The application of multi-dimensional analysis techniques to the processing of event data from large spectrometers

In order to be able to handle the expected event data-taking rates of the near future, techniques capable of rapid data reduction and analysis will be required. The conventional approach to track recognition and geometrical reconstruction would require unrealistically large amounts of computer time, even with the fastest computers available at present. In addition, future detector layouts that attempt to minimise construction cost and maximise data rates, might not be well suited to conventional track following and geometrical fitting techniques. In this paper, we shall consider the usefulness of multi-dimensional analysis techniques for fast data processing. Their application to some existing large spectrometers will be described, as well as the possible constraints imposed on the detectors and detector layouts of the future. In this, therefore, multi-dimensional analysis can be used to guide the hardware in a rather general way, rather than simply expecting the software to solve all problems imposed by the hardware often at great cost in computer time.