Marigold Maripol

trying to figure something out for audio spatialization, not exactly sure how i would go about it. maybe someone here can help.

i have a stereo/two-channel input that i'm expanding into 8-channels.

the stereo is turned into a mono mix through rand~ crossfading. i'm using separate rand~s to modulate this, 8 times.

what i'd like to do, though - is randomly move the stereo pair, four times , so that each L channel had a complimentary output with the R channel.

the distance between L and R would be randomly variable for each.

attached:
i have a stereo/two-channel input that i'm expanding into 8-channels.
the stereo is turned into a mono mix through rand~ crossfading.  i'm using separate rand~s to modulate this, 8 times. 

what i'd like to do, though - is randomly move the stereo pair, four times , so that each L channel had a complimentary output with the R channel.
the distance between L and R would be randomly variable for each. 

a-spatial-reasoning-problem

Practically anything is possible... you just have to figure out how.

I am not sure I entirely understand your question, but I thought I could give you a few pointers. Consider pasting a patch you are working on for more specific help.

3D spatialization of sound can be achieved simply with mono panning as you describe, especially with a large number of speakers configured properly, but a more realistic spatialization can be achieved with fewer speakers by creating a system that interpolates weighted values for multiple adjacent speakers. Ideally you want the spatialization to be achieved by a combination of 3 different levels in adjacent speakers. There is an awesome object in Max that takes the math out of this. Its called the "nodes" object. You set the position of each node to correspond with the position of the speaker, and the size of each node to be the weighting you want that speaker to have in the nodal system. Generally, for sound you want large overlapping nodes that will give you 3 levels from adjacent speakers, but exclude levels from speakers at the far side of the room. Positioning of the slider in the nodes object can be automated any number of ways using pattr, line, or coll, or the automove subpatch included in the help file.

For Stereo files, you could have two nodes, and tie together and offset the position of two sliders in the nodes or give different weightings to the speakers. You could vary the distance between the two sliders to adjust the stereo spread.

Practically anything is possible... you just have to figure out how.  

I am not sure I entirely understand your question, but I thought I could give you a few pointers.  Consider pasting a patch you are working on for more specific help.  

3D spatialization of sound can be achieved simply with mono panning as you describe, especially with a large number of speakers configured properly, but a more realistic spatialization can be achieved with fewer speakers by  creating a system that interpolates weighted values for multiple adjacent speakers.  Ideally you want the spatialization to be achieved by a combination of 3 different levels in adjacent speakers.  There is an awesome object in Max that takes the math out of this.  Its called the "nodes" object.  You set the position of each node to correspond with the position of the speaker, and the size of each node to be the weighting you want that speaker to have in the nodal system.  Generally, for sound you want large overlapping nodes that will give you 3 levels from adjacent speakers, but exclude levels from speakers at the far side of the room.  Positioning of the slider in the nodes object can be automated any number of ways using pattr, line, or coll, or the automove subpatch included in the help file.

For Stereo files, you could have two nodes, and tie together and offset the position of two sliders in the nodes or give different weightings to the speakers.  You could vary the distance between the two sliders to adjust the stereo spread.


a spatial reasoning problem