I just made a pair of tapeop omnis and stuck them in my ears... I guess that makes me a dummy head. This is not music-- just a shaker and a bell to test it out. You have to use headphones or earphones to get the 3D effect, and how well it works may depend on how similar your pinnae are to mine.
http://www.ubertar.com/creot/binaural.MP3
binaural recording
Moderator: cgarges
-
spankenstein
- re-cappin' neve
- Posts: 639
- Joined: Sat Apr 17, 2004 8:58 pm
-
ubertar
- ears didn't survive the freeze
- Posts: 3775
- Joined: Wed Feb 18, 2004 7:20 pm
- Location: mid-Atlantic US
- Contact:
The odd thing I'm finding is that while up and down seems to work well (and right/left), front/back is not right-- everything sounds like it's from behind. That's better than the other way around, since it's not something you can do otherwise, but it's not fully 3D. I wonder why sounds in front still seem to be from the back. Anyone have any ideas?
-
spankenstein
- re-cappin' neve
- Posts: 639
- Joined: Sat Apr 17, 2004 8:58 pm
-
Professor
- ghost haunting audio students
- Posts: 3307
- Joined: Wed May 07, 2003 2:11 pm
- Location: I have arrived... but where the hell am I?
I recently did some messing around with taking 5.1 mixes and runninng them through a binaural mapper program to place surround-simulations in headphones. My thought was that since the iPod phenomenon has so many people listening on a very controlled and predictable system that binaural surround might just be a reasonable delivery format. I posted some files and details here and will put up more as I get a chance to play around a bit more.
I also ran into the heavier rear than front sensation with the 5.1 to binaural mapper.
I'm not sure what would make the rear seem more active, or more distant that the fronts and will have to look into some HRTF research to explain that. I do know there is a problem called the "cone of confusion" inherent in binaural recording. The cone is this problem where a sound that appears at a particular location from left to right appears at your ears/microphones with the same timing and volume differences regardless of whether it is in front of you, above, behind, or below. So if the sound is 45? to your right and 5-ft in front, and is rotated up to be 45? to the right and 5-ft above you, then behind, then below, the sound will be pretty much the same. Reverse that and deliver a sound to your ears that is located at one of these points and your ear won't know quite which direction it's coming from.
This is primarily a problem for spherical binaural mics like the ones from THE, Schoeps, or mics on a Jecklin or Schneider disk. When you move to a dummy head, things improve, but not completely.
Getting the microphones to the proper depth inside the dummy ears will help the head to perform more EQ changes to assist your ear in differentiating between front and rear. Having well-shaped-, yet average pinnae (outer ears) will help translate better. Having hair, simulated muscle, simulated skin, etc. will also help to improve the accuracy.
It's actually really fascinating to read some of this research though some of it can be a little math intensive. There's an article available here as part of a free sample of the AES on-line journal.
Now obviously you were using your own head which is going to be a better dummy head than anything else, but the microphones probably weren't deep enough in your ears to get the full effect of your pinnae, muscle and hair. And don't try to put them deeper!!
As it happens, a lot of research has gone into the shape of ears for binaural dummies because it seems that some ear shapes are better than others for reproducing directional information. For what should be obvious reasons, your own ears translate the best into your own head, but may not translate well for me or anyone else. So considerable experimenting went into finding a good 'average' shape that would translate well for all kinds of ear shapes.
One thought for the front to back issue might be to move the mics further to the front of your head, like by taping them onto the sides of sunglasses or something. Since your ears sit slightly further back than right at the middle of your head, moving forward might cause more absorbtion from hair, skin, etc. and allow the EQ effects to cancel more sounds from the rear and less from the front, so the images from the rear seem more distant while the ones from the front are more clearly placed. Of course, it could also be the exact opposite.
Either way, this is fun stuff to play with and you should keep posting your tests. Maybe I'll hop on a drumset later with a couple tiny omnis stuck to my ears to see what I can capture - should be interesting while I turn my head to face the instruments I'm playing.
-Jeremy
I also ran into the heavier rear than front sensation with the 5.1 to binaural mapper.
I'm not sure what would make the rear seem more active, or more distant that the fronts and will have to look into some HRTF research to explain that. I do know there is a problem called the "cone of confusion" inherent in binaural recording. The cone is this problem where a sound that appears at a particular location from left to right appears at your ears/microphones with the same timing and volume differences regardless of whether it is in front of you, above, behind, or below. So if the sound is 45? to your right and 5-ft in front, and is rotated up to be 45? to the right and 5-ft above you, then behind, then below, the sound will be pretty much the same. Reverse that and deliver a sound to your ears that is located at one of these points and your ear won't know quite which direction it's coming from.
This is primarily a problem for spherical binaural mics like the ones from THE, Schoeps, or mics on a Jecklin or Schneider disk. When you move to a dummy head, things improve, but not completely.
Getting the microphones to the proper depth inside the dummy ears will help the head to perform more EQ changes to assist your ear in differentiating between front and rear. Having well-shaped-, yet average pinnae (outer ears) will help translate better. Having hair, simulated muscle, simulated skin, etc. will also help to improve the accuracy.
It's actually really fascinating to read some of this research though some of it can be a little math intensive. There's an article available here as part of a free sample of the AES on-line journal.
Now obviously you were using your own head which is going to be a better dummy head than anything else, but the microphones probably weren't deep enough in your ears to get the full effect of your pinnae, muscle and hair. And don't try to put them deeper!!
As it happens, a lot of research has gone into the shape of ears for binaural dummies because it seems that some ear shapes are better than others for reproducing directional information. For what should be obvious reasons, your own ears translate the best into your own head, but may not translate well for me or anyone else. So considerable experimenting went into finding a good 'average' shape that would translate well for all kinds of ear shapes.
One thought for the front to back issue might be to move the mics further to the front of your head, like by taping them onto the sides of sunglasses or something. Since your ears sit slightly further back than right at the middle of your head, moving forward might cause more absorbtion from hair, skin, etc. and allow the EQ effects to cancel more sounds from the rear and less from the front, so the images from the rear seem more distant while the ones from the front are more clearly placed. Of course, it could also be the exact opposite.
Either way, this is fun stuff to play with and you should keep posting your tests. Maybe I'll hop on a drumset later with a couple tiny omnis stuck to my ears to see what I can capture - should be interesting while I turn my head to face the instruments I'm playing.
-Jeremy
Who is online
Users browsing this forum: No registered users and 52 guests