Thanks to both you and Michael for the "thumbs up" about Ambisonics.

The opcode docs don't show recommended or default values for the various
input parameters when encoding. Do you have any? (I'm just looking for
reasonable starting points.)

Art Hunkins

----- Original Message ----- 
From: "David Ogborn" 
To: "Art Hunkins" ; "Developer discussions"

Sent: Saturday, October 28, 2006 12:46 AM
Subject: Re: [Cs-dev] [Csnd] MIDI Pan Curve


> Absolutely - I routinely do live electronic pieces with Ambisonics
> functioning as an "abstraction layer" for the spatialization, decoding
> to 2, 4, 6, or 8 channels depending on the gig.  The decoding math is
> just a multiplication or two and add per channel.  The encoding math is
> not much more involved.  Very cheap operations.
>
> Yours truly,
> David
>
> Art Hunkins wrote:
> > Any idea whether an ambisonic B encode/decode for stereo or quad could
be
> > done in Csound in *real time*?
> >
> > Art Hunkins
> >
> > ----- Original Message ----- 
> > From: "Victor Lazzarini" 
> > To: "Developer discussions" ; "Art
> > Hunkins" 
> > Sent: Friday, October 27, 2006 7:17 PM
> > Subject: Re: [Csnd] MIDI Pan Curve
> >
> >
> > I'd say ambisonics. Even stereo reductions are
> > quite impressive.
> >
> >> A general question about pan opcodes (including UDO's):
> >>
> >> Which of the available options give the clearest spatial
> >> localization?
> >>
> >> I understand that many acoustical factors are involved in
> >> achieving two- or three-dimensional imaging, but would
> >> like advice as to which Csound opcodes create the *most
> >> precise* positioning in the stereo (or quad) field.
> >>
> >> I've always been disappointed that my attempts to position
> >> sounds have largely fallen flat, and not been perceptually
> >> apparent in any very specific sense. (I've used the pan
> >> opcode.) In a nutshell, I'm looking for Csound code that
> >> can better project perceptual space.
> >>
> >> Thanks for sharing your experience.
> >>
> >> Art Hunkins
> >>
> >> ----- Original Message -----
> >> From: "Steven Yi" 
> >> To: 
> >> Sent: Friday, October 27, 2006 3:27 PM
> >> Subject: Re: [Csnd] MIDI Pan Curve
> >>
> >>
> >> Hi All,
> >>
> >> Just wanted to note that I did a bug fix of this panning
> >> UDO today (hard left gave NaN for right channel signal)
> >> and have renamed it to pan_gm2 as it's a recommendation
> >> for General Midi 2 implementors.  If you've downloaded
> >> this UDO please try downloading again here:
> >>
> >>
> > http://www.csounds.com/udo/displayOpcode.php?opcode_id=90
> >> Thanks!
> >> steven
> >>
> >>
> >> On 10/26/06, Steven Yi  wrote:
> >>> Hi All,
> >>>
> >>> I came across a document today called "Default Pan
> >> Curve" on the MIDI website:
> >>> http://www.midi.org/about-midi/specshome.shtml
> >>>
> >>> and have implemented it as a UDO here:
> >>>
> >>>
> > http://www.csounds.com/udo/displayOpcode.php?opcode_id=90
> >>>> It's described in the spec as:
> >>> Over the course of the development of MIDI, different
> >>> manufacturers have implemented
> >>> their own responses to CC#10. In addition, various
> >>> AMEI/MMA Specification (such as GM2 and
> >>> DLS2) have recommended response curves that are not all
> >>> the same. The different pan curves produce
> >>> an equal power distribution between the left and right
> >>> channels, but in some cases, a true center could
> >>> not be achieved by following the recommended equation
> >>> directly. The formula in this RP does achieve a
> >>> true center and uses an equal power distribution, which
> >> is desirable. >
> >>> The formula given is:
> >>>
> >>> Left Channel Gain [dB] = 20*log (cos (Pi/2*
> > max(0,CC#10
> >>> - 1)/126) Right Channel Gain [dB] = 20*log (sin (Pi
> > /2*
> >> max(0,CC#10 - 1)/126) >
> >>> I realized just now that I've rescaled the values to be
> >>> between -1 and 1 which is what I'm accustomed to, but
> >>> then naming it pan_midi doesn't make that much sense
> >>> then.  Would anyone have a recommendation on what to
> >> call this?  And should a pan_midi which takes in 0-127 be
> >> added? >
> >>> Thanks!
> >>> steven
> >>>
> >> --
> >> Send bugs reports to this list.
> >> To unsubscribe, send email to
> >> csound-unsubscribe@lists.bath.ac.uk
> >>
> >> --
> >> Send bugs reports to this list.
> >> To unsubscribe, send email to
> >> csound-unsubscribe@lists.bath.ac.uk
> >
> >
>
> -------------------------------------------------------------------------
> > Using Tomcat but need to do more? Need to support web services,
security?
> > Get stuff done quickly with pre-integrated technology to make your job
easier
> > Download IBM WebSphere Application Server v.1.0.1 based on Apache
Geronimo
> > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> > _______________________________________________
> > Csound-devel mailing list
> > Csound-devel@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/csound-devel
>
> -------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job
easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel


-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

Hi Art,

I don't use the "built-in" opcodes for Ambisonics encoding/decoding for
two/three reasons:

1a) Both for my own learning, and for teaching others, I'd rather rely
on building and understanding things out of a small reliable, core set
of opcodes.

1b) If Csound is ever optimized for SSE2/3/3DNOW/etc-type machine
instructions, those core opcodes will surely be the most likely
candidates for optimization, and so things built from them rather than
coded in C will benefit immediately.

2)  "Special situations" arise almost as a rule, particularly for
decoding.  Even with encoding, I find it useful to create opcodes fit to
the requirements of a particular piece or situation within a piece.

Here is an example of a horizontal-only encoding (to the edge of the
unit circle) opcode.  If the k-rate azimuth argument (range 0 to 2pi; 0
is front and centre, pi is back and centre) changes quickly, the
parallel interpolation of the sin/cos functions will introduce momentary
spatial distortions - but then to move things quickly around the room is
already a kind of "distortion":

; usage: aw,ax,ay spatialize aMySignal, kposition
opcode spatialize,aaa,ak
ain,kazimuth xin
acos interp cos(kazimuth)
asin interp sin(kazimuth)
xout ain*0.707,ain*acos,ain*asin
endop

And here is an example of a decoding instrument that takes
horizontal-only input (W,X,Y) and decodes to a rig consisting of a
certain arrangement of six speakers plus two (identical) subwoofer feeds:

instr 1
aw inch 1
ax inch 2
ay inch 3
ax2 = ax * 0.707
ay2 = ay * 0.707
a1 = aw + ax
a2 = aw + ax2 - ay2
a3 = aw - ay
a5 = aw - ax
a7 = aw + ay
a8 = aw + ax2 + ay2
as1 butterlp aw, 100
as2 butterlp as1, 100
as3 butterlp as2, 100
asub balance as3,aw
outc a1,a2,a3,asub,a5,asub,a7,a8
endin

Hope that helps - I don't mind continuing to chat about this off-list
also...

Yours truly,
David


Art Hunkins wrote:
> Thanks to both you and Michael for the "thumbs up" about Ambisonics.
> 
> The opcode docs don't show recommended or default values for the various
> input parameters when encoding. Do you have any? (I'm just looking for
> reasonable starting points.)
> 
> Art Hunkins
> 
> ----- Original Message ----- 
> From: "David Ogborn" 
> To: "Art Hunkins" ; "Developer discussions"
> 
> Sent: Saturday, October 28, 2006 12:46 AM
> Subject: Re: [Cs-dev] [Csnd] MIDI Pan Curve
> 
> 
>> Absolutely - I routinely do live electronic pieces with Ambisonics
>> functioning as an "abstraction layer" for the spatialization, decoding
>> to 2, 4, 6, or 8 channels depending on the gig.  The decoding math is
>> just a multiplication or two and add per channel.  The encoding math is
>> not much more involved.  Very cheap operations.
>>
>> Yours truly,
>> David
>>
>> Art Hunkins wrote:
>>> Any idea whether an ambisonic B encode/decode for stereo or quad could
> be
>>> done in Csound in *real time*?
>>>
>>> Art Hunkins
>>>
>>> ----- Original Message ----- 
>>> From: "Victor Lazzarini" 
>>> To: "Developer discussions" ; "Art
>>> Hunkins" 
>>> Sent: Friday, October 27, 2006 7:17 PM
>>> Subject: Re: [Csnd] MIDI Pan Curve
>>>
>>>
>>> I'd say ambisonics. Even stereo reductions are
>>> quite impressive.
>>>
>>>> A general question about pan opcodes (including UDO's):
>>>>
>>>> Which of the available options give the clearest spatial
>>>> localization?
>>>>
>>>> I understand that many acoustical factors are involved in
>>>> achieving two- or three-dimensional imaging, but would
>>>> like advice as to which Csound opcodes create the *most
>>>> precise* positioning in the stereo (or quad) field.
>>>>
>>>> I've always been disappointed that my attempts to position
>>>> sounds have largely fallen flat, and not been perceptually
>>>> apparent in any very specific sense. (I've used the pan
>>>> opcode.) In a nutshell, I'm looking for Csound code that
>>>> can better project perceptual space.
>>>>
>>>> Thanks for sharing your experience.
>>>>
>>>> Art Hunkins
>>>>
>>>> ----- Original Message -----
>>>> From: "Steven Yi" 
>>>> To: 
>>>> Sent: Friday, October 27, 2006 3:27 PM
>>>> Subject: Re: [Csnd] MIDI Pan Curve
>>>>
>>>>
>>>> Hi All,
>>>>
>>>> Just wanted to note that I did a bug fix of this panning
>>>> UDO today (hard left gave NaN for right channel signal)
>>>> and have renamed it to pan_gm2 as it's a recommendation
>>>> for General Midi 2 implementors.  If you've downloaded
>>>> this UDO please try downloading again here:
>>>>
>>>>
>>> http://www.csounds.com/udo/displayOpcode.php?opcode_id=90
>>>> Thanks!
>>>> steven
>>>>
>>>>
>>>> On 10/26/06, Steven Yi  wrote:
>>>>> Hi All,
>>>>>
>>>>> I came across a document today called "Default Pan
>>>> Curve" on the MIDI website:
>>>>> http://www.midi.org/about-midi/specshome.shtml
>>>>>
>>>>> and have implemented it as a UDO here:
>>>>>
>>>>>
>>> http://www.csounds.com/udo/displayOpcode.php?opcode_id=90
>>>>>> It's described in the spec as:
>>>>> Over the course of the development of MIDI, different
>>>>> manufacturers have implemented
>>>>> their own responses to CC#10. In addition, various
>>>>> AMEI/MMA Specification (such as GM2 and
>>>>> DLS2) have recommended response curves that are not all
>>>>> the same. The different pan curves produce
>>>>> an equal power distribution between the left and right
>>>>> channels, but in some cases, a true center could
>>>>> not be achieved by following the recommended equation
>>>>> directly. The formula in this RP does achieve a
>>>>> true center and uses an equal power distribution, which
>>>> is desirable. >
>>>>> The formula given is:
>>>>>
>>>>> Left Channel Gain [dB] = 20*log (cos (Pi/2*
>>> max(0,CC#10
>>>>> - 1)/126) Right Channel Gain [dB] = 20*log (sin (Pi
>>> /2*
>>>> max(0,CC#10 - 1)/126) >
>>>>> I realized just now that I've rescaled the values to be
>>>>> between -1 and 1 which is what I'm accustomed to, but
>>>>> then naming it pan_midi doesn't make that much sense
>>>>> then.  Would anyone have a recommendation on what to
>>>> call this?  And should a pan_midi which takes in 0-127 be
>>>> added? >
>>>>> Thanks!
>>>>> steven
>>>>>
>>>> --
>>>> Send bugs reports to this list.
>>>> To unsubscribe, send email to
>>>> csound-unsubscribe@lists.bath.ac.uk
>>>>
>>>> --
>>>> Send bugs reports to this list.
>>>> To unsubscribe, send email to
>>>> csound-unsubscribe@lists.bath.ac.uk
>>>
>> -------------------------------------------------------------------------
>>> Using Tomcat but need to do more? Need to support web services,
> security?
>>> Get stuff done quickly with pre-integrated technology to make your job
> easier
>>> Download IBM WebSphere Application Server v.1.0.1 based on Apache
> Geronimo
>>> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
>>> _______________________________________________
>>> Csound-devel mailing list
>>> Csound-devel@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>> -------------------------------------------------------------------------
>> Using Tomcat but need to do more? Need to support web services, security?
>> Get stuff done quickly with pre-integrated technology to make your job
> easier
>> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
>> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
> 
> 
> -------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

>>>>> "Steven" == Steven Yi  writes:

 Steven> Hi Art,
 Steven> You might want to look at Jan Jacob Hoffman's website for information
 Steven> on Ambisonics. The information there is very well written and he has
 Steven> Csound code there.  His music is also fantastic and you can hear what
 Steven> an ambisonics reduction to 2 channel would sound like.

 Steven> http://www.sonicarchitecture.de/

just to endorse this; Jan Jacob Hoffman's use of spacial sound is
excellent and well worth the effort to hear.  Ambisonic performances
are even better!

==John ffitch

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net