Josh McDermott, teaching MIT's course 9.35 Perception, examines how the auditory system extracts pitch and timbre from sound and how it separates a single incoming waveform into the distinct sound sources that produced it. The lecture covers the basics of pitch perception, including how the brain identifies a fundamental frequency from complex harmonic mixtures, and moves into auditory scene analysis, the problem of figuring out which parts of a sound belong to which source when multiple sounds overlap, such as several voices or instruments heard at once. McDermott draws on psychoacoustic experiments and demonstrations to illustrate how listeners group frequency components into coherent perceptual streams. Running eighty minutes as the fourth lecture in the course, it gives a detailed, evidence based account of a problem that underlies everything from speech perception in noisy rooms to music listening.