Best field binaural rig

[rav]

I´m going to purchase a handheld recorder and a set of binaural mics

I´ve already have some mics for field recording to use whith it.

I want it to record sound atmospheres fx.

Here my question goes.

Which handheld recorder you prefer microtrack II, zoom H4 or marrantz pmd620 or....?

Which set of in ear binaural microphones that translates well to speakers?

okm? Wich ones for sound atmospheres recording?

Many questions.

Thanks in advance.

Best.

[aaron_in_sf]

My own requirements might be different than yours, but I do discuss this very topic somewhat exhaustively on my site's DIY section:

quiet american

The short answer is, for ease of use and good translation for conventional playback (e.g. non-true-binaural) I myself prefer Sonic Studios DSM-6S/EHs mounted in the WHB headband/windscreen (with double socks for true wind).

They're electrets powered by plug-in power so you will need an adapter or outboard powering module if you use them with a deck that does not provide it. Sound Professionals make an inline adapter that will provide plug-in power off a 48V (or 30V) phantom supply, but I haven't used it myself.

They're not the quietest -- their noise floor makes them not a good choice for truly quiet ambiances -- but in all other respects I love them. Hours of examples on my site, almost entirely untouched. See the piece 'Annapurna' in 'related projects' for a good sense of their range.

The Core Sound HEBs based on matched DPA 4060 capsules come with their own powering module; they are slightly quieter but more colored than the DSMs (they have a bump for the DPA's intended speech reinforcement application). Out of the box they do not offer any obvious way to mount them for true or near-binaural use; they are small enough to tuck in your ear canals if you're gonzo -- that's why I got them -- but that's not a long-term field use solution unless you do some engineering for a comfortable mount. If you intend to mount them in an artificial head that's not as much of an issue, of course. (Speaking of, Sonic Studios sells a portable head for use with their mics...).

On the portable recorder front there are a host of handy small boxes these days, there are those on this forum that prefer the Korg DSD recorders. In the Naturerecordists list circle that I'm in the Fostex FR-2LE is quite popular. You might look at the new Marantz PMD620 and Sony PCM-D50 as well. If you can afford one I have had good results using my DSMs with my Sound Devices 722.

On the cheap end, I should say I do really like my Zoom H2 -- as long as it's fed through its line input. The plug-in power providing external mic in is not recommended. Btw its own internal mics are quite good enough for scratch recording.

best,
aaron

[ambo]

No in ear mics will translate well to speakers. If you want to do both I'd suggest mounting the mics in a hat or on glasses slightly forward of your ears. It's not true binaural but sounds good on speakers and headphones. I just bought a Microphone Madness set of Sennheiser in ear omni's but they haven't arrived yet. For small omni's I'm using Church Audio mics with a Church preamp. Very nice and affordable. Oh yes and i use an Edirol R09 recorder.

[jnorman]

go over to taperssection.com forums - there are many threads about this topic, and lots of good info.

[boojum]

Binaural is for headphones.

I have used the OKM II's and am happy with them. Aaron liked a track of mine, too, and published it. He lalso kindly sent me a copy of the CD it was included in. I use the OKM II's with their battery box and run it to an MD recorder, currently a SONY MZ RH1. It records in WAV format and is rugged. It has a great set of pre-amps. YMMV MD is a dying format, but I have plenty of blank MD's. ;o)

[rav]

Thanks everybody,

I have already purchased microtrack II and okm II Klassic Studio and waiting for delivering.

I hope i didn´t make a mistake

Best.

[MBishopSFX]

The Neumann KU-100 binaural head (aka "Fritz") does indeed translate very well to speaker playback.

[bcgood]

Originally Posted by MBishopSFX

The Neumann KU-100 binaural head (aka "Fritz") does indeed translate very well to speaker playback.

I think it would be cool to record an entire album with fritz. Do a kind of Cowboy Junkies thing. Record in a really nice space and use baffles to precisely control the sound of the acoustic space and reverberation. Record it in a way that it would sound good on speakers and mind blowing awesome on headphones.

[Remoteness]

FRITZ!!!

[MBishopSFX]

Originally Posted by bcgood

I think it would be cool to record an entire album with fritz. Do a kind of Cowboy Junkies thing. Record in a really nice space and use baffles to precisely control the sound of the acoustic space and reverberation. Record it in a way that it would sound good on speakers and mind blowing awesome on headphones.

Here's a recording made just as described above:

ftp://telmedia.telarc.com/telarc/mp3/80536-320.mp3

[boojum]

To me the Telarc track sounds straight stereo in that the sounds are all "inside" my head. A visit to quietamerican.com allows tracks that are "inside" and "outside" my head. I thought that binaural always had the impression that the events were not just inside our heads but all around us. Is this because Fritz is set far enough back that all the sound events are in front? That could be an explanation. Does anyone know the distance back?

I wonder, too, how Fritz sounds alongside a Jecklin disc? Are there available tracks comparing that?? This is an area I am very much interested in and would appreciate any help here possbile.

Thanks

[aracu]

Although a microphone that looks like a manniken's head seems bizarre, that recording gives an extremely realistic sense of relative near/far and horizontal distances. The sound is impressively warm and well balanced.

[bcgood]

Originally Posted by MBishopSFX

Here's a recording made just as described above:

ftp://telmedia.telarc.com/telarc/mp3/80536-320.mp3

I just got through listening to that on my studio headphones. That sounded awesome! Thanks for posting that. I wonder why more recordings aren't done this way, that sounded phenomenal.

As good as it sounded check out this bcgood enhance "O" fest 2008 version...

80536-320 II.mp3

[mikesilence]

Thank you so much for the audio file... thumbsupthumbsupthumbsup

Unbelievable good, I like it. Is there an album that can be bought??

[Dave C]

Originally Posted by MBishopSFX

Here's a recording made just as described above:

ftp://telmedia.telarc.com/telarc/mp3/80536-320.mp3

It should be noted that the lossy psychoacoustic model used for mp3 and minidisc atrac encoding will most likely discard the psychoacoustic spatial information that is gained by true binaural recording.

Whereas on the original recording (assuming full bandwidth PCM or better) the horizontal, vertical, and (to a lesser extent) front to back directional and distance cues are 'encoded' by the binaural recording process, the lossy compression assumes that this information is not needed and dumps it.

True spatially encoded binaural needs non-lossy recording and reproduction.


Regards,

Dave C.

[aaron_in_sf]

Originally Posted by Dave C

It should be noted that the lossy psychoacoustic model used for mp3 and minidisc atrac encoding will most likely discard the psychoacoustic spatial information that is gained by true binaural recording.

I agree that this should, in theory, be the case...

...but it might bear mentioning that my experience in practice has been otherwise.

I myself have made many recordings using near- and true-binaural recordings direct to minidisc using ATRAC, and also uncompressed recordings later compressed to high-bitrate MP3.

In both cases, I would say that to the resolution of most peoples' playback situation, almost all of the binaural properties are preserved. Or perhaps a better way of saying it is, by the time those things are noticeably compromised, it's the case that other aspects of the recording have been too. So my $.02 would be, don't worry about that too much. (Of course, I don't claim to have particularly golden ears...)

Generally speaking front-back discrimination is always the Achilles' heel, incidentally, in my recordings, even when recorded to 24/96 on my SD722... :/

best,
aaron

[Dave C]

Originally Posted by aaron_in_sf

I agree that this should, in theory, be the case...

...but it might bear mentioning that my experience in practice has been otherwise.

I myself have made many recordings using near- and true-binaural recordings direct to minidisc using ATRAC, and also uncompressed recordings later compressed to high-bitrate MP3.

In both cases, I would say that to the resolution of most peoples' playback situation, almost all of the binaural properties are preserved. Or perhaps a better way of saying it is, by the time those things are noticeably compromised, it's the case that other aspects of the recording have been too. So my $.02 would be, don't worry about that too much. (Of course, I don't claim to have particularly golden ears...)

Generally speaking front-back discrimination is always the Achilles' heel, incidentally, in my recordings, even when recorded to 24/96 on my SD722... :/

best,
aaron

Agreed on front to back issue. What binaural mic system are you using for your tests?

By True Binaural I mean utilising at least a dummy head with pinnae and embedded mics in the dummy's ear canal. Head and torso system would be even better. The discrete reflections from the curves of the pinnae and the shading effects and shape of the head (Head Related Transfer Function (HRTF), and the fact that the mics are in the ear canal are what cause true binaural recordings to reproduce directional and spatial cues (reasonably) accurately, when listened to on headphones. I don't just mean a reasonable stereo image, I mean the ability to hear the exact direction in 3D space and a perception of the distance of the source in that direction.

16/44.1 is perfectly adequate resolution to perceive these spatial cues.
Lossy compression removes the spatial cue artifacts (tiny delays from the pinnae curves, HRTF artifacts at low level) from the signal leaving it as merely a 'stereo' recording. L to R discrimination should not normally be reduced, but the perception of spatial depth and distance will be removed.

I have done the research with a Fritz(KU-100) as the test mic.
The KU-100 to a mic pre, ADC, and straight to PCM and atrac storage. The atrac lossy compression (based on similar psychoacoustic models as MP3) removes these low level psychoacoustic spatial cues. The 3 dimensionally perceived soundfield collapses to something approximating regular stereo when encoded.

If anyone wants to try this and doesn't have a dummmy head mic, there are some CD's available recorded with one which can then be lossily encoded. It is then easy to compare the PCM and atrac/MP3 versions in a DAW and swapping back and forth between versions (A/B'ing) while listening on headphones will illustrate this.

David Tudor's binaural work is a good test source for this type of test.



Regards,

Dave C.

[boojum]

^^^ Dave, interesting and informative. I had not thought that MP3 compression would damage the binaural sound field. OTOH, I have listened to Aaron's files in MP3 and find the spatial cues better than the orchestral piece from the head which was posted above. I do not know why this is, but remarked on it as soon as I listened to it. As usual, YMMV

[MBishopSFX]

Originally Posted by Dave C

It should be noted that the lossy psychoacoustic model used for mp3 and minidisc atrac encoding will most likely discard the psychoacoustic spatial information that is gained by true binaural recording.

Whereas on the original recording (assuming full bandwidth PCM or better) the horizontal, vertical, and (to a lesser extent) front to back directional and distance cues are 'encoded' by the binaural recording process, the lossy compression assumes that this information is not needed and dumps it.

True spatially encoded binaural needs non-lossy recording and reproduction.


Regards,

Dave C.

The posted example is Ensemble Galilei and both their Telarc releases are out on SACD made from DSD source and kept 1-bit all the way to the SACD author. The surround program on their second SACD is made with two KU-100s back-to-back.

There's also a Robert Jr. Lockwood SACD made with the KU-100. I use the KU-100 as the centerpiece of all my orchestral and jazz sessions.

[aaron_in_sf]

Originally Posted by Dave C

Agreed on front to back issue. What binaural mic system are you using for your tests?

By True Binaural I mean utilising at least a dummy head with pinnae and embedded mics in the dummy's ear canal. Head and torso system would be even better.

Hi Dave,

Ah. Yes. Well, by those standards, perhaps the reason I haven't noticed a distracting loss of queues is that they're absent to begin with!

One important disclaimer, most of the material on my site was not made with a true binaural setup, but with Sonic Studios DSM mics typically positioned on my temples, slightly *forward* of my ears, sometimes on my ears. In neither configuration woul the elements be in a position to receive any of the more subtle HRTF queues you describe.

I was thinking specifically myself of my experiments using my DPA 4060-based Core Sound HEBs for 'binaural' by inserting them with care into my own ear canals. My experience with that has been that the spatialization I have gotten with that approach is discernibly improved over that of my trusty DSMs, but at the expense of impaired translation to speakers (still useable, but more lifeless...).

I'm aware with this approach though is not actually getting the 4060's elements into 'proper' position, but also had reflected (bad pun?) that the elements were on the other hand close enough (for gov't work) to the position that the drivers of my Etymotic ER4s earcanalphones would be on playback... which seemed a Good Thing.

Honestly, though, I haven't performed the rigorous A/Bing you describe with either type of recordings -- and now I'm curious. I was going more on my awareness that the compressed versions have always contained enough information for me to not be overtly concerned that I was losing much of their ability to render space.

None of my work is at the level of sophistication you describe, though... it's an avocational love, not my livelhood...

...yet!

best,
aaron

[boojum]

Originally Posted by aaron_in_sf

Hi Dave,

<snip>
I have gotten with that approach is discernibly improved over that of my trusty DSMs, but at the expense of impaired translation to speakers (still useable, but more lifeless...).
<snip>

best,
aaron

It has always been my understanding that "true" binaural does not work at all well on speakers. The simple stuff I have done with OKM II's has not sounded at all good on speakers; lifeless as Aaron describes. But on the earphones I get the feeling of depth and distance I do not get as well with plain stereo.

[uosdwis]

While I agree that lossy compression can be detrimental to spatial cues, especially for binaural recordings, for clarification I'd just like to point out that the MP3/ATRAC psyschoacoustic models don't really 'remove spatial cues' per se. That may well be how the end result is perceived, but the actual encoding process isn't nearly as sophisticated as that.

I, too, am very fond of the concept of binaural recordings. It's the most democratic surround format there is: the effect is can be powerful even on cheap earbuds but naturally scale up with better headphones. I'm very interested in experimenting with binaural ambiences in combination with dry elements and see how far you can take the blend. It's tricky, any non-binaural elements or mixing of different spaces easily break the illusion. Re-recording speaker playback binaurally is one way, I guess.

I'd love a Fritz or two, but they sure are pricey. Good thing you at least don't have to feed him (anything other than P48, that is!).

[Dave C]

Hi All,

Micheal, thanks for the info on the original releases. I imagine they should sound wonderful and very true to life. However, the link was to an mp3, which by the nature of its lossy encoding process does not give the full 'true binaural' listening experience as I have outlined above in a previous post. You're lucky to have a KU-100, they really are great. I only had access to one where I worked for a time, but do have a Soundfield MKV on the way....



Regarding making surround recordings with 2 x KU-100 back to back...
I don't see the sense in this, as to properly appreciate the binaural recording (i.e. the 3d spatial cues) you must listen on headphones. Listening on speakers will not give the effect of aural immersion into the recording space, although the KU-100 does function as a reasonable stereo mic when heard on speakers.
However, putting 2 of them back to back and playing on a surround speaker system will give no more advantage than a good stereo pair of mics back to back. In fact, a directional stereo pair of mics would have some sort of rear rejection so the pair assigned to the front playback speakers would have little rear information in them and vice versa for the rear pair of mics.
The dummy head does not have any rear rejection when used for speaker playback (the spatial cues only work on headphones) and so is effectively 2 omnis seperated by a barrier (the head part) for improved L-R seperation. If we put 2 omni pairs back to back, the second pair is more or less redundant, as they have no (or very little; depends on frequency I suppose) rear rejection. So I can see no point in putting 2 KU-100 back to back for surround recording. Maybe a single binaural head mic could be processed through an inverse HRTF model (or de-convolved, whatever...) to give single point loudspeaker feeds?



Aaron, I would imagine that the DPAs inserted at the opening of your ear canal should record more or less true binaural. You will be getting the HRTF and pinnae reflections. Listening to those recordings on headphones should give you a 3d listening experience, provided they are not lossy recordings in the first place. Binaural recordings really are meant for headphone listening, although there is ongoing research into L to R loudspeaker crosstalk removal. I don't think it's quite there yet though...

Uosdwis, you are right in that the MP3/ATRAC process is not actually sophisticated enough to remove the micro delays from the pinnae reflections and de-convolve the HRTFs, but what it does do is remove a lot of low level (i.e. low amplitude) components from the signal. The ones it thinks you can't hear because it assumes they are masked by nearby larger amplitude frequency components. Unfortunately, the parts of the signal that make up the spatial cues are generally low amplitude components of the signal, and as such they get discarded by the MP3/ATRAC algorithms.


Best Regards,

Dave C.

[aaron_in_sf]

Originally Posted by Dave C

Aaron, I would imagine that the DPAs inserted at the opening of your ear canal should record more or less true binaural. You will be getting the HRTF and pinnae reflections. Listening to those recordings on headphones should give you a 3d listening experience, provided they are not lossy recordings in the first place. Binaural recordings really are meant for headphone listening, although there is ongoing research into L to R loudspeaker crosstalk removal. I don't think it's quite there yet though...

Oh yes, that tactic definitely gives me what I'm used to thinking of as a "true binaural" experience with an "immersive," natural, three-dimensional soundstage (or rather, "inhabited sound space")... even when compressed .

I was just contemplating that they elements aren't actually in the position I believed dictated for text-book binaural, at the eardrum; I recall reading somewhere that the reflections within the ear canal can provide additional information. In this case those would be munged again by the odd reflections around the mics (and cables) themselves...

best,
aaron

[bobkeiser@ieee.o]

Originally Posted by Dave C

Hi All,

Regarding making surround recordings with 2 x KU-100 back to back...
I don't see the sense in this...So I can see no point in putting 2 KU-100 back to back for surround recording. Maybe a single binaural head mic could be processed through an inverse HRTF model (or de-convolved, whatever...) to give single point loudspeaker feeds?

Best Regards,

Dave C.

Dave,

While you make many points here, I would recommend you listen to this surround recording before you pass judgment. It may not make sense, but it sounds terrific! EG put on a great performance, and Michael & the rest of the Telarc crew did a great job capturing it with this setup. One of my favorites!

[Plush]

Two dummy heads back to back does not give you surround. The reason is that there is not enough difference in placement in the hall. Both mics are omni pick-ups so the rear facing mic is not resolving hall ambience any differently than the front facing mic.

That is, of course, if the mics are truly back to back. (Almost touching)


I agree that the KU-100 is the best binaural mic out there and worth every bit of it's price. It is also a very good stereo mic for playback on loudspeakers.

[Mark A. Jay]

Originally Posted by bcgood

I think it would be cool to record an entire album with fritz. Do a kind of Cowboy Junkies thing. Record in a really nice space and use baffles to precisely control the sound of the acoustic space and reverberation. Record it in a way that it would sound good on speakers and mind blowing awesome on headphones.

BC: I can't recall if you downloaded the tracks or not, but I recently put them back up on my FTP site as some folks had asked about them. So, if you still want those cowboy junkies tracks (recorded at The Ark in Ann Arbor, MI) you can use an FTP client to get there. Host: immersifi.com Username: client1.immersifi.com Password: binaural. The tracks are in a folder called Cowboy_Junkies.

Otherwise, you can use this string: ftp://client1.immersifi.com:binaural@immersifi.com

Feel free to download any or all tracks. Again, beware the the gain shift on track 05 - it was beyond my control.

[dallas simpson]

Originally Posted by uosdwis

I'm very interested in experimenting with binaural ambiences in combination with dry elements and see how far you can take the blend. It's tricky, any non-binaural elements or mixing of different spaces easily break the illusion. Re-recording speaker playback binaurally is one way, I guess.

Multiple binaural through the use of mixes of binaural and 'dry elements' is a simple matter in the electronic mixing environment, but as you suggest considerable care must be taken to preserve a good differential for 'inside' (mono) and 'outside' (binaural) the head sound localisations.

Re-recording speaker replay is an excellent way of creating binaural choreography, but again has its own problems. Room reflections reduce spatial accuracy and size of speaker can affect the spatial definition of the sound source - large speakers are more 'diffuse' in localisation. Using a speaker outdoors can eliminate most of the reflection issues, but brings other challenges. But all in all this is an exciting challenge.

I posted some links on another thread, which may be of interest - here they are:


The headphone only audience format - band playing in a separate room to the audience, binaural audio and live video feeds to the listening space in the form of a 'chill-out' zone - from our binaural Mini Tour last year (2009), that we both collaborated on, is up on YouTube:


YouTube - 1of4 Swimming + Dallas Simpson// There's A Mountain (Binaural Live)// Listen with HEADPHONES ONLY
YouTube - 2of4 Swimming + Dallas Simpson// The Fireflow Trade (Binaural Live)// Listen with HEADPHONES ONLY
YouTube - 3of4 Swimming + Dallas Simpson// Crash the Current (Binaural Live)// Listen with HEADPHONES ONLY
YouTube - 4of4 Swimming + Dallas Simpson// Improv (Binaural Live)// Listen with HEADPHONES ONLY

And here is an environmental binaural music remix:
SWIMMInG// Tigershark Binaural Woods Remix (Headphone Only) on Vimeo

So it is possible to do live binaural headphone concerts. Obviously it could be done with any music genre - jazz, classical, metal, rock...

There is a short showcase of binaural drama with Richard O'Brien in a post of mine on the 'What is your favourite binaural recording?' thread. This uses a mix of mono narrative and binaural soundscape.