Virtual Barber Shop

[ra7]

Hey guys,

I don't know if you've heard this amazing piece
on the ability of our ears.

Put your most expensive headphones on and have a go!
The end is the best part!
http://www.youtube.com/watch?v=IUDTlvagjJA

Regards,
Rahul

P.S.: I have high quality version which sounds way better.

[Atilla]

This is great fun but you should really find the original version. The youtube one doesn't do it justice

[andy_c]

Is there a lossless version of this available? The best I have found so far is a 192k MP3 here http://r-1.ch/VB1.mp3

[ra7]

I think I have the one you refer to above.
But yes, it is amazing how well they have
done it. Any idea how it was recorded?

He does say during the clip that they put
two mikes where our ears would be. But
what mikes are they? Bloody good ones
I imagine.

[andy_c]

I don't know what they used. Back in the mid 1970s, in Audio magazine, there was a column called Audio Etc. by a guy named Edward Tatnall Canby. He did a series of columns about binaural recordings that were fun reading. He apparently used a dummy head containing microphones made by Sennheiser. I found something on the Sennheiser site about them here.

I vaguely remember Canby saying that two microphones without the dummy head worked pretty well also. That Sennheiser page also shows microphones that are made to be warn on one's head like headphones. I don't know if any of these are still being made.

There's also an article by Michael Gerzon on the topic here (PDF file).

[Key]

You can the same effect by getting a dummy head that is approximately the same size as your head and sticking high quality omni-directional mics in the ears. The closer the dummy is to being exactly the same as your head the better the results.

I have a cheapo one I made with some Radioshack omnis and a Louis Armstrong singing doll. http://i.ebayimg.com/12/!B(R2FbgBGk~$(KGrHgoH-D4EjlLlzw8!BKbe9JuPHw~~_12.JPG

Anyway I have a couple problems with that demo that I can point out. They use a little bit of trickery to get it to work well imho.

For instance they cover up the weak point in HRTF with suggestion. One of the biggest problems I have with bi-naural is that the front image can invert and become the back image depending on how you focus on the sounds. If you listen closely they aid it along with suggestion by saying "around the back" when the sounds go around the back. If you were to play this same exact sound only with the words "around the front" overdubbed you would hear it as wrapping around the front.

The inclusion of a static acoustic guitar provides a base for you to contrast the other sounds in the recording. Of course if they recorded this more as a natural one take binaural recording there would be ambient noise that could provide this same base.

The sound of the bag over the head seems to make the mics very obvious. My ears don't sound like mics being rubbed unless I stick my ear on a desktop or a diaphragm of some sort.

And beyond that the recording just sounds overproduced and edited. It has what I call the "theme park" effect. Where everything sounds like I am on a ride at Disney Land.

Maybe I don't like it as much because the first time I did this it was live with someone doing the scissors to the dummy head in the next room. It seemed much more realistic without all the voice over artists and production value.

[andy_c]

Originally Posted by Key

And beyond that the recording just sounds overproduced and edited. It has what I call the "theme park" effect. Where everything sounds like I am on a ride at Disney Land.

Well yeah, but I think it was meant as more of an entertainment thing, rather than being purely educational. I found it pretty entertaining anyway. I was laughing through most of it.

It would be interesting to hear some purely acoustic music recorded in this way.

[janneman]

Hmmm. I thought this was extensively DSP'd. Playing with level, delay times and phases. Isn't this what Bose does with their docking stations?

jd

[ra7]

I agree with andy. It is funny. First time I heard something done this way.

[Key]

I didn't want to poo poo the whole experience I still enjoyed it actually. I just thought like any demo it has a bit of subtle goosing to make the results better than they should be. And the trickery is not obvious to someone who hasn't experimented with the techniques.

I really don't get how Cetera even fits into this demo. They tell you that they use a bi-naural head but they also whisper that the Cetera Algorithm is what makes it work better. So maybe it's a mixture of binaural and dsp simulated binaural mixing? I thought Cetera was for hearing aids but it would make sense that the same thing which works in a hearing aid would work for encoding sound to 2 channels the way we actually hear it.