Some libraries have samples with a few milliseconds of silence before the attack. For short, accented articulations (e.g., staccato or marcato), this can cause them not to play exactly on the beat.

To compensate, KSEM can trigger the notes of these articulations a few milliseconds earlier.

Note: Make sure Delay Compensation is enabled to automatically offset the resulting latency, or adjust it manually in the Track Delay section.