphone placement that does its thing on its own. It does it automatically because of the design of the system.

Analogies are a great way to illustrate a point or argument! You're incorrect, Plush. 3D Photography doesn't manipulate images. It takes two 2D images at a set spacing and presents one to each eye. Same with binaural audio. It records two audio signals positioned around an acoustical barrier modeled after a human head. It then plays back one signal for each ear! I couldn't think of a more clear analogy to illustrate my point. I don't see how someone with your understanding of the medium could think this is invalid.

You don't have to lecture me on how binaural works or shoot me underhanded comments. Even if you considered my first response reason enough to lodge a forum complaint and get it removed, let me say in more civil terms that your post was not constructive in its criticism and also did not address the main points of the original poster which was what I was trying to do.

Originally Posted by Plush

I am a practitioner not a theorist.

Why make a full critique of my theoretical ideas? The OP intended to provoke some thoughts! Why is it your goal to shoot them all down?

[bcgood]

Here's a binarual recording I made outside of an airplane flying overhead near a creek with birds singing in the background.

[Plush]

Hello R. Scott,

Thank you for your full post. I DO value back and forth discussion and I DO read and listen to what posters are adding to the dialog. I don't skim the post and do a knee-jerk reaction posting. May I say that because I value discussion I did not report your or anyone else's post to a moderator here.

In my posting I was not responding to the OP's posts, I was commenting on your own posts in posting #'s 71 and 77.

There you proposed that binaural needs to "ditch the niche" and that hearing binaural on headphones was too limiting a factor for wide audience acceptance.

That thought was what caught my attention.

Of course I will object to this theoretical musing since you call for ruining the
binaural experience in your post. (because you urge abandoning headphones)

I don't agree that binaural is judged on a traditional "sales volume" basis and I certainly do not buy your assertion that most record releases make "hardly any money these days."

My records make all involved a lot of money and that includes binaural recordings.

So my objections to your posts go right to the very core of the language you used. My thoughts are different than opinions. I attempt to make informed commentary from working with binaural for many decades.

[RScott]

I never meant to suggest not using headphones. I was suggesting to move from 'only for headphones' to something like 'enhanced for headphones' or 'best on headphones'.

Your practice of including binaural recordings as a supliment to more traditional recordings is also a great idea along these lines. While most people would not jump at a release biled as headphones only, they may really enjoy a special bonus track in the binaural method.

I guess my whole point is that binaural may, for now, need something to "open the door" to audiences who may very well find out that they love the experience.

Sent from my SPH-D700 using Gearslutz.com App

[klaukholm]

Originally Posted by Mark A. Jay

Emil: AT first, I merely glanced at the picture of the orchestra. However, I have to say that of the orchestras that I have recorded, not one of them had the risers that are present in that photo. This is very interesting as what I am used to hearing here in the states is the orchestra assembled on a flat stage - there is no change in elevation, at least not for the orchestras I have listened to or recorded.

Very, very interesting.

This may sound foolish, but is this typical for the orchestra there to practice as well as perform situated in this manner?

The vast majority of professional orchestras in the US as well as europe use various forms of riser systems. It is a better visual for the audience, it projects the back of the orchestra better, it serves as a reflector for the string sound and it allows everyone in the back to see and hear better.

Emil, I may have seen you in the old radio concerthall about a year ago. I was rehersing with a copenhagen orchestra when a group of student like people were putting up a sphere with dozens on membranes mounted on it. Once in the balcony and once on the floor in front. Were you part of this?