Csound Csound-dev Csound-tekno Search About

[Cs-dev] bformdec / bformenc angles & manual

Date2008-02-19 14:50
FromVictor Lazzarini
Subject[Cs-dev] bformdec / bformenc angles & manual
As far as I can see, the expected format for angles in
bformenc  is counterclockwise, starting from the right (0),
as the X direction is cos(A)cos(E)  and Y is sin(A)cos(E).

If  we say hard right, then cos(0) = 1 and sin(0) = 1 looks
the right coefficients to me, for the X and Y directions. Then
front cos(90) = 0 and sin(90) = 1 also do.

The manual says angles are *clockwise*, but as far as I can see
the code uses the conventions above. I expect the manual to
be wrong. Can anyone confirm/correct this?

Now, as far as bformdec is concerned, I see a similar thing. The
coefficients for say, 8-channel ring are for angles (described
correctly in the manual) increasing counterclockwise from the
right. But as the manual does not say anything, we are left wondering.
So can this be confirmed and the manual updated?

Also, have these opcodes been fixed? I have fixed them locally,
but did not commit, as they are not my code. Should I do so?

Thanks

Victor



Victor Lazzarini
Music Technology Laboratory
Music Department
National University of Ireland, Maynooth


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

Date2008-02-19 16:08
FromAnthony Kozar
SubjectRe: [Cs-dev] bformdec / bformenc angles & manual
Counterclockwise from the positive X axis is certainly standard so I would
just change the manual.

You could try contacting Samuel Groner as I do not think he is on this list.
I will send you his email address privately.

Anthony

Victor Lazzarini wrote on 2/19/08 9:50 AM:

> As far as I can see, the expected format for angles in
> bformenc  is counterclockwise, starting from the right (0),
> as the X direction is cos(A)cos(E)  and Y is sin(A)cos(E).
> [...]
> The manual says angles are *clockwise*, but as far as I can see
> the code uses the conventions above. I expect the manual to
> be wrong. Can anyone confirm/correct this?
> 
> Now, as far as bformdec is concerned, I see a similar thing. The
> coefficients for say, 8-channel ring are for angles (described
> correctly in the manual) increasing counterclockwise from the
> right. But as the manual does not say anything, we are left wondering.
> So can this be confirmed and the manual updated?
> 
> Also, have these opcodes been fixed? I have fixed them locally,
> but did not commit, as they are not my code. Should I do so?


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

Date2008-02-20 11:53
FromJan Jacob Hofmann
SubjectRe: [Cs-dev] bformdec / bformenc angles & manual
Please also note that in Ambisonic convention positive x (x =1 in cartesian
coordinates) is infront and positive y (y =1 in cartesian coordinates) is
90° left to the listener.  The counting according to that convention goes
anti-clockwise.  Usually sounds are encoded that way for ambisonic and as
far as I can see, the source-code for bformenc works that way too. It is a
matter of definition where to put x= 0° in space, though. The maths of the
code works the same, no matter where place x= 0° in real space.

Sound has to be decoded according to the same convention as the encoding was
done. As one does not sometimes know, according to wich convention the sound
has been encoded, if there are several flavors, decoding should follow the
usual convention.  

The manual for beformenc does not follow this usual convention. 
For beformdec the stereo-setup decode is a good example to see that counting
is starting infront of the listener with 0°. Counting there should be anti-
clockwise, to align with the general convention. In the manual there is no
trace, if counting is clockwise or anti-clockwise, though it has to align
with the way, the position of sound was encoded.

I wonder if the bformdec source is independent of any convention, as long
the decode is consistent with the way the sound has been encoded. I'll leave
this proof to a mathematician. 

In practice, besides the above,  the bformdec-opcode is not usable for me:
The employed "Spherical Harmonic" decoding equation works well by definition
only for a large number of speakers, which might only be  given for the
8-chnl - circle setup. John ffitch has added the "Controlled opposites"
equations as far as I can remember. This is useful for the square setup, but
fails for asymmetric speaker placement as 5.1 and the asymmetric and
distorted cube given in the decoding presets of bformdec.

see

http://www.york.ac.uk/inst/mustech/3d_audio/secondor.html
and
http://www.muse.demon.co.uk/ref/speakers.html



Jan Jacob






Anthony Kozar-2 wrote:
> 
> Counterclockwise from the positive X axis is certainly standard so I would
> just change the manual.
> 
> You could try contacting Samuel Groner as I do not think he is on this
> list.
> I will send you his email address privately.
> 
> Anthony
> 
> Victor Lazzarini wrote on 2/19/08 9:50 AM:
> 
>> As far as I can see, the expected format for angles in
>> bformenc  is counterclockwise, starting from the right (0),
>> as the X direction is cos(A)cos(E)  and Y is sin(A)cos(E).
>> [...]
>> The manual says angles are *clockwise*, but as far as I can see
>> the code uses the conventions above. I expect the manual to
>> be wrong. Can anyone confirm/correct this?
>> 
>> Now, as far as bformdec is concerned, I see a similar thing. The
>> coefficients for say, 8-channel ring are for angles (described
>> correctly in the manual) increasing counterclockwise from the
>> right. But as the manual does not say anything, we are left wondering.
>> So can this be confirmed and the manual updated?
>> 
>> Also, have these opcodes been fixed? I have fixed them locally,
>> but did not commit, as they are not my code. Should I do so?
> 
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel
> 
> 

-- 
View this message in context: http://www.nabble.com/bformdec---bformenc-angles---manual-tp15568299p15587389.html
Sent from the Csound - Dev mailing list archive at Nabble.com.


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net
h