The light source in the PDS is a conventional 150 Watt halogen lamp. A photodiode is placed in the beam used to measure the plate transmission, and its output is used to servo the lamp intensity. The beam is then passed through an aperture and an image of this is formed on the emulsion surface of the photographic plate. By varying the shape of the aperture and the focal length of the eyepiece used, a range of shapes and size of the measuring spot can be selected. An identical set of apertures and eyepieces are then employed to refocus the spot onto the photocathode of a photomultiplier. With this arrangement it is possible to eliminate any scattered light from reaching the photomultiplier. This makes it possible to achieve a dynamic range in excess of 4 units of optical density. This arrangement is possible as it is the photographic plate which is moved, the measuring beam being fixed. The output of the photomultiplier is then digitised to 16 bits in an analogue-to-digital converter(ADC). To distribute the ADC bits uniformly in density space, the photomultiplier signal passes through a logarithmic amplifier before being digitised. The log amp displays good linearity over 6 decades, but does impose restrictions on the bandwidth. For small signals, corresponding to the central region of astronomical objects, it is only of the order of a few kHz.