I would like to explore the feasibility of an LSL feature which allows emission of sound at an x,y,z position remote from an originating scripted source, such as :
llTriggerSoundPos(vector pos, float volume, float falloff);
'pos' would be region co-ordinates (though it could possibly obey the same situational co-ordinate frame rules for linksets as llSetPos)
The idea is to be able to cast sound emission locations away from the scripted source at defined x,y,z points with a defined volume, and (preferably) a customisable falloff to aid situational tuning.
Definition by region position is of particular interest in creating scripted solutions for region and area soundscapes from a centralised scripted orchestrator, rather than positioning a large number of fixed boxes. This opens a wide potential for automation of moving sources, sensor targeted asset-defined sound locations and reducing investment in LI budget for individual scripted sound sources.
Notes :
(1)To avoid any griefing possibilities, perhaps a limitation could be applied where sound will only emit at non-zero volume where the script owner has rez rights.
(2) Although similar results might perhaps be achieved by moving the source dynamically with llSetRegionPos, this concept would be far more robust and elegant.