« Return to Thread: Some Ambisonics questions
Re: Some Ambisonics questions
by Josh Parmenter
::
Rate this Message:
Hi Dan... just to join in a bit...
On Mar 8, 2008, at 2:07 AM, Dan Stowell wrote:
> Hi -
>
> I've never used Ambisonics before, but I'm wondering about how
> possible it would be to implement the following setup using SC's
> Ambisonics stuff:
>
> I have 20 or 30 point sources, each of which has a location in the x-y
> plane (and may move about in this plane). I'd like to encode them into
> an Ambisonic representation, then decode this to a ring-of-8-speakers.
> Then for convenience I'd also like to be able to project this
> ring-of-8 down to stereo headphones using a HRTF.
>
> Looking back over the mailing list archives it looks like this is
> probably possible - looks like I could use BFEncode2 to encode the
> sources, then BFDecode1 to decode onto the speaker ring. But I have
> some questions:
>
> * How do I turn many point sources into one single B-format feed? Do I
> encode each one separately and then just add the feeds together? (Or
> multiply the feeds together? Or...)
as said - summing is the way to go. As a note, summing and scaling
(with the same scale over all the channels) is pretty much the only
'basic' thing you should do to the Ambisonic signal. e.g., no
filtering or reverbs. If you are going to do anything after the source
is encoded, it should be a transform (such as Rotate, Tilt or Tumble)
or you should first turn the signal into A-format (with B2A, or back
with A2B). Transforms generally take 4 signals, and return 4 (or, for
higher orders, 9, etc.). A-format signals can have filters and reverbs
applied to them. Convert back to B, and you have the localization
encoded again.
>
>
> * In terms of the audio outcome, is the only difference between the
> different "orders" of Ambisonics the sound quality?
As said before, better localization, and perhaps a wider 'sweet spot'.
>
>
> * For projecting the ring down to headphones, what is needed? In the
> archive there's discussion of downloading the Kemar HRTFs from MIT -
> fine. But is there a specific UGen that would take my 8 channels and
> apply the HRTFs? Or otherwise how would it work?
The ambIEM Kemar stuff does work well. It doesn't work easily with my
UGens though. Since ambIEM is all class driven, it is easier to pass
information between the classes (order, etc.). I need to figure out
how to make my UGens a little friendlier in that regard. The main
advantage in my UGens are there are ways to decode for 'any' rig
(including non symmetrical ones), and ways to compensate for these
differences. This may be where the discussion of delays came in (if
you have an 8 channel rectangle, with delays and scaling of specific
speakers, you can make it sound like an equilateral octagon). I think
the Convolutions and Kemat files could be used still with my UGens,
but the ambIEM stuff has it all ready to go. In general, I don't think
the two systems are really inter-changeable, since my encoding and
decoding methods are different then ambIEM (I use a variable scaler on
'W' as mentioned on: http://www.york.ac.uk/inst/mustech/3d_audio/ambis2.htm
where I don't think ambIEM does).
>
I often use B2Ster for stereo mixes for headphones. It is missing the
HRTF data, but is actually a pretty nice stereo decode. UHJ is also
available in my lib, but I actually don't like how it sounds (a bit
phasey). But it can be a handy way to 'transport' BF data without
height.
Hope that helps,.
Josh
>
> I'd be grateful for any tips
> Dan
> _______________________________________________
> sc-users mailing list
> sc-users@...
> http://lists.create.ucsb.edu/mailman/listinfo/sc-users
******************************************
/* Joshua D. Parmenter
http://www.realizedsound.net/josh/
“Every composer – at all times and in all cases – gives his own
interpretation of how modern society is structured: whether actively
or passively, consciously or unconsciously, he makes choices in this
regard. He may be conservative or he may subject himself to continual
renewal; or he may strive for a revolutionary, historical or social
palingenesis." - Luigi Nono
*/
_______________________________________________
sc-users mailing list
sc-users@...
http://lists.create.ucsb.edu/mailman/listinfo/sc-users
On Mar 8, 2008, at 2:07 AM, Dan Stowell wrote:
> Hi -
>
> I've never used Ambisonics before, but I'm wondering about how
> possible it would be to implement the following setup using SC's
> Ambisonics stuff:
>
> I have 20 or 30 point sources, each of which has a location in the x-y
> plane (and may move about in this plane). I'd like to encode them into
> an Ambisonic representation, then decode this to a ring-of-8-speakers.
> Then for convenience I'd also like to be able to project this
> ring-of-8 down to stereo headphones using a HRTF.
>
> Looking back over the mailing list archives it looks like this is
> probably possible - looks like I could use BFEncode2 to encode the
> sources, then BFDecode1 to decode onto the speaker ring. But I have
> some questions:
>
> * How do I turn many point sources into one single B-format feed? Do I
> encode each one separately and then just add the feeds together? (Or
> multiply the feeds together? Or...)
as said - summing is the way to go. As a note, summing and scaling
(with the same scale over all the channels) is pretty much the only
'basic' thing you should do to the Ambisonic signal. e.g., no
filtering or reverbs. If you are going to do anything after the source
is encoded, it should be a transform (such as Rotate, Tilt or Tumble)
or you should first turn the signal into A-format (with B2A, or back
with A2B). Transforms generally take 4 signals, and return 4 (or, for
higher orders, 9, etc.). A-format signals can have filters and reverbs
applied to them. Convert back to B, and you have the localization
encoded again.
>
>
> * In terms of the audio outcome, is the only difference between the
> different "orders" of Ambisonics the sound quality?
As said before, better localization, and perhaps a wider 'sweet spot'.
>
>
> * For projecting the ring down to headphones, what is needed? In the
> archive there's discussion of downloading the Kemar HRTFs from MIT -
> fine. But is there a specific UGen that would take my 8 channels and
> apply the HRTFs? Or otherwise how would it work?
The ambIEM Kemar stuff does work well. It doesn't work easily with my
UGens though. Since ambIEM is all class driven, it is easier to pass
information between the classes (order, etc.). I need to figure out
how to make my UGens a little friendlier in that regard. The main
advantage in my UGens are there are ways to decode for 'any' rig
(including non symmetrical ones), and ways to compensate for these
differences. This may be where the discussion of delays came in (if
you have an 8 channel rectangle, with delays and scaling of specific
speakers, you can make it sound like an equilateral octagon). I think
the Convolutions and Kemat files could be used still with my UGens,
but the ambIEM stuff has it all ready to go. In general, I don't think
the two systems are really inter-changeable, since my encoding and
decoding methods are different then ambIEM (I use a variable scaler on
'W' as mentioned on: http://www.york.ac.uk/inst/mustech/3d_audio/ambis2.htm
where I don't think ambIEM does).
>
I often use B2Ster for stereo mixes for headphones. It is missing the
HRTF data, but is actually a pretty nice stereo decode. UHJ is also
available in my lib, but I actually don't like how it sounds (a bit
phasey). But it can be a handy way to 'transport' BF data without
height.
Hope that helps,.
Josh
>
> I'd be grateful for any tips
> Dan
> _______________________________________________
> sc-users mailing list
> sc-users@...
> http://lists.create.ucsb.edu/mailman/listinfo/sc-users
******************************************
/* Joshua D. Parmenter
http://www.realizedsound.net/josh/
“Every composer – at all times and in all cases – gives his own
interpretation of how modern society is structured: whether actively
or passively, consciously or unconsciously, he makes choices in this
regard. He may be conservative or he may subject himself to continual
renewal; or he may strive for a revolutionary, historical or social
palingenesis." - Luigi Nono
*/
_______________________________________________
sc-users mailing list
sc-users@...
http://lists.create.ucsb.edu/mailman/listinfo/sc-users
« Return to Thread: Some Ambisonics questions
| Free embeddable forum powered by Nabble | Forum Help |
