The effect of the spatial distribution of a radioactive source on timing measurements has been investigated with particular consideration toward the focal plane of recoil separators. The work conducted during this thesis is a precursor to understand the magnitude of such effects for the upcoming fast timing array (FATIMA) at FAIR. An experiment was undertaken at the University of Jyvaskyla using the K130 cyclotron to accelerate a 36-Ar beam to 190 MeV, directed onto a 106-Cd target, to produce recoils of 138-Gd and 136-Sm via fusion-evaporation reactions. Recoils directed using RITU to the focal-plane DSSSD of GREAT were distributed over the majority of the 124-mm by 40-mm extension of the DSSSD. A new array consisting of eight lanthanum bromide detectors was used to measure the time between coincident prompt-gamma rays emitted following the de-excitation of isomeric recoil states implanted into the DSSSD. Lifetimes were measured to be 213(20) ps and 200(100) ps for the first-excited 2+ states in 138-Gd and 136-Sm, respectively. Positional information, extracted from the DSSSD, was used to correct for the difference in the time-of-flight of gamma rays as they travelled from the implantation position to the lanthanum bromide detectors. When accounted for, the lifetimes were remeasured to be 217(20) ps and 210(90) ps, respectively, showing no significant change in value or error. A method of quantifying the increase in uncertainty of a lifetime measurement due to the spatial distribution of the source and the position of the surrounding detectors, supported by simulation, has been provided to explain these observations. A new technique for extracting the time-walk from each of the CFDs in a multi-detector array has been presented. The new technique offers a reduced complexity in calculations by accounting for the correlated time-walks present in time measurements from different detector-pairs sharing a common CFD. Work towards a technique for extracting lifetimes from time data has been presented. Dubbed the Symmetrised-Convolution Lifetime Measurement (SCLM) method, this technique essentially applies a model-dependent convolution of the prompt-response with nuclear exponential decay on both time spectra, obtained by inverting the start and stop conditions of a TAC, simultaneously and draws parallels to the Mirror Symmetric Centroid Difference method.