3 The method and data sources

The basis of this comparison is the use of these instruments as redshift engines. In this context, a suitable figure of merit is their speed - defined as the number of galaxies measured (at a given magnitude and resolution) per unit of observing time.

Therefore, to compare the relative speed of the designs, two characteristics need to be studied in some detail:

• the relative multiplex gain, and
• the relative exposure time for a given source magnitude and s/n

In order to restrict the large number of parameters we will consider such a comparison for R=150 although we will comment on the effect of varying this choice.

It is assumed that MOS reaches this resolution with the 300 l/mm grating and a slit width of 0.2 arcsec (ie, two lines of micro-mirrors).