In the synchrotron-based facilities (KEK and RAL), muons come in a pulse, with a spread of 50 ns and a pulse-to-pulse interval of 20 ms. Since the muon arrival time (t=0) is known from the timing signal of the synchrotron, SR measurements are performed by taking the time spectra of decay positrons relative to the muon pulse. The timing resolution of this `pulse-SR' method is limited by the muon pulse-width (50 ns), but the experimental time window is virtually infinite (s). The long experimental time window makes this method convenient for measurements of slow muon spin relaxation. The pulse-SR method is also convenient to introduce extreme conditions, such as high-magnetic fields  and optical radiations [31,32], using a pulse magnets/lasers synchronized to the muon pulse.
The cyclotron-based facilities (TRIUMF and PSI) provide a continuous muon beam. As a result, one needs a muon counter on the beam path right before the sample, so that one knows a muon arrival time (t=0). The timing resolution of this `continuous-beam' SR method is theoretically infinitesimally small; with carefully tuned electronics and small counters, sub-nanosecond resolution ( ns) may be achieved . The experimental time window is typically 12 s, which is limited by the random background and the pile-up of second muon arrival (see section 2.2.3).
Since all the data in this thesis were obtained at TRIUMF, I will explain more details of the continuous-beam SR method in the following sections.