We analyze active and passive seismic data recorded by the Stanford distributed acoustic sensing array (SDASA) located in conduits under the Stanford University campus. For the active data we used low-energy sources (betsy gun and sledge hammer) and recorded data using both the DAS array and 98 three-component nodes deployed along a 2D line. The joint analysis of shot profiles extracted from the two data sets shows that some surface waves and refracted events are consistently recorded by the DAS array. In areas where geophone coupling was suboptimal because of surface obstructions, DAS recordings are more coherent. In contrast, surface waves are more reliably recorded by the geophones than the DAS array. Because of the noisy environment and weak sources, neither data set shows clear reflections. We demonstrate the repeatability of DAS recordings of local earthquakes by comparing two weak events (magnitude 0.95 and 1.34) with epicenters 100 m apart that occurred only one minute from each other. Analyzing another local, and slightly stronger, earthquake (magnitude 2.0) we show how the kinematics of both the P-arrival and S-arrival can be measured from the DAS data. Interferometric analysis of passive data shows that reliable virtual-source responses can be extracted from the DAS data. We observe Rayleigh waves when correlating aligned receivers, and Love waves when correlating receivers belonging to segments of the array parallel to each other. Dispersion analysis of the virtual sources shows the expected decrease in surface-wave velocity with increasing frequency.