i noticed i use single math operator objects all the time instead of combining them long chains in an expr object. in a number of big patches of me this methode would reduce the number of objects and patchcords. will this also save some cpu cycles and/or some bytes in memory? or is this nothing worth to think about with todays computers?
Probably nothing to worry about as far as CPU is concerned. There’s a (small) price for every object and patch cord, but the last time I saw source code for how [expr] worked, there was some internal overhead for every operation as well.
Barry’s point about tidying up patches is well taken, but you can also encapsulate.
One other thing might be of interest: [expr] may have higher precision for intermediate calculations, compared to sending data through patch cords. Someone who’s actually looking at the source code once in a while would need to verify that hypothesis, though.