The invention relates to a method for optically detecting at least one entity which is arranged on a substrate. The at least one entity is scanned with a measuring volume using at least one radiation source and a confocal optic. During a scanning process an auxiliary focus is generated by means of at least one second radiation source and a second optic. Radiation generated by the first radiation source is collimated by a first optic and radiation generated by the second radiation source is collimated by a second optic. A retroreflection from the auxiliary focus is detected by at least one detector and is used to measuring the position of an interface and, thus, for indirectly positioning the measuring volume. The position of the auxiliary focus relative to the measuring volume is adjustable in a defined manner.