An encoder apparatus comprising a scale comprising scale features in at least one track which define a series of incremental scale marks and at least one reference mark and a readhead comprising a light source, a diffraction grating, an incremental photodetector and a reference photodetector comprising at least two detector channels. The features defining the at least one reference mark are configured such that the at least one reference mark as resolvable by each of the at least two detector channels comprises a single feature. The optical arrangement of the encoder is such that a non-imaged representation of at least the track in which the at least one reference mark is contained falls onto the detector.