A scanned optical beam is divided so as to form a set of scanned subbeams. To compensate for scan errors, a portion of at least one subbeam is detected and a scan error estimated based on the detected portion. A beam scanner is controlled according to the estimated error so as to adjust a propagation direction of some or all of the set of scanned subbeams. The scanned subbeams with adjusted propagation directions are received by an f-theta lens and directed to a work piece. In typical examples, the portion of the at least one subbeam that is detected is obtained from the set of scanned subbeams prior to incidence of the scanned subbeams to the f-theta lens.