Fun 3d soundstage audio test!
Jan 4, 2006 at 9:50 PM Thread Starter Post #1 of 44

Luminair

Head-Fier
Joined
Nov 23, 2003
Posts
69
Likes
10
I enjoyed listening to this: http://media.putfile.com/Cereni---Holophonic

I don't know if anyone has played around with 3d audio tests here before, but since threads sink so quickly I figure whoever is reading this right now will enjoy this too
smily_headphones1.gif


FYI, at the start the guy walks in a circle around you, starting from behind. Then he shakes the matches up and down on either side of you.

My impression of it using AKG K240s is that the sounds in front seem to be up above my eyebrows, and the sound behind seem to be a bit below me. There is less definition behind than in front, possibly just because the recording was done that way. The high height sounds seem fine, and the low height sounds seem to be a bit behind me.

What really amazed me is that when the sound goes up and down, changing your headphone orientation doesn't change the up and down sense of the sound. What I mean is, rotating the cups to be upside down doesn't change the direction of the sound.

This is quite interesting to me as someone who has never put any thought into the way our hearing sense works.
 
Jan 4, 2006 at 9:54 PM Post #2 of 44
hey thats kind of neat... Could certainly hear some sense of vertical height with my MS2.

Garrett
 
Jan 4, 2006 at 10:02 PM Post #3 of 44
Quote:

Originally Posted by Luminair
What really amazed me is that when the sound goes up and down, changing your headphone orientation doesn't change the up and down sense of the sound. What I mean is, rotating the cups to be upside down doesn't change the direction of the sound.


Of course. Your brain assumes the sound to have been processed by your outer ear and interprets it according to that. If this, purely hypothetically, were replaced with a vertically mirrored copy, you'd probably hear "upside down".
 
Jan 5, 2006 at 8:14 AM Post #7 of 44
Quote:

Originally Posted by sgrossklass
Of course. Your brain assumes the sound to have been processed by your outer ear and interprets it according to that. If this, purely hypothetically, were replaced with a vertically mirrored copy, you'd probably hear "upside down".


This is not so easy. Up-down localisation is very ambiguous (esp. below 5kHz) and mostly supported by visual cues. In absence of visual feedback, the auditory system cannot unambiguously resolve, whether the sound comes from up or down.

The up-down (elevation) localization is not also easily trainable to a higher accuracy with 3D HRTF systems.

Fortunately our familiarity with sound types also helps in this. For example, if we hear a sound of an airplane, we are more likely to localise it up than down, if the spectral cues don't mismatch.
 
Jan 5, 2006 at 8:44 AM Post #9 of 44
awesome test............. although...................


Scared the crap outta me...... OMG reminded me of the "emily rose" and the girl in the ring.......... GOD
evil_smiley.gif


only tested on ymy sony's... really wanted to test it on my 485's.. but no.. Ill have nightmares... anything more spacelike and not umm so scary? lol
 
Jan 5, 2006 at 9:33 AM Post #10 of 44
interesting, I have a percussion sound track recorded similar to that.

I didnt get any in front or vertical soundstage.

I heard some guy walking behind me farther away, and then getting closer and backing up again.

Unlike others here though I didnt here anything above me or inside my head?

It was on the same ground plain of my ears except all the audio was movig about me behind me with tangible sound stage in regards to horizontal distance.

I was using my deep depth woodied hd600s however.
 
Jan 5, 2006 at 10:22 AM Post #11 of 44
Quote:

Originally Posted by Luminair
What really amazed me is that when the sound goes up and down, changing your headphone orientation doesn't change the up and down sense of the sound. What I mean is, rotating the cups to be upside down doesn't change the direction of the sound.


Why would that matter? Dynamic drivers are symmetrical (mostly) and don't give a damn what the recorded material is.

These are called binaural recordings. They are acheived with the use of a dummy head and a microphone in each ear. Thus, the sound is recorded very similarly to the way we hear it. So when you play it back with headphones, it's as if you were inplace of the dummy head during the recording.
 
Jan 5, 2006 at 11:51 AM Post #13 of 44
Quote:

Originally Posted by RnB180
interesting, I have a percussion sound track recorded similar to that.

I didnt get any in front or vertical soundstage.

I heard some guy walking behind me farther away, and then getting closer and backing up again.

Unlike others here though I didnt here anything above me or inside my head?

It was on the same ground plain of my ears except all the audio was movig about me behind me with tangible sound stage in regards to horizontal distance.

I was using my deep depth woodied hd600s however.



It's hard to get sound in front of you while your eyes are open because your brain says "What i cant SEE anything in front of me", so it places the sound behind you

you shouldn't hear anything inside your head (becasue that's the point of binaural) but there should be a little bit of up-and-down movement (but i didn't hear anything directly above me or anything)
 
Jan 5, 2006 at 12:51 PM Post #14 of 44
I've heard many binaural recordings and have always been puzzled by the fact on why sound that is supposed to be sounding in front of you always either sound right smack on the surface of your face or behind your head. Seems that headphones or our brains cannot visaulize properly sound coming from in front of us in the binaural recordings...Anyone care to explain why? seems that sound from behind is always easy to be heard and "visualized"...
confused.gif
 
Jan 5, 2006 at 1:08 PM Post #15 of 44
Quote:

Originally Posted by hugz
It's hard to get sound in front of you while your eyes are open because your brain says "What i cant SEE anything in front of me", so it places the sound behind you.


Actually localisation BEHIND is even more difficult than in front, because spectral difference due to masking for sound sources from behind the listener is so completely different.

However, front-to-back localisation is trainable in virtual audio systems.

That is, after you train to believe your ears not your eyes, you can localize a 3D HRTF sound position fairly accurately in front-back plane.

This, as stated above, is not so accurate for elevation (up-down) and cannot be trained almost at all for 3D audio systems.
 

Users who are viewing this thread

Back
Top