When running a multi-dimensional NMR experiment you have to decide how many scans and how many slices (or rows, or increments) to use. If your time is limited, as it normally is, increasing one means reducing the other. So which is better more scans or more slices?