This is a 5 second pre-intro to the track that I mentioned earlier, where you can hear a voice go "one, two, one-two-three" and when I sit in my normal listening position, even though the speakers are at about the 11 o'clock and 1 o'clock position to me, I can hear several "knock" sounds right around at the 3-o'clock position. When I turn my head to the right it still appears to come from the same place. I got my wife to have a listen and she heard exactly the same.phase_accurate said:
There was a bit of 25 Hz rumble in the recording so I filtered it out, and also brought up the level a bit. I hope this hasn't damaged the stereo illusion I am talking about.
Also, there is no music in this wave file though it was taken from a nominally copyrighted CD, so if anyone feels there is a copyright problem I will take it down.
http://www.circlotron.haydenpratt.com/intro.wav
I find it funny how people talk about how such and such amplifier or source or whatever improves the soundstage, when the root of the problem is geometric and has to do with the way the sound is recorded.
Even using a cheap headphone system, as long as the frequency response and phase aren't too messed up, a binaural recording can provide a pretty amazing soundstage. The limits of binaural recordings, however, are that each person has a different HRTF so the generic dummy head has limitations. Most dummy heads completely eliminate the pinnea and only have an ear canal, as the outer parts of the ear vary a lot. On some binaural recordings where the microphones were actually in someone's ear, however, the result may be better for some listeners. If you could use your own HRTF the imaging would be perfect, with even clear up/down, but that's not practical.
With speakers the main problem is crosstalk from each speaker to the opposite ear (while this happens with any sound source, it's already included in a recording so you're getting double crosstalk). If you put a soundproof barrier in between you could listen to binaural recordings with speakers properly, and some have indeed tried that, but it's not practical so software crosstalk cancellation can help, and there are a number of demos out there.
One of the demos I have found around the net is the stereo dipole. This demo is quite dated, but the effect is still quite an experience for those not familiar with these types of techniques. The way it works is that with two speakers placed just inches from each other, a listener some distance away (10 degree seaker separation from listener's position) can get amazing side to side imaging going off to infinity in the left or right with properly processed sound:
Positioning the speakers (or it won't work); the sweet spot is very small and sharply defined:
http://www.isvr.soton.ac.uk/FDAG/vap/html/readme.html
Processed sound files:
http://www.isvr.soton.ac.uk/FDAG/vap/html/download.html
The explanation:
http://www.isvr.soton.ac.uk/FDAG/vap/html/sd.html
Even using a cheap headphone system, as long as the frequency response and phase aren't too messed up, a binaural recording can provide a pretty amazing soundstage. The limits of binaural recordings, however, are that each person has a different HRTF so the generic dummy head has limitations. Most dummy heads completely eliminate the pinnea and only have an ear canal, as the outer parts of the ear vary a lot. On some binaural recordings where the microphones were actually in someone's ear, however, the result may be better for some listeners. If you could use your own HRTF the imaging would be perfect, with even clear up/down, but that's not practical.
With speakers the main problem is crosstalk from each speaker to the opposite ear (while this happens with any sound source, it's already included in a recording so you're getting double crosstalk). If you put a soundproof barrier in between you could listen to binaural recordings with speakers properly, and some have indeed tried that, but it's not practical so software crosstalk cancellation can help, and there are a number of demos out there.
One of the demos I have found around the net is the stereo dipole. This demo is quite dated, but the effect is still quite an experience for those not familiar with these types of techniques. The way it works is that with two speakers placed just inches from each other, a listener some distance away (10 degree seaker separation from listener's position) can get amazing side to side imaging going off to infinity in the left or right with properly processed sound:
Positioning the speakers (or it won't work); the sweet spot is very small and sharply defined:
http://www.isvr.soton.ac.uk/FDAG/vap/html/readme.html
Processed sound files:
http://www.isvr.soton.ac.uk/FDAG/vap/html/download.html
The explanation:
http://www.isvr.soton.ac.uk/FDAG/vap/html/sd.html
Circlotron said:
Interesting - Not sure if the filtering caused it, but my son and I hear the beat counting dead ahead and the knocking about 430. (using cheap "computer speakers.")
With my Sony projector TV it arms with virtual surround sound option that uses only the left and right speakers sometimes the effect is quite suprising. How do they do that?
Chris
Chris
Another way to reduce the interaural crosstalk (signal from right speaker reaching left ear, for example) is to use the old Carver "Sonic Hologram Generator" such as the C-9. This is an analog way of creating a signal in the right speaker which is a filtered and delayed version of the left stereo channel. When this optimally delayed sound reaches the right ear it cancels the unwanted signal from the left speaker, effectively making the left speaker inaudible to the right ear. The effect is to move some sounds WAY off to the left or right of the speaker positions. It requires precise head positioning so that the timing is just right.
Polk made speakers that achieved the same end by having a second array of drivers positioned next to the primary drivers, displaced by a distance chosen to achieve the desired signal delay. Of course this can now be done in the digital domain as well (as mentioned previously).
I've often wondered whether speakers with poorly controlled diffraction (which implies a secondary radiator of sound) could yield some spurious imaging effects much like those designed into the Polk speakers. Perhaps room reflections (either in the original recording venue or in the listening room) can have a similar effect? When reviewers rave about speakers that present images beyond the speaker boundaries I'm often left wondering if that is necessarily a figure of merit.
Polk made speakers that achieved the same end by having a second array of drivers positioned next to the primary drivers, displaced by a distance chosen to achieve the desired signal delay. Of course this can now be done in the digital domain as well (as mentioned previously).
I've often wondered whether speakers with poorly controlled diffraction (which implies a secondary radiator of sound) could yield some spurious imaging effects much like those designed into the Polk speakers. Perhaps room reflections (either in the original recording venue or in the listening room) can have a similar effect? When reviewers rave about speakers that present images beyond the speaker boundaries I'm often left wondering if that is necessarily a figure of merit.
To see it in practice, using a tweeter like the one in the pic, Focal TN51, which has dispersion control flaps, is it better to use it
in the car in a vertical flap orientation ( flaps eliminating left and right radiation ) or in a horizontal orientation ( flaps eliminating upper and lower radiation, giving a wide angled radiation to left and right ) to have better/precise soundstage ?
http://www.in-akustik.com/focal_jmlab/Focal_audiomobile/Focal2000/drivers/techdat.asp?typ=TN51
in the car in a vertical flap orientation ( flaps eliminating left and right radiation ) or in a horizontal orientation ( flaps eliminating upper and lower radiation, giving a wide angled radiation to left and right ) to have better/precise soundstage ?
http://www.in-akustik.com/focal_jmlab/Focal_audiomobile/Focal2000/drivers/techdat.asp?typ=TN51
An externally hosted image should be here but it was not working when we last tested it.
- Status
- Not open for further replies.