|
In simplest terms and to directly mimic an analog sum, it is the simultaneous add of multiple digital input data streams, into a single digital output stream.
I would expect differences between floating point and fixed point math implementations, and perhaps time offsets due to serial vs. parallel organization of much digital hardware. Processing complexity can increase if sample rate conversions and fader/pan multiplies are involved. The gain tweak will be fast (in digital clock time) but the conversion may get messy.
In theory such a sum should be loss-less and arbitrarily accurate, even better than analog. In practice, results and opinions vary.
JR
|