SenseCam is a wearable digital camera that automatically takes pictures due to timers or sensor-based triggers. We propose using a compass sensor to trigger picture-taking, and detail our implementation of this idea as an add-on module for the SenseCam. We also describe a method for comparing triggering algorithms for SenseCam using a video recorder and video annotation software. We use this method to conduct a preliminary evaluation of compass-based triggering, and show a significant increase in the number of “independent views” captured.