Agreed. Dr. Griesinger suggests not relying mutch on phantom sources and reverb must be decorrelated between channels. (Link not suggested for 'purists')
Quite elegant solution for multiseat AND multiorientation listening - slightly less complex then WFS.
but what is the solution? I have read the presentation but I cannot find in it any solution WRT specific method of audio reproduction at home
I would say they are errors of different severity. It is commonly hypothesised in the literature that pinna localisation is primarly because of spectral tracking of notches. The HRTF is 'smoother' for sounds directly in front than from +/- 30 degrees side. Then a claim can be made that it is a greater error (from the perception point of view) to have wrong notches than to not to have notches at all, because perception is based on available cues and 'fuzzy' cues are neglected. In this light wrong pinna cues in stereo are worse than wrong pinna cues of Ambiophonics.
Also on the other hand, propably 90% of the sound is intended to come from center location so using Ambiophonics at least 90% of the time should be satisfactory
- Elias
Also on the other hand, propably 90% of the sound is intended to come from center location so using Ambiophonics at least 90% of the time should be satisfactory
- Elias
It's in the presentation. A common 3/2 setup was used. His conclusion: "It is possible to make videos of music performances that re-create the excitement and involvement of a live performance."
this 3/2 setup is 5-channel multichannel - 2 front, center and 2 rear? sorry for lame question but multichannel is not my cup of tea
Last edited:
I'm not sure why the error introduced by stereophony is more severe than the error introduced by Ambiophonics (which probably introduces even more errors when used with recordings made for stereophony).
The simple solution is to add a center speaker to stereophony. It solves the problem of compatibility with existing recordings and size of sweet spot too.
The simple solution is to add a center speaker to stereophony. It solves the problem of compatibility with existing recordings and size of sweet spot too.
I would say they are errors of different severity. It is commonly hypothesised in the literature that pinna localisation is primarly because of spectral tracking of notches. The HRTF is 'smoother' for sounds directly in front than from +/- 30 degrees side. Then a claim can be made that it is a greater error (from the perception point of view) to have wrong notches than to not to have notches at all, because perception is based on available cues and 'fuzzy' cues are neglected. In this light wrong pinna cues in stereo are worse than wrong pinna cues of Ambiophonics.
Also on the other hand, propably 90% of the sound is intended to come from center location so using Ambiophonics at least 90% of the time should be satisfactory
- Elias
Last edited:
here:
Problem is single-seat-head-in-the-vise only imaging and (poor)envelopment with standard 2 channel listening.
(one) solution lies in the recording process described in the presentation and using many discreet sources in playback. Others are already mentioned: WFS and higher-order ambisonics.
Last edited:
yes, but this is solution on the side of recording process I was asking about proposed solution on the side of reproduction process
A psychoacoustically optimised upmixing algoritm (to minimise phantom imaging and replay reverb from optimal location)
Last edited:
yes, trinaural is interesting as is Dynavector's Superstereo, a four channel solution: http://web.onetel.net.uk/~dynavector/
Try moving your speakers to +/- 20 degrees or about one third the normal spacing. Also move forward and back to see if you can get a stage you are more comfortable with. You can also adjust the delay and attenuation parameters to taste.
This is an e-mail I just got from an Ambiophonics user. He is to my knowledge likely the third person in the world to have a four loudspeaker Ambiophonic system going. He is not a member of this forum.
I’m listening to Handel’s Water Music Suite using a 4-channel Ambiophonics system. Using Winamp and the Ambiophonics Winamp plugin as a digital signal processor I get a truly remarkable sense of clarity, localization, and space despite the low-cost of the components. I doubt the total cost of my 4 speakers exceeds $400, for example. The result is a sound quality that could not be obtained using conventional stereo and state-of the art components in my opinion. Although one might get better test results using expensive equipment, I doubt that a stereo-based system could ever create the ambience I’m getting using the crosstalk cancellation algorithm implemented in Ambiophonics. Many thanks for making it possible to hear more music than I ever thought possible.
Jeffrey Gimprich
Email: gimprich@optonline.net
Ralph Glasgal
Home Page
I’m listening to Handel’s Water Music Suite using a 4-channel Ambiophonics system. Using Winamp and the Ambiophonics Winamp plugin as a digital signal processor I get a truly remarkable sense of clarity, localization, and space despite the low-cost of the components. I doubt the total cost of my 4 speakers exceeds $400, for example. The result is a sound quality that could not be obtained using conventional stereo and state-of the art components in my opinion. Although one might get better test results using expensive equipment, I doubt that a stereo-based system could ever create the ambience I’m getting using the crosstalk cancellation algorithm implemented in Ambiophonics. Many thanks for making it possible to hear more music than I ever thought possible.
Jeffrey Gimprich
Email: gimprich@optonline.net
Ralph Glasgal
Home Page
Now I'm confused. "Normal" stereo spacing is usually considered to be 60° or +/-30 from the center line. You're suggest a 40° total arc? I thought it was meant to be about 20-25 total for Ambiophonics. Or do you mean "about" 20 degrees total? That is what I tried.
Let me explain a bit about pinna localization issues.
Humans localize by sensing ITD and ILD at lower frequencies and use a single pinna at frequencies from say 2000 Hz and up. There may also be a two pinna mechanism especially for height. The pinna act as direction finders by coloring the sound that reaches the ears before it can get to the ear canal.
Obviously, a sound that comes from the far sides has a clear shot at the ear canal and so is not much influenced by the pinna. Most other angular sources are converted by the pinna into a very complex pattern that like fingerprints are unique to each individual.
Thus if speakers, as in Ambiophonics, are more central than in stereo, the pinna error will be minimized for the center stage and soloists. Now if you have say a left channel signal only, the image should be at 90 degrees to the left. The large ITD and ILD will place it there but the pinna, in contrast, are saying that the sound is coming from the front speakers. Often the result is that the sound is sensed as coming from say 80 degrees instead of 90.
There are two fixes for this. One is described in the previous message. Use two additional speakers behind the listening position also RACE cancelled. The pinna pattern from the rear mixes with the pattern from the front. Since such a mixed pattern is unnatural, the brain ignores this and can better localize to the extreme sides, since as already mentioned, the far side pinna pattern is of lesser complexity and importance.
Another method is to put two speakers to the sides and feed them from a Prologic decoder so they are mostly only active when a signal is all left or all right. But the rear speaker option allows viewing 360 degree surround movies or sports events where direct sound images at the sides can be quite enthralling.
Ralph Glasgal
Home Page
Humans localize by sensing ITD and ILD at lower frequencies and use a single pinna at frequencies from say 2000 Hz and up. There may also be a two pinna mechanism especially for height. The pinna act as direction finders by coloring the sound that reaches the ears before it can get to the ear canal.
Obviously, a sound that comes from the far sides has a clear shot at the ear canal and so is not much influenced by the pinna. Most other angular sources are converted by the pinna into a very complex pattern that like fingerprints are unique to each individual.
Thus if speakers, as in Ambiophonics, are more central than in stereo, the pinna error will be minimized for the center stage and soloists. Now if you have say a left channel signal only, the image should be at 90 degrees to the left. The large ITD and ILD will place it there but the pinna, in contrast, are saying that the sound is coming from the front speakers. Often the result is that the sound is sensed as coming from say 80 degrees instead of 90.
There are two fixes for this. One is described in the previous message. Use two additional speakers behind the listening position also RACE cancelled. The pinna pattern from the rear mixes with the pattern from the front. Since such a mixed pattern is unnatural, the brain ignores this and can better localize to the extreme sides, since as already mentioned, the far side pinna pattern is of lesser complexity and importance.
Another method is to put two speakers to the sides and feed them from a Prologic decoder so they are mostly only active when a signal is all left or all right. But the rear speaker option allows viewing 360 degree surround movies or sports events where direct sound images at the sides can be quite enthralling.
Ralph Glasgal
Home Page
I am sorry. I mean one third which of course is 20 degrees total but you can try going up to 30 degrees total if there is some problem or your speakers are unusual. Moving backwards and forwards can help in deciding this for a given parameter setting.
Ralph