Hello

Now I have sort of a request please. (Although it may not be viable.)

The surround panner in protools has a nice feature that allows you to control the 'leakage' of the signal into different speakers. Basically it's similar to the 'spread' variable of vbap but there are three of them: one for front/rear, one for front left/right, and one for rear left/right 'divergence'. I've attached a screenshot: you can see the rear left/right (78%) and front/rear (86%) divergences are graphically reflected in the picture by the purple frame.  

Is it possible to have something like this in Csound? Speakers are rarely actually symmetrically placed and in any case it doesn't really make sense to have the same spread for all directions.

Thanks
Peiman 

 

hi peiman,

so, if i understand correctly, the purple box effectively limits the
range of XY coordinates to leave a greater or lesser spread in
different regions?

i would think that this might be best accomplished outside the vbap
opcode, treating spatialization parameters as coordinates before
converting then to aziumuth&spread.  i would suggest using the `limit'
opcode where the low and high limits are interpolated according to the
x or y positions themselves (depending on whether its F/R, L/R, etc.)

best,
--  ben

On Thu, Aug 2, 2012 at 6:00 AM, peiman khosravi
<peimankhosravi@gmail.com> wrote:
> Sorry, obviously this isn't a spam. I think it must be the attachment.
>
> On 2 August 2012 12:41, peiman khosravi <peimankhosravi@gmail.com> wrote:
>>
>> Hello
>>
>> Now I have sort of a request please. (Although it may not be viable.)
>>
>> The surround panner in protools has a nice feature that allows you to
>> control the 'leakage' of the signal into different speakers. Basically it's
>> similar to the 'spread' variable of vbap but there are three of them: one
>> for front/rear, one for front left/right, and one for rear left/right
>> 'divergence'. I've attached a screenshot: you can see the rear left/right
>> (78%) and front/rear (86%) divergences are graphically reflected in the
>> picture by the purple frame.
>>
>> Is it possible to have something like this in Csound? Speakers are rarely
>> actually symmetrically placed and in any case it doesn't really make sense
>> to have the same spread for all directions.
>>
>> Thanks
>> Peiman
>>
>>
>
>

Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Hi,

I don't think you can do this with the "normal" vbap algorithm. What
vbap does is find the two speakers (or three when in 3D) where the
desired direction is, and then does amplitude panning between them.
So, spread is a concept completely outside the algorithm. You would
have to extend the algorithm so that it looked at the "second"
neighbors and then would assign a smaller amount of the signal to
them. This is not too hard on 2D, but it gets complicated for 3D.

However, don't forget the precedence effect! You will likely not hear
the spread with most signals and directions.

Cheers,
Andrés

On Thu, Aug 2, 2012 at 4:41 AM, peiman khosravi
<peimankhosravi@gmail.com> wrote:
> Hello
>
> Now I have sort of a request please. (Although it may not be viable.)
>
> The surround panner in protools has a nice feature that allows you to
> control the 'leakage' of the signal into different speakers. Basically it's
> similar to the 'spread' variable of vbap but there are three of them: one
> for front/rear, one for front left/right, and one for rear left/right
> 'divergence'. I've attached a screenshot: you can see the rear left/right
> (78%) and front/rear (86%) divergences are graphically reflected in the
> picture by the purple frame.
>
> Is it possible to have something like this in Csound? Speakers are rarely
> actually symmetrically placed and in any case it doesn't really make sense
> to have the same spread for all directions.
>
> Thanks
> Peiman
>
>

Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Hi,

Does ProTools do decorrelation when doing spread?

Cheers,
Andrés

On Sun, Aug 5, 2012 at 7:55 AM, peiman khosravi

<peimankhosravi@gmail.com> wrote:
> Hi Andres,
>
> Thanks, that clarifies it. Infact, right now I'm dreaming about a panner
> similar in to that in protools where the spread is felt by giving the user
> some control over decorrelating the signals. I might try making a non vbap
> UDO panner and see if I can take it from there.
>
> In protools I use the panner with multiple copies of decorrelated files
> (thanks to you) but I'm on a mission to make my tools as I like them to be,
> rather than always hacking. I've also noticed that it's much easier to deal
> with numbers rather than graphic points, changing a number by a fraction can
> have huge perceptual repercussions and it makes sense to deal with the
> numbers to begin with.
>
> Thanks
> Peiman
>
> On 5 August 2012 15:25, Andres Cabrera <mantaraya36@gmail.com> wrote:
>>
>> Hi,
>>
>> I don't think you can do this with the "normal" vbap algorithm. What
>> vbap does is find the two speakers (or three when in 3D) where the
>> desired direction is, and then does amplitude panning between them.
>> So, spread is a concept completely outside the algorithm. You would
>> have to extend the algorithm so that it looked at the "second"
>> neighbors and then would assign a smaller amount of the signal to
>> them. This is not too hard on 2D, but it gets complicated for 3D.
>>
>> However, don't forget the precedence effect! You will likely not hear
>> the spread with most signals and directions.
>>
>> Cheers,
>> Andrés
>>
>> On Thu, Aug 2, 2012 at 4:41 AM, peiman khosravi
>> <peimankhosravi@gmail.com> wrote:
>> > Hello
>> >
>> > Now I have sort of a request please. (Although it may not be viable.)
>> >
>> > The surround panner in protools has a nice feature that allows you to
>> > control the 'leakage' of the signal into different speakers. Basically
>> > it's
>> > similar to the 'spread' variable of vbap but there are three of them:
>> > one
>> > for front/rear, one for front left/right, and one for rear left/right
>> > 'divergence'. I've attached a screenshot: you can see the rear
>> > left/right
>> > (78%) and front/rear (86%) divergences are graphically reflected in the
>> > picture by the purple frame.
>> >
>> > Is it possible to have something like this in Csound? Speakers are
>> > rarely
>> > actually symmetrically placed and in any case it doesn't really make
>> > sense
>> > to have the same spread for all directions.
>> >
>> > Thanks
>> > Peiman
>> >
>> >
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>             https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>> csound"
>>
>


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

as someone who has worked extensively with vbap in my own work for
several years, i have to say that i rarely work with sonic materials
that don't yield perceptual differences with differing amounts of
spread (with the only exception being very periodic signals).  working

a lot in lager spaces, i like using spread to aid audience coverage.

in fact i usually have the opposite problem with precedence -- that
higher spreads often dilute the perceptual acuity of phantom
locations.  as a result i sometimes add different delays to each

output channel such that the channels closest to the virtual source
arrive first.

cheers,
--  ben

On Sun, Aug 5, 2012 at 7:25 AM, Andres Cabrera <mantaraya36@gmail.com> wrote:
> Hi,
>
> I don't think you can do this with the "normal" vbap algorithm. What
> vbap does is find the two speakers (or three when in 3D) where the
> desired direction is, and then does amplitude panning between them.
> So, spread is a concept completely outside the algorithm. You would
> have to extend the algorithm so that it looked at the "second"
> neighbors and then would assign a smaller amount of the signal to
> them. This is not too hard on 2D, but it gets complicated for 3D.
>
> However, don't forget the precedence effect! You will likely not hear
> the spread with most signals and directions.
>
> Cheers,
> Andrés
>
> On Thu, Aug 2, 2012 at 4:41 AM, peiman khosravi
> <peimankhosravi@gmail.com> wrote:
>> Hello
>>
>> Now I have sort of a request please. (Although it may not be viable.)
>>
>> The surround panner in protools has a nice feature that allows you to
>> control the 'leakage' of the signal into different speakers. Basically it's
>> similar to the 'spread' variable of vbap but there are three of them: one
>> for front/rear, one for front left/right, and one for rear left/right
>> 'divergence'. I've attached a screenshot: you can see the rear left/right
>> (78%) and front/rear (86%) divergences are graphically reflected in the
>> picture by the purple frame.
>>
>> Is it possible to have something like this in Csound? Speakers are rarely
>> actually symmetrically placed and in any case it doesn't really make sense
>> to have the same spread for all directions.
>>
>> Thanks
>> Peiman
>>
>>
>
>
> Send bugs reports to the Sourceforge bug tracker
>             https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

as someone who has worked extensively with vbap in my own work for
several years, i have to say that i rarely work with sonic materials
that don't yield perceptual differences with differing amounts of
spread (with the only exception being very periodic signals).  working

Interesting. I think in a real-world context the acoustics of the venue can makes a huge difference, and sometimes working to our advantage. I find that sounds that are prone to reflection, mostly high frequency noise-based textures tend to spread more readily, even without much of a diffusion. Even in a stereo situation some sounds seem to move beyond the stereo field, almost enveloping the listener. It all depends on the sound as well as the space.
 

a lot in lager spaces, i like using spread to aid audience coverage.

I don't know the exact context but surely more speakers would be the obvious solution. Lacking that, I would make use of the acoustics of the room and point speakers in such a way as to increase reflection.    
 

in fact i usually have the opposite problem with precedence -- that
higher spreads often dilute the perceptual acuity of phantom
locations.  as a result i sometimes add different delays to each

Yes. When you consider all the acoustic reflections and phasing artefacts caused by diffusing the sound. It certainly does make a difference, but the point is that it doesn't have the promised effect, which is to increase the perceived width of the acoustic image. Or more precisely it doesn't do it for everyone. 

One easy way to add some decorrelation when diffussion stereo pieces is simply to swap the stereo image in the back. And as you say, adding less than 20ms delay to the back also helps if you're sitting close to the back: it sort of compensates for the precedence effect so you still perceive the sound as being front-orientated. I always do these two things by default when diffusing.      

Best,
Peiman
 

output channel such that the channels closest to the virtual source
arrive first.

cheers,
--  ben

On Sun, Aug 5, 2012 at 7:25 AM, Andres Cabrera <mantaraya36@gmail.com> wrote:
> Hi,
>
> I don't think you can do this with the "normal" vbap algorithm. What
> vbap does is find the two speakers (or three when in 3D) where the
> desired direction is, and then does amplitude panning between them.
> So, spread is a concept completely outside the algorithm. You would
> have to extend the algorithm so that it looked at the "second"
> neighbors and then would assign a smaller amount of the signal to
> them. This is not too hard on 2D, but it gets complicated for 3D.
>
> However, don't forget the precedence effect! You will likely not hear
> the spread with most signals and directions.
>
> Cheers,
> Andrés
>
> On Thu, Aug 2, 2012 at 4:41 AM, peiman khosravi
> <peimankhosravi@gmail.com> wrote:
>> Hello
>>
>> Now I have sort of a request please. (Although it may not be viable.)
>>
>> The surround panner in protools has a nice feature that allows you to
>> control the 'leakage' of the signal into different speakers. Basically it's
>> similar to the 'spread' variable of vbap but there are three of them: one
>> for front/rear, one for front left/right, and one for rear left/right
>> 'divergence'. I've attached a screenshot: you can see the rear left/right
>> (78%) and front/rear (86%) divergences are graphically reflected in the
>> picture by the purple frame.
>>
>> Is it possible to have something like this in Csound? Speakers are rarely
>> actually symmetrically placed and in any case it doesn't really make sense
>> to have the same spread for all directions.
>>
>> Thanks
>> Peiman
>>
>>
>
>
> Send bugs reports to the Sourceforge bug tracker
>             https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"