The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Tags:

Audio Reality - An Introduction to Binaural Sound and Space Recorders, Players & Tape Machines
Old 27th January 2017
  #1
Gear Maniac
 

Thread Starter
Audio Reality - An Introduction to Binaural Sound and Space

Hey all,

I have been gathering impulse responses throughout Europe, Asia and Canada, for my continuing research into "directional binaural impulse response reverb." I published and presented a paper on this last year at AES in Paris, and feel the initial results are very promising.

I just uploaded a video demo'ing some of the results, and thought you may find it interesting.

https://youtu.be/onuVjhZ_MWs

Matthew
Old 28th January 2017
  #2
Lives for gear
 
jnorman's Avatar
very fun, matthew - thanks.
Old 28th January 2017
  #3
Lives for gear
Despite Matthew's contention that 85% of listeners use headphones/earbuds (which I can't verify, but am sure the proportion has been rising and will continue to grow), so little popular, classical or other genres are recorded using the KU100 or similar methods.

There are some CD's available with dual layers (eg SACD plus conventional stereo 16/44.1) but again in a minority...there would seem to be a market opportunity to record and release binaural and conventionally recorded layers on a single disc, to give listeners 2 perspectives of the same material ?

The major labels would find a multitude of reasons (mainly production cost related) to NOT do this, but from an editing standpoint I'm guessing it would not be insurmountable. Maybe it's already being done...does anyone know of examples ?

I'm referring of course to the (ever-shrinking) CD market..whereas download delivery could explode this resistance in an instant.
Old 28th January 2017
  #4
Lives for gear
 

I'm glad I watched and listened to this Youtube video!
Old 28th January 2017
  #5
Quote:
Originally Posted by mlien View Post
Hey all,

I have been gathering impulse responses throughout Europe, Asia and Canada, for my continuing research into "directional binaural impulse response reverb." I published and presented a paper on this last year at AES in Paris, and feel the initial results are very promising.

I just uploaded a video demo'ing some of the results, and thought you may find it interesting.

https://youtu.be/onuVjhZ_MWs

Matthew
I enjoyed the video, and am off to the AES library to read your paper.
Old 28th January 2017
  #6
Lives for gear
 

Quote:
Originally Posted by studer58 View Post
Despite Matthew's contention that 85% of listeners use headphones/earbuds (which I can't verify, but am sure the proportion has been rising and will continue to grow), so little popular, classical or other genres are recorded using the KU100 or similar methods.
Yes, headphones are essential for mobile sound which probably accounts for that huge percentage, but I prefer loudspeakers for serious music listening.

Liked the video very much. Nice use of the that wonderful FritzHead microphone and convolution reverb.
Old 28th January 2017
  #7
Lives for gear
Quote:
Originally Posted by David Spearritt View Post
Yes, headphones are essential for mobile sound which probably accounts for that huge percentage, but I prefer loudspeakers for serious music listening.

Liked the video very much. Nice use of the that wonderful FritzHead microphone and convolution reverb.
Do you think that the perceived current binary split between enjoying on headphones (esp. binaural ) and enjoying on speakers, and material being unable to serve both replay methods equally well is destined to remain ?

Plush claims to use the KU100 quite often, and not in the context of binaural-only releases ?
Old 28th January 2017
  #8
Gear Nut
 
Peter Allison's Avatar
the video certainly "opened my eyes" as to the reasons a lot of people, both here and off the board, have been "banging" on about. I am going to give it a try, just have to source an alternative microphone, as the dummy head you use is out of my league, as I just record organs in English cathedrals and smaller churches , York Minster been close to me. Any suggestion will be gratefully received. recorder is a Tascam DR 100 mk 3
Old 28th January 2017
  #9
Leaving the issues with CD and other physical media format behind, wouldn't it be nice to be able to choose (perhaps automatically when the playback system detects headphones) between a traditional stereo version for loudspeakers and a binaural version for headphones? Two additional channels of audio, even if both were at 24/96K, is no burden for current networks and storage.

Echoing Studer58s's comments; The issue for the producer and engineer, I think, is how to create the two versions in the most efficient manner. Shall we in$i$t on multiple head$ for spot miking, or can we use a head for the main pair, with "regular" spot mics and processing to blend them in to a pleasing binaural image?

May I ask what software or process you used for your binaural reverb, after you'd recorded your impulses?
Old 28th January 2017
  #10
Great video! To place sounds in the stereo field using these impulses, I'm assuming you'd want to use a binaural stereo encoder for your source first, right? Otherwise you'd be pasting level-based stereo onto binaural stereo. (That's assuming we don't have access to a binaural head, like you do.)
Old 29th January 2017
  #11
Gear Maniac
 

Thread Starter
Quote:
Originally Posted by studer58 View Post
Despite Matthew's contention that 85% of listeners use headphones/earbuds (which I can't verify, but am sure the proportion has been rising and will continue to grow), so little popular, classical or other genres are recorded using the KU100 or similar methods.
The 85% stat is from AES 2016 in Paris where my paper was published and presented. I was surprised at first, but then not when looking around at the new music consumers everywhere, streaming their music via their phones.

I believe one reason binaural production has remained rare despite this prominence (or should I say dominance) of earphones, is the perceived inability to multitrack and process tracks individually where reverb is concerned (which I am working to resolve). I think this perception has held binaural hostage of the live recording genre where it's often too realistic in the end (even for classical... many spot mics, and thus control over the hall and added reverb yields a better, super-realistic result). Beyond reverb, some other challenges remain, primarily being panning which (in a "native binaural" production) needs to be determined at the time of tracking... but I actually appreciate this need to thoroughly pre-contemplate a project before hitting record.

Quote:
Originally Posted by studer58 View Post
There are some CD's available with dual layers (eg SACD plus conventional stereo 16/44.1) but again in a minority...there would seem to be a market opportunity to record and release binaural and conventionally recorded layers on a single disc, to give listeners 2 perspectives of the same material ?

The major labels would find a multitude of reasons (mainly production cost related) to NOT do this, but from an editing standpoint I'm guessing it would not be insurmountable. Maybe it's already being done...does anyone know of examples ?
Well, funny you should ask. Not content to capture binaural IR's only, I have just completed the first in-studio experiment to track two songs, one upbeat, and one mellow, entirely in native binaural and implementing my binaural IR reverb technique.

The KU-100 was used for every instrument, so panning positions (and movement) needed to be determined in advance.

In the case of the drums (until now, a totally elusive beast for binaural recording unless you like the sound of a drum kit in your room), this required three stereo passes -- one for kick/snare/hihat; one for toms which were extended by adding RotoToms to the rack for a wider binaural image; and one for cymbals which were also added to, for more tonal and imaging variety. This approach allowed us to re-position the KU-100 for optimal imaging, tonality and balance. But this also meant the drummer has to leave holes where tom fills and cymbal crashes or rides would occur. Very tricky, but I think very successful.

I did spot-mic certain instruments as well, but this was to send isolated sounds to binaural reverbs when two or more instruments shared a pass. For example, after recording a french horn and trumpet simultaneously at 9:00 and 3:00 panning positions, I could then send their spot-mic signals to the impulse responses originally captured at the same positions.

Regarding a dual-format production, I did do this when tracking two other songs for a typical stereo album in those same sessions... meaning, I also put the KU-100 into position so I would have an option later to produce binaural versions. But I must say, for the two complex binaural songs I produced (links below), following a philosophical discussion with my engineer about how we handle this project and whether we should also check the mix on near fields, we decided in the end to mix ONLY for headphones and for the binaural result. There needed to be a do-or-die approach to binaural producing on this scale, without compromise. Remember 20 years ago when 85% of listeners used stereo speakers? Did we sweat the headphone mix? Not really... we only used it as a tool to confirm the stereo speaker mix. We didn't set out to ensure the headphone crowd would be pleased... and I felt we likewise needed to commit to headphones in order to prove or disprove binaural's potential in mainstream music production for earphones.

Please have a listen to the preliminary results, and let me know what you think. (These are still in need of a bit of tweaking, but they're close.)

Angel at the Station
https://soundcloud.com/matthewlien/angel

Hallelujah
https://soundcloud.com/matthewlien/hallelujah

I plan to record 8 more tracks and release this as an album in mid-2017.

Quote:
Originally Posted by studer58 View Post
I'm referring of course to the (ever-shrinking) CD market..whereas download delivery could explode this resistance in an instant.
Agree... this binaural album is most-likely meant for download delivery (and possibly via an exclusive platform so a realistic return on investment can be pursued... streaming kills). I guess most people that would still purchase a disc would likely not be listening to that disc on headphones.
Old 29th January 2017
  #12
Gear Maniac
 

Thread Starter
Quote:
Originally Posted by Peter Allison View Post
the video certainly "opened my eyes" as to the reasons a lot of people, both here and off the board, have been "banging" on about. I am going to give it a try, just have to source an alternative microphone, as the dummy head you use is out of my league, as I just record organs in English cathedrals and smaller churches , York Minster been close to me. Any suggestion will be gratefully received. recorder is a Tascam DR 100 mk 3
I think you'll be pleased, Peter. But I don't know of any binaural microphone which can achieve such natural and pleasing results as the KU-100.

I did try putting a pair of DPA 4060 omni mini-mics into a DIY head with realistic silicone ears, and you could certainly pursue that for fun. I've heard of people using lesser expensive Countryman mics and others. The mics just need to be omni-directional mini (lavaliere-style) mics, placed into a pair of silicone ears (so the screen is facing out, and is flush with the inner base of the concha), with the ears attached to a mannequin head. Lots of DIY info on the internet.

Your recorder should be fine. I use the Tascam DR-60D MkII for field recording when I am not really sweating the perfection of the recording too much, or when it's too dangerous to bring my primary field recorder.

You might like these two organ videos I made while capturing impulse responses:

https://youtu.be/zBg4IBg81V0

https://youtu.be/O3KA8ewt4OM

Cheers and good luck!
Old 29th January 2017
  #13
Gear Maniac
 

Thread Starter
Quote:
Originally Posted by TMetzinger View Post
May I ask what software or process you used for your binaural reverb, after you'd recorded your impulses?
I am using Altiverb. I use the sweep tones they provide for a specified reverb time and sample/bit rate, then playback and record the tone using Nuendo, and finally import the impulse responses into Altiverb again to create the binaural reverb in the studio.
Old 29th January 2017
  #14
Gear Maniac
 

Thread Starter
Quote:
Originally Posted by daniel@mineral View Post
Great video! To place sounds in the stereo field using these impulses, I'm assuming you'd want to use a binaural stereo encoder for your source first, right? Otherwise you'd be pasting level-based stereo onto binaural stereo. (That's assuming we don't have access to a binaural head, like you do.)
You are right, binaural reverb is meant to be applied to binaural stereo tracks... although I find binaural reverb "opens up" typical stereo tracks when heard on headphones... so there is a benefit to binaural reverb even for non-binaural production if you believe most people will listen on earphones.

At this time, I am focused entirely on native binaural using the KU-100. But I am experimenting with simulated binaural imaging. The best binaural simulator I know of is SPAT from Flux / IRCAM. I am currently beta-testing the new version, called Spat Revolution, which is pretty amazing! I think it will be a great approach to combine native with simulated binaural in the near future.

One area I am also exploring, is the creating of sample library instruments in binaural, for example, an ancient Chinese stone xylophone where every stone is sampled, one by one, around the head (and therefore, the playback sounds like that!). Such VST instruments, blended with native and sim binaural, is going to be a blast!
Old 29th January 2017
  #15
Lives for gear
 
Boogaju's Avatar
 

Great stuff, very nive video
Old 1st February 2017
  #16
Lives for gear
 
Plush's Avatar
When I am working with the Neumann head, the stone and wood church is my friend. Placing instrumentalists and singers around the head is the same as other 2 mic stereo recordings. Distance and height of the playas are changed to get the balance.

Currently I am pursuing setting up a binaural streaming channel in cooperation with a big national classical broadcaster. It will offer music and radio plays on a stream so that the listener may put on headphones to hear it correctly.

Binaural is recorded with a modified KU-100 head and also derived from A format and B format Ambisonics mics using special software.

At least for music, working in a reverberant space gets rid of the need for artificial reverb. I do believe however that artificial binaural reverbs are a quite interesting area of research and enhancement
Old 1st February 2017
  #17
Lives for gear
 

The Unanswered Question by Charles Ives seems to be a perfect candidate for a KU-100 binaural recording if it has not already been recorded this way. Anyways, I googled the topic and found this: American Mavericks: Charles Ives and Spatial Music

I think it is relevant to mlien's comment of:
"I believe one reason binaural production has remained rare despite this prominence (or should I say dominance) of earphones, is the perceived inability to multitrack and process tracks individually where reverb is concerned (which I am working to resolve). I think this perception has held binaural hostage of the live recording genre where it's often too realistic in the end (even for classical... many spot mics, and thus control over the hall and added reverb yields a better, super-realistic result). Beyond reverb, some other challenges remain, primarily being panning which (in a "native binaural" production) needs to be determined at the time of tracking... but I actually appreciate this need to thoroughly pre-contemplate a project before hitting record."
Old 3rd February 2017
  #18
Gear Addict
Matthew - I have been following your binaural experiments off and on via YouTube with great interest. Like you, I believe that it is worthwhile to explore this style of recording, as I think the majority of young audiophiles are turning to high end head fi hear. You have only to observe the activity in discussion forums. Those catering to large systems with monoblock amps and large floorstanding speakers are filled with older people whose numbers are dwindling. By contrast, the head fi sites are vibrant and booming with a new generation of audiophiles. Head-fi.org gets about 50k unique visitors per day. And if you spend any time there, you'll know that binaural source material is a perennial topic of interest.

David Chesky has figured this out and has fully committed to the format. I have no idea what kind of sales they have, but they do seem to keep producing them, and at an ever more rapid pace.

The reality is that people who are just now entering their peak earning years grew up on white earbuds. Listening to music on the go is an integral part of their behavior pattern. As a group, they also live in smaller spaces, and often in closer proximity to neighbors in high density communities. So, a large and powerful hi fi system is often out of the question. And the dominant political views in this group see a large and expensive hi fi system as ostentatious and conspicuous consumption. That is, of course, unless the system is "vintage", in which case it's acceptable.

To me, all this adds up to a growing potential demand for recordings that are made specifically for head fi. I think you're on the right track.

Last edited by bwanajim; 3rd February 2017 at 10:32 PM..
Old 7th February 2017
  #19
Gear Maniac
 

Thread Starter
Quote:
Originally Posted by bwanajim View Post
... I think you're on the right track.
Thanks for the response and confidence. Once the new binaural album using these techniques is complete, I'll need to engage forums focused on binaural to get the word out.

Here's a couple tracks nearly complete.

https://soundcloud.com/matthewlien/angel

https://soundcloud.com/matthewlien/hallelujah

Another fun little YouTube video is going up in a couple days, a nearly a cappella version of Down in the River.
Old 2nd March 2017
  #20
Lives for gear
 
jnorman's Avatar
The 3dio pro semi-head uses DPA 4060 capsules and is about $2000. I haven't used it but it looks interesting.
Old 2nd March 2017
  #21
Gear Maniac
 

Thread Starter
Quote:
Originally Posted by jnorman View Post
The 3dio pro semi-head uses DPA 4060 capsules and is about $2000. I haven't used it but it looks interesting.
Before purchasing the KU-100, I experimented with mannequin heads, anatomically accurate silicone ears, and the DPA 4060 microphones you mention. If you wanted to go this route, I would recommend just buying the DPA microphones and silicone ears, and experimenting yourself. The results aren't bad and it's lots of fun.
Old 2nd March 2017
  #22
Gear Addict
Can't help wondering, for loudspeaker playback of a binaural/dummyhead recording, do people us a shuffler circuit/device as AlanBlumlein proposed in his patent?
Old 2nd March 2017
  #23
Lives for gear
 
James Lehmann's Avatar
 

A wonderful video Matthew, and great work you are doing!

I'm listening to the stuff on your SoundCloud now.

As a side note, if you or anyone else reading this would like to experience a piece of 'Binaural Theatre' go see The Encounter by Complicite; in this performance the audience are all given headphones and the solo performer (Simon McBurney) roams around a stage with a KU100 as the centrepiece. I saw it in Amsterdam and was blown away - a really interesting and I think unique use of binaural technology.
Old 10th April 2017
  #24
Gear Head
 

Quote:
Originally Posted by Peter Allison View Post
I am going to give it a try, just have to source an alternative microphone, as the dummy head you use is out of my league, as I just record organs in English cathedrals and smaller churches , York Minster been close to me. Any suggestion will be gratefully received. recorder is a Tascam DR 100 mk 3
If you are the handy type, DIY binaural microphones are your best option.
In general, I believe that properly built DIY microphones with a proper recording process (there are EQ considerations to take into account when using DIY binaural recordings, that I suppose the KU 100 takes care of with internal circuitry) can achieve as good as results, if not better, than professional microphones.
Plus, to your point, it can be done for less than 1/10th of the price.

I actually created an improvement for DIY microphones.

https://www.shapeways.com/shops/bina...icrophone-ears

The concept is to improve the HRTFs encoded into each channel, to recreate a more realistic sound stage during playback. This is done with the use of an anatomically accurate ear canal replica, rather than the straight tubes that most microphones use (including the KU 100).
Old 10th April 2017
  #25
Lives for gear
 
Plush's Avatar
I would rather use Soundman in-ear mics than a home made set up.
Why reinvent the wheel, baby? I do not trust home made pinnae measurements.

The reason is that they sit in real ears (your ears) and pick up with excellent fidelity.

Then, on playback, you hear the spatial cues which work the best for you since you are listening back with your own ears.

Soundman is made in Germany and is recommended.
Old 10th April 2017
  #26
Gear Head
 

Quote:
Originally Posted by Plush View Post
I would rather use Soundman in-ear mics than a home made set up.
Why reinvent the wheel, baby? I do not trust home made pinnae measurements.
That's fine, but it's actually non proper binaural recordings you make with in-ear mics.
If one prefers Soundman to, for example, Neumann KU 100, then my ear models are not for them.
However, if one would like to get same or better proper binaural recording quality than with a KU 100, for much less money, my ears aim at getting better HRTFs encoded into the two channels.

Quote:
Originally Posted by Plush View Post
The reason is that they sit in real ears (your ears) and pick up with excellent fidelity.

Then, on playback, you hear the spatial cues which work the best for you since you are listening back with your own ears.

Soundman is made in Germany and is recommended.
I followed a different philosophy, although I see your point.
My aim was to create a product able to let you make binaural recordings that would work for the majority of the people.
So I opted to use an ear canal that is averaged over hundreds of real canal measurements.
Old 10th April 2017
  #27
Lives for gear
 
Plush's Avatar
DIY'ers will have a very big challenge fitting ANY microphones that are as good as the Neumann mics in the KU100. It does not matter where the maker situates the mics within the ear canal.

So making a head with your ears is a huge challenge. Really an endless pursuit.

Also the head must be of the proper human density and weight. Ideally you would want a torso that can wear a sweater and a tie!

My point is that you can indeed make great binaural recordings that are correct and proper for the LISTENER with an inexpensive engineered system like Soundman.

One's ears are already correct. They are not a model.
Old 10th April 2017
  #28
Gear Head
 

I'd like to comment on one point that is fundamental to binaural recording technique.
It DOES matter where you put the microphone.
A microphone located at the opening of the canal, for example, like many do with their DIY rigs (and professional microphones), actively changes the sound waveform reaching the capsule. The reason is because putting a solid boundary (the capsule) where there would be an opening (the canal) forces the velocity of the sound wave to be identically 0 where it shouldn't be. In other words, the capsule changes the HRTF by the mere fact of being there.

Putting the capsule at the eardum's location is fundamental to encode proper HRTFs on the recording.
Capsules up to 6 mm can be used with my models.
While those capsules are usually not that quiet when used in traditional ways, one huge indirect benefit of binaural microphones is that there is an embedded 15-20 dB A-weighted S/N increase, which makes it possible to use noisier capsules.
4mm Primo capsules would be my suggestion.
High SPL handling is required, given the added 15-20 dB boost applied by the HRTF (which is the cause of the S/N increase).

Making a DIY binaural microphone, with or without my ear models, is not as 'plug-and-play' as using professional microphones. There is a mandatory EQ to apply in the mixing phase before the recording can be played back, for one.
But the savings make up for the effort of a little DIY work, in my opinion. Someone might even have fun doing it!

As far as the microphone material, it would be great if it were made with something closer to human flesh. However, the shape of the microphone is far more important than the material.
Old 10th April 2017
  #29
Lives for gear
It's quite laughable really that either of you, Plush or sax, claim to have definitive knowledge of the inner workings of the ear/brain interface...or even to suggest that the eardrum is somehow analogous with a condensor mic element.

The total hearing mechanism, not to mention the inner ear and the associated neuronal transfer processes, are far more complex than simply the shape of the pinnae, or even the dimensions of the hearing canal. For a start, there is transconductance and impedance matching between the tympanic membrane and the cochlear fluid...so that leaves a conventional mic diaphragm at the starting gate.

Frequency selective hair trigger cells, stapedius, round and oval windows...where's the equivalent in a microphone capsule ?

Do some research about the workings of the ear, evolved over millennia, and a complete integrated system...and perhaps you'll put aside childish comparisons with a crude electro mechanical device we've had for less than 100 years.

A simplified jumping off point for you both: https://en.wikipedia.org/wiki/Hearing
Old 10th April 2017
  #30
Gear Head
 

Quote:
Originally Posted by studer58 View Post
It's quite laughable really that either of you, Plush or sax, claim to have definitive knowledge of the inner workings of the ear/brain interface...or even to suggest that the eardrum is somehow analogous with a condensor mic element.

The total hearing mechanism, not to mention the inner ear and the associated neuronal transfer processes, are far more complex than simply the shape of the pinnae, or even the dimensions of the hearing canal. For a start, there is transconductance and impedance matching between the tympanic membrane and the cochlear fluid...so that leaves a conventional mic diaphragm at the starting gate.

Frequency selective hair trigger cells, stapedius, round and oval windows...where's the equivalent in a microphone capsule ?

Do some research about the workings of the ear, evolved over millennia, and a complete integrated system...and perhaps you'll put aside childish comparisons with a crude electro mechanical device we've had for less than 100 years.

A simplified jumping off point for you both: https://en.wikipedia.org/wiki/Hearing
One can only approximate nature as best as they can.
All that you are saying is absolutely correct, but should we stop trying to improve things just because the microphone capsule is not made out of eardrum tissue? I don't think so.

I never claimed to have definitive knowledge of anything. With my models I just humbly tried to walk another step forward in the constant and never ending quest for a truly accurate audio recording.
Peace!
New Reply Submit Thread to Facebook Facebook  Submit Thread to Twitter Twitter  Submit Thread to LinkedIn LinkedIn  Submit Thread to Google+ Google+  Submit Thread to Reddit Reddit 
 
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump