Binaural Head

[Dreadhead]

Hi All,

I'm thinking of buying two measurement mikes and turning my new mannequin head Bob into a binaural head with ear canals etc. Once I do that I will also send it off to be calibrated and then the fun will really begin

To do this though I need a dremel or something that will let my router out a couple access ports etc into the back of Bob. Anyone have something suitable I can borrow?

[Dusty Chalk]

What about finding a local body shop that works with fiberglass? You can't just drill fiberglass (well, you can, but...I wouldn't).

Also, a binaural head mic actually makes an attempt at modeling the acoustic properties of the ear flaps, reflective properties of the surface, etc. I mean, it'll probably be fine for measurements, I'm just not sure about using it for a binaural mic.

[grawk]

I was expecting something much more lurid in this thread

[Dreadhead]

What about finding a local body shop that works with fiberglass? You can't just drill fiberglass (well, you can, but...I wouldn't).

Also, a binaural head mic actually makes an attempt at modeling the acoustic properties of the ear flaps, reflective properties of the surface, etc. I mean, it'll probably be fine for measurements, I'm just not sure about using it for a binaural mic.

Dusty, thanks for the advice re the body shop but I actually have quite of a bit of experience with fiberglass through my work on various sailboats through the years.

The head I'm using has ears etc so the flaps are covered mostly (the Aachen head doesn't even have real ears, HEAD acoustics - NVH Division - Binaural recording systems - Overview). The newest binaural heads (Aachen/Head, Neumann) seem to have a hard surface with a matte finish which Bob is pretty close to as well. Also the methods for binaural vary a bit because some even use mics outside the head like: http://en.wikipedia.org/wiki/File:Dummyhead.jpg

I figure if they will calibrate it it should be damn close after the equalization is applied. The professional ones have the equalization built in.

[Hopstretch]

I was expecting something much more lurid in this thread

He didn't say anything about not drilling the mouth.

[Dusty Chalk]

Dusty, thanks for the advice re the body shop but I actually have quite of a bit of experience with fiberglass through my work on various sailboats through the years.

Okay, so you know what to literally watch out for. Good, that was my main concern.

I am very interested in what results you can obtain, will be monitoring your progress. Heck, if it works out, I might have to try to do the same thing myself.

[Dreadhead]

He didn't say anything about not drilling the mouth.

[Dreadhead]

Okay, so you know what to literally watch out for. Good, that was my main concern.

I am very interested in what results you can obtain, will be monitoring your progress. Heck, if it works out, I might have to try to do the same thing myself.

I emailed the guy who I bought my calibrated mic from to see if he would calibrate the head and for how much if he will. He usually charges $50 or so for a mono mic so I would assume $100 or so.

If I get the calibration done for that price and including the mics I may get out for under $300 not bad considering the Neumann is over 5 grand

[grawk]

Not to say that your measurement mics in a manequin head would be comparable to a neumann dummy head.

[Dusty Chalk]

How much do the Aachen systems run? I mean, are there even any binaural microphones for <US$2+K? (And I'm pulling that number out of my ass -- I haven't seen any reasonably priced binaural mics at all.) If it comes even close to sounding like a binaural recording, <$300 is definitely more my style.

[Dreadhead]

Not to say that your measurement mics in a manequin head would be comparable to a neumann dummy head.

Why not? I am actually being serious. The distortion numbers on the measurement mics is very low and the transducers in the KU100 are pretty small too. I guess I could go nuts and do it with Earthworks M30s instead but that would drive the cost way up. I long ago abandoned that because it has a brand name on it is something special.

How much do the Aachen systems run? I mean, are there even any binaural microphones for <US$2+K? (And I'm pulling that number out of my ass -- I haven't seen any reasonably priced binaural mics at all.) If it comes even close to sounding like a binaural recording, <$300 is definitely more my style.

The Aachen is even more expensive (I think north of 20k) because they pretty much only sell it as a measurement device. Long term I'd love to find a KU100 but I could never really justify it but if I keep this under $500 it's a lot easier to justify.

I have a paper from the guys at Aachen that I have to read on the auralization of spaces goes rummaging through his briefcase.

[John Buchanan]

I would have a go at inserting microphones into the canals, but leave the microphones poking out of the canals with their diaphragms at the same position as the diaphragm of the replay headphone. It seems odd that most dummy heads have the microphones nestled into an ear canal, where the only correct replay for this recording will be with an in-ear-canal headphone, because if you are using an around ear headphone, the sound will be modified by the dummy head pinna on recording and your own pinna on replay. It would be interesting to try a dummy head with the microphones in several positions for different headphones.

1. buried in the ear canals for IEM replay

2. with the microphones poking out of the canal slightly for on ear phones

3. with the microphones as mentioned above for around pinna headphones

4. at the side of the head and facing forward for something like a Stax Sigma or an AKG K1000 (with the microphones placed at the centre of a prospective headphone transducer)

[Duggeh]

Neumann had the mics down a canal on the KU81i, and close to the entrance of the ear on the KU100 now. Of course, they've also got a shitload of clever EQ going on in both cases.

[Dreadhead]

Yup the EQ and research into that EQ is what you're paying for.

I plan to apply the EQ using an impulse response function which is a lot easier if you're dealing with the data digitally and doesn't have the phase issues (I think Aachen may do it digitally too).

Heard back from the calibration guy and he is interested but needs to think about it a bit because of concerns about the HRTF effecting the results but I only want the calibration with the head facing the source anyway.

[schulein]

Hi "Dreadhead" from Bob Schulein,

I have just joined Head Case

As an individual building a dummy head for binaural experimentation, you may be interested in work that I have been doing based on this technology.

I have developed an experimental binaural audio with HD video technique for capturing a point of view "you are there" experience from a musical performance. I call the process ImmersAV, in that you become immersed in an Audio/Video experience.

If interested, I have uploaded two examples to You Tube:

http://www.youtube.com/watch?v=97M1-fCBkx4

and

There are about 10 others examples as well. If you search You Tube for ImmersAV you can find them all.

Some examples have an audio video sync issue due to the You Tube up load process. I am currently working on a solution for that problem.

Other postings will be coming as I have ramped up my production schedule.

My interest at this point is in obtaining feedback as to the entertainment value of such productions.

I have many yet to be tried production techniques involving multiple video and audio points of view, for the future.

Also if you download the HQ version from You Tube and use Quick Time to create an

i pod file version, the productions will be preserved with good audio and video fidelity.

Best if viewed on an iPod Touch or iPhone.

As an additional point, I have been presenting a master class on Binaural Audio Technology at conventions of the Audio Engineering Society over the past year (Audio Engineering Society (AES)).

Later this fall, my class presented in Munich Germany will be available on line free to members of the AES.

All comments and suggestions welcome.

Bob Schulein

[Donald North]

This looks to be a fun project!

Back in college in the early 90s I did a lot of binaural experimentation. For a dummy head I used a styrofoam head with silly putty pinnae and flushed Panasonic microphone capsules (purchased from Digi-Key) at what would be the entrance to the ear canals. No electronic EQ was applied. Replay was through beyer DT990 headphones. This combo sounded better and more realistic than the Sennheiser 2002 dummy head which I also owned at the time and later sold. With this kunstkopf, I made several recordings that were startlingly realistic to my friends/classmates. Listening through beyer DT990 yielded the best results, bettering the K240M, K240DF, K340, and Stax SR84 at the time.

For best personal results, I made these hangers for the Panasonic elements that fit over my ears, again positioning the mic at the entrance to the ear canal.

[Dusty Chalk]

...the Sennheiser 2002 dummy head...

What is this? I googled, and all I could find on Sennheiser's website was microphones ("stethoset"), but the picture shows a dummy head. link

On an unrelated note, I like the picture of Oskar as well:

Edited October 3, 2009 by Dusty Chalk

[Donald North]

The MKE 2002 was an affordable binaural microphone made by Sennheiser. It looked like a stethoscope which you could fit on the included head "Fritz"-I believe or wear yourself. Price was around $800 when I bought it new in 1991. I liked the formfactor, but sound quality was not nearly as good as my homemade using the Panasonic elements.

[LFF]

You have the basic idea right. However, Bob is not suitable to make a "correct" binaural dummy head. Trust me, I have read into this stuff for over a year and have made a few mics.

I say ditch Bob and make a holophonic mic. It's cheaper and it is speaker compatible. If you truly want to make a dummy head, you can't use Bob as is. Your are NOT just paying for the EQ and research. The more expensive heads use materials that actually mimic the acoustic response of the skin and "average" facial features. They also weigh close to what an "average" head weighs. The capsules are also of very high quality. The mic circuit you use will also affect the sound quality.

Anyway, PM me if you have any more questions or feel free to post them here.

[jgazal]

The head related transfer function (HRTF) is extremely idiosyncratic.

Why don

[Duggeh]

The thing abotu sticking mics in your own ears, is that all the noises from inside your head end up in the mics. Chewing or swallowing, breathing noises, plus keeping your own head rock still is a pita and makes it hard to even look at your recorder.

I've got my head set up now. I could make a wee matchbox style recording if anyone is interested.

[Dusty Chalk]

Oui, Ja, Da, Yes please, &c.

[LFF]

The thing abotu sticking mics in your own ears, is that all the noises from inside your head end up in the mics. Chewing or swallowing, breathing noises, plus keeping your own head rock still is a pita and makes it hard to even look at your recorder.

YUP!

Also HRTF algorithms can't get close to replicating a true binaural recording. I have tried a lot of them and none can get even close to even a holophonic mic.

However, for those who don't have any experience with binaural, it's quite easy to fool them with HRTF algorithms.

[LFF]

Here is a recording with a Grado Holophonic mic.

You don't always need a dummy head to get a "you are there" sound. As I said before, the benefit of this is that it sounds great on speakers and even better on headphones.

[Dreadhead]

You have the basic idea right. However, Bob is not suitable to make a "correct" binaural dummy head. Trust me, I have read into this stuff for over a year and have made a few mics.

I say ditch Bob and make a holophonic mic. It's cheaper and it is speaker compatible. If you truly want to make a dummy head, you can't use Bob as is. Your are NOT just paying for the EQ and research. The more expensive heads use materials that actually mimic the acoustic response of the skin and "average" facial features. They also weigh close to what an "average" head weighs. The capsules are also of very high quality. The mic circuit you use will also affect the sound quality.

Anyway, PM me if you have any more questions or feel free to post them

here.

I disagree. As far as weight and surface finish: The Gras KEMAR head is smooth (so smooth it shines) weighs 17.1 lbs for a head, neck, shoulders and torso and is another standard in Etymotic research.

http://www.grasinfo.dk/documents/pd_45BA_ver_12_10_07.PDF

The soft pinnae are cool but I could likely get that to work later if I tried. The KU100 is around 8lbs which is in the range for the human head though but the fact that the KEMAR is lighter points to this being not much of an issue. Also I could easily add weight anyway.

Mic quality is an issue as you say but if I set it up this way I can easily switch to earthworks M30s which I argue are about as high quality as it gets and the ECM8000s I was starting with are very well reviewed for sound quality and low distortion at least on the websites I've visited.