Hello all,
I'd like to share with you my personal contribution to the world of DIY binaural microphone recording technique.
If accuracy and realism is what you are after in your recordings, you will find my 3D printed ears very interesting.
https://www.shapeways.com/shops/binaural-microphone-ears
All the best and great recordings to you!
I'd like to share with you my personal contribution to the world of DIY binaural microphone recording technique.
If accuracy and realism is what you are after in your recordings, you will find my 3D printed ears very interesting.
https://www.shapeways.com/shops/binaural-microphone-ears
All the best and great recordings to you!
A little info, from my shop's page:
I started this venture for personal reasons. I am a classical guitar player and wanted to make a recording of myself playing my favorite pieces.
I have always found classical recordings, even from the major labels, to be lacking in realism.
They can sound great and bigger than life when well recorded, but never like a realistic representation of the music.
There is always a sense of artificiality, and that's probably the reason why many people think that stereo systems never quite sound like the real thing.
So I started investigating various recording techniques, and ended up choosing binaural for the best possible results in terms of accuracy.
I believe if you are visiting this page you have come to the same conclusion.
I have worked on my DIY binaural microphone for years, perfecting it in every possible aspect, from head dimensions, to microphone capsule choice, to the most important part of it: the ears.
Why anatomically accurate ear canals?
Evolution provided us with our ears and ear canals to be able to hear our surroundings.
Our brain's constant processing of left and right ear inputs is what makes us hear. These inputs vary depending on the nature of the source and its location in respect to the listener, and are knows as Head Related Transfer Functions (HRTFs).
The ears and the ear canals play a fundamental role in providing our brain with the HRTFs to interpret. They literally shape the sound wave before it hits the eardrum, so that the brain is able to interpret the signals correctly.
Differences between left and right inputs are specifically what makes us able to locate sounds.
Without ears and ear canals, localization and sound recognition are compromised.
Binaural recordings aim at encoding the HRTFs directly on the record, to be able to be retrieved by our brain during playback.
It is essential to use anatomically accurate replicas of the ears as well as the ear canals in a binaural microphone. This ensures that the HRTFs encoded in the left and right microphone signals are the correct ones, for our brain to interpret them and let us enjoy the ultimate in sound accuracy and realism.
I started this venture for personal reasons. I am a classical guitar player and wanted to make a recording of myself playing my favorite pieces.
I have always found classical recordings, even from the major labels, to be lacking in realism.
They can sound great and bigger than life when well recorded, but never like a realistic representation of the music.
There is always a sense of artificiality, and that's probably the reason why many people think that stereo systems never quite sound like the real thing.
So I started investigating various recording techniques, and ended up choosing binaural for the best possible results in terms of accuracy.
I believe if you are visiting this page you have come to the same conclusion.
I have worked on my DIY binaural microphone for years, perfecting it in every possible aspect, from head dimensions, to microphone capsule choice, to the most important part of it: the ears.
Why anatomically accurate ear canals?
Evolution provided us with our ears and ear canals to be able to hear our surroundings.
Our brain's constant processing of left and right ear inputs is what makes us hear. These inputs vary depending on the nature of the source and its location in respect to the listener, and are knows as Head Related Transfer Functions (HRTFs).
The ears and the ear canals play a fundamental role in providing our brain with the HRTFs to interpret. They literally shape the sound wave before it hits the eardrum, so that the brain is able to interpret the signals correctly.
Differences between left and right inputs are specifically what makes us able to locate sounds.
Without ears and ear canals, localization and sound recognition are compromised.
Binaural recordings aim at encoding the HRTFs directly on the record, to be able to be retrieved by our brain during playback.
It is essential to use anatomically accurate replicas of the ears as well as the ear canals in a binaural microphone. This ensures that the HRTFs encoded in the left and right microphone signals are the correct ones, for our brain to interpret them and let us enjoy the ultimate in sound accuracy and realism.
Last edited:
Hi rdf.
Thank you for the kind words. I did put a lot of time and effort into making my ears. I think it paid off, though.
Here is a link to some recordings I made.
https://www.dropbox.com/sh/strkpn3r6ewcy2w/AAAKQI5RfG0Z4l8aMrlvp48la?dl=0
I did end up having to use capsules with higher SPL handling. Panasonic WM-61A capsules start to saturate with quite low SPL material, like piano.
The reason is that the canal hypes the frequencies that our ears are most sensitive to.
This requires a capsule with higher SPL handling, but it also lowers the microphone noise floor signal at those frequencies by the same amount, after processing through the necessary equalization filter.
The increase in S/N ratio is easily 15 dB, given that the peaks of the replica response are in the order of 15-20 dB, at the frequencies humans are most sensitive.
Thank you for the kind words. I did put a lot of time and effort into making my ears. I think it paid off, though.
Here is a link to some recordings I made.
https://www.dropbox.com/sh/strkpn3r6ewcy2w/AAAKQI5RfG0Z4l8aMrlvp48la?dl=0
I did end up having to use capsules with higher SPL handling. Panasonic WM-61A capsules start to saturate with quite low SPL material, like piano.
The reason is that the canal hypes the frequencies that our ears are most sensitive to.
This requires a capsule with higher SPL handling, but it also lowers the microphone noise floor signal at those frequencies by the same amount, after processing through the necessary equalization filter.
The increase in S/N ratio is easily 15 dB, given that the peaks of the replica response are in the order of 15-20 dB, at the frequencies humans are most sensitive.
Always enjoy seeing seeing DIY innovation at the other end of the program chain. Some of the work nature recorders do with arrangements like Jecklin, SASS and PIBO can be startling.
The different HRTFs built into my daily travel in-ears - the LG/AKG Quadbeat 3 and Grado GR10 - aren't ideal for binaural recordings but your samples already sound excellent. Somewhere in the pile at home are pairs of ruler flat Audio Technica CK9 to try. Interesting note about noise, it's what I had in mind with the Panny capsules. In stock form they're not know for being very quiet. It makes sense the natural binaural boost in that range would help mitigate it. What capsule was used for these recordings?
The different HRTFs built into my daily travel in-ears - the LG/AKG Quadbeat 3 and Grado GR10 - aren't ideal for binaural recordings but your samples already sound excellent. Somewhere in the pile at home are pairs of ruler flat Audio Technica CK9 to try. Interesting note about noise, it's what I had in mind with the Panny capsules. In stock form they're not know for being very quiet. It makes sense the natural binaural boost in that range would help mitigate it. What capsule was used for these recordings?
Each HTRF is very different, in vivo measurements always shows an large variance and the ear concha is also very implicated.
But ,IMHO, these way of recording can be interesting in certain circumstances (depending on the sound engeneer experience) escpecially in term of soundstaging.
I've performed recordings with two large capsules microphones (in a ORTF configuration) of external sounds (open window) and played it RAW on my dipolar loudspeakers with my head at the mic location.
I can't explain why (experimented sound engeneers probably can) but the localisation is amazingly good and i'm totally unable to differentiate the real sound and the recorded one.
@ rdf
I'm glad to hear you like my recordings.
I used Primo capsules.
@cbdb and silversprout
Yours is a very good point. That is why I used an ear canal averaged from hundreds of real measurements.
The good news is that the differences of each individual's sets of ears and canals still come into play during the playback, although now the sources are in one specific position for all the sounds, that is the speaker's or headphone driver's location.
But yes, you are absolutely right, it is an approximation, as many things in the whole audio chain.
That being said, this is a product meant to make the recordings sound realistic and appealing to a majority of people.
I suppose one could make their own binaural recordings with an exact replica of their own head and equalize the microphone specifically to their own HRTFs. They would have a perfect recording, but it would work just for themselves.
About the use of a small mics in one's ears, its results are not as realistic as with the use of ear replicas with canals.
The ear canal's part in the HRTFs is just as important as the pinna's, and adding a microphone near the canal opening, no matter how small, changes the pressure wave drastically.
For example, using a capsule on the opening as if it was an earbud changes the boundary conditions of the pressure wave forcing its velocity to be 0 at the capsule's location.
I'm glad to hear you like my recordings.
I used Primo capsules.
@cbdb and silversprout
Yours is a very good point. That is why I used an ear canal averaged from hundreds of real measurements.
The good news is that the differences of each individual's sets of ears and canals still come into play during the playback, although now the sources are in one specific position for all the sounds, that is the speaker's or headphone driver's location.
But yes, you are absolutely right, it is an approximation, as many things in the whole audio chain.
That being said, this is a product meant to make the recordings sound realistic and appealing to a majority of people.
I suppose one could make their own binaural recordings with an exact replica of their own head and equalize the microphone specifically to their own HRTFs. They would have a perfect recording, but it would work just for themselves.
About the use of a small mics in one's ears, its results are not as realistic as with the use of ear replicas with canals.
The ear canal's part in the HRTFs is just as important as the pinna's, and adding a microphone near the canal opening, no matter how small, changes the pressure wave drastically.
For example, using a capsule on the opening as if it was an earbud changes the boundary conditions of the pressure wave forcing its velocity to be 0 at the capsule's location.
With the ORTF microphones configuration (cardioid pattern capsules) you have to set the angle between the two capsules in order to ajust your focus point. If you go far away of this point the localisation is lost (it is like photography because images are blurred if you are not at the focal distance).
My loudspeakers sweet spot is placed at the focus point of my microphone setup, this is the only way that i've found to make the recorded sound impossible to ABX.
I'm glad you are happy with the results of your ORTF configuration, silverprout.
Personally, on the other hand, I have never heard any other technique that is capable to come close to the binaural 360 soundstage representation.
I also think the timber of the instruments is recreated more precisely with a well equalized binaural microphone.
My thinking is that this has to do with the strong correlation between left and right channel, not always guaranteed with other recording techniques.
As a side benefit, binaural recordings also sound a little louder (and clearer) to me than other recordings at same dBFS, when played on regular speakers. I believe it is because this correlation makes cross talk cancellation less destructive.
Personally, on the other hand, I have never heard any other technique that is capable to come close to the binaural 360 soundstage representation.
I also think the timber of the instruments is recreated more precisely with a well equalized binaural microphone.
My thinking is that this has to do with the strong correlation between left and right channel, not always guaranteed with other recording techniques.
As a side benefit, binaural recordings also sound a little louder (and clearer) to me than other recordings at same dBFS, when played on regular speakers. I believe it is because this correlation makes cross talk cancellation less destructive.
Last edited:
Binaural
Le plus important c'est est la distance entre les oreilles( +/-18cm ) de la tête artificielle et pour être bon l écoute doit se faire au casque . Si vous avez de de l argent test avec des micros cravate sanken ou DPA .
The most important is the distance between the ears (+/- 18cm) of the artificial head and to be good listening must be done with headphones. If you have money, test with sanken or DPA lavalier mics.
Le plus important c'est est la distance entre les oreilles( +/-18cm ) de la tête artificielle et pour être bon l écoute doit se faire au casque . Si vous avez de de l argent test avec des micros cravate sanken ou DPA .
The most important is the distance between the ears (+/- 18cm) of the artificial head and to be good listening must be done with headphones. If you have money, test with sanken or DPA lavalier mics.
Last edited by a moderator:
IMHO sound engineers are able of outstanding recordings, but as they are technicians, they only do what they can with a lot of constraints.
In the audio world it seems that we must deal with the "made to sound good on everything" everytime, everywhere forever.
There is a risk of recording your breathing with a helmet (or eventually your farts
) IMHO.@ schlomoff
Dimensions of the head are important, but I wouldn't say they are the most important thing, let alone the ONLY important thing.
I did put particular attention in choosing a mannequin head that would have realistic measurements. Most of them are too little.
That being said, to get the full benefits of binaural recordings, you need replicas of the pinnae and canals, in my own experience.
I can guarantee you that binaural recordings can sound good on speakers as well, if done correctly.
@ silverprout
It may be the case, but binaural recordings have the declared intent of making things sound accurate.
My ears were definitely built with that intent in mind, I can assure you.
Dimensions of the head are important, but I wouldn't say they are the most important thing, let alone the ONLY important thing.
I did put particular attention in choosing a mannequin head that would have realistic measurements. Most of them are too little.
That being said, to get the full benefits of binaural recordings, you need replicas of the pinnae and canals, in my own experience.
I can guarantee you that binaural recordings can sound good on speakers as well, if done correctly.
@ silverprout
It may be the case, but binaural recordings have the declared intent of making things sound accurate.
My ears were definitely built with that intent in mind, I can assure you.
Have you heard this album :
https://www.amazon.fr/gp/product/B0024CP5IE/ref=oh_aui_d_detailpage_o02_?ie=UTF8&psc=1
Yes. That was done with a Neumann KU 100 binaural microphone.
Another proof of what can be achieved with this recording technique.
That microphone has straight tubes as ear canals, though.
I simply went a step further and made the canals more realistic.
In my experience this makes recordings sound even more natural.
It makes sense to me, since nature provided us with ear canals shaped like that instead of straight tubes.
Another proof of what can be achieved with this recording technique.
That microphone has straight tubes as ear canals, though.
I simply went a step further and made the canals more realistic.
In my experience this makes recordings sound even more natural.
It makes sense to me, since nature provided us with ear canals shaped like that instead of straight tubes.
Here is normal ear canal moldings (yellow) and integral ear canal moldings (blue)
You probably have to find a 3D model based on a large panel of integral ear canal moldings... not easy to find IMHO.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.