Time-lapse is a technique that compresses an extended period of real time into a short video sequence by capturing frames at intervals much longer than normal playback, then playing those frames back at standard speed. Events that unfold over hours, days, or even longer appear to happen in seconds, revealing patterns of movement and change that are imperceptible at normal speed.

The technique works by photographing a scene at regular intervals - every few seconds, minutes, or hours depending on the subject - and then assembling the resulting still images into a video sequence. Flowers blooming, clouds racing across a sky, a city moving through a full day and night cycle, construction progressing over months, or seasons transforming a landscape are all classic time-lapse subjects because the natural time scale of these events makes them difficult to observe meaningfully at real speed. The resulting footage creates a sense of time flowing differently, giving viewers access to temporal scales outside their normal perceptual range. Hyperlapse, a variant that combines time-lapse with camera movement through space, adds another layer of compressed motion to the technique.

When prompting AI video generation for time-lapse aesthetics, describing the subject, the time scale being compressed, and the characteristic visual qualities of time-lapse footage - accelerated cloud movement, sped-up foot traffic, fast-moving shadows - communicates the intended effect. Some AI video generation tools can produce time-lapse style motion natively, while others may require post-production speed adjustment to achieve the characteristic compressed-time quality.