make definition for `audinterface.Process.sampling_rate` more explicit

[reichelu]

In the docs it says Sampling rate in Hz. One could evtl. make clear that this is not the sampling rate to be expected from the input signal y e.g. in audinterface.Process.process_signal(y, ...) but the sampling rate that is required by the underlying process (feature extractor or model). Indirectly this becomes clear, since for .process_signal() one separately needs to specify the sampling rate. Nevertheless, this difference could be made more explicit simply by replacing Sampling rate in Hz with Required sampling rate in Hz.