ALL of us play chords slightly 'late' all the time (no-one is THAT precise!). But the arranger does some VERY clever tricks to fool us that nothing wrong happened.

Just imagine the case where you played as close to 'on the beat' you possibly could, but ended up playing one or two 'ticks' late... Completely indistinguishable to the ear, but to a computer, definitely late. What happens next varies from manufacturer to manufacturer, but essentially, the old 'incorrect' notes are cut off (derived from the previous chord) somehow, new notes are generated, and pitch bent or portamento'd (with a time of zero) to the new 'correct' notes. All of this with as little 'glitching' as possible.

Now , we've ALL heard those little 'glitches' from time to time, especially on stripped down styles, and when we are just listening to the arranger, not playing on top of it (which tends to mask the problem). But HOW each arranger does this is a bit of a mystery, and obviously, if one manufacturer's arranger uses a different system to another, link them together and the slave module may not use the same codes that the master does to avoid the glitches.

The result...? Usually, dropped notes at the chord changes, or weird little glitches, or sometimes wrong notes. Drum tracks don't suffer from this, because they never transpose as you change chords, but all pitched tones do... So, hence my thread about trying to find out how each manufacturer deals with this problem, so we can compile a list of what works (and doesn't work) with what...

I hope this explains it better...
_________________________
An arranger is just a tool. What matters is what you build with it..!