Key frame

• 1. Introduction

  ›

  Key frames are the hidden architecture of almost every moving image you have ever watched. In animation, in film, in the video on your phone, two or three drawings placed at precise moments in time carry the entire logic of motion. Everything in between is filled in later. The key frame does not depict the action; it marks where the action begins and ends. That deceptively simple idea has shaped how animators, editors, and software engineers understand movement itself. How does placing just a handful of drawings across a strip of film create the convincing illusion of fluid motion? And what happens when that same principle migrates from the drawing table into the digital machinery of video editing and compression?

• 2. The Logic of Inbetweens

  ›

  A strip of film or a digital video timeline measures position in time by the frame. Placing a key frame at the start of an action and another at its end defines both what moves and when it moves. The timing of the movement is set entirely by where those key frames sit on the timeline. Because only two or three key frames across a single second of footage cannot produce the illusion of movement on their own, the frames in between, called "inbetweens," must fill the gap. In traditional animation those inbetween drawings were produced by a separate team of artists working from the key frames drawn by senior animators. The key frame is not decoration; it is instruction. Every frame the viewer eventually sees flows from the logic the key frames establish.

• 3. Parameters and 3D Light

  ›

  In software packages that support animation, particularly 3D graphics, a key frame is not limited to a drawing of a character or object. Nearly any parameter attached to any object in the scene can be keyed. A light is a useful example. In 3D graphics, lights function similarly to their real-world counterparts: they produce illumination, cast shadows, and create specular highlights. A single light carries multiple distinct parameters, among them intensity, beam size, color, and the texture it casts. If an animator wants the beam size to shift smoothly from one value to another within a set period of time, two key frames accomplish that. A value is set at the start of the animation and a different value is set at the end. The software then automatically interpolates between those two values, producing a smooth transition without the animator drawing every intermediate state by hand.

• 4. Editing and Audio

  ›

  Non-linear digital video editing software carries the key frame concept into picture and sound work. In this context, a key frame marks the beginning or end of a change made to a parameter rather than a change in visual position. An audio fade is a direct example: a key frame can be placed to indicate the exact point at which the audio level will have risen or fallen to a specific value. Video compositing software operates on the same principle. The editor sets two points in time and the software handles everything in between. The animators drawing by hand and the editors shaping sound share the same underlying logic, even though one works in pencil and the other in waveforms.

• 5. Compression and Seeking

  ›

  In video compression, the key frame takes on a third and distinct meaning. Also called an intra-frame, it is a frame in which a complete image is stored in the data stream. Standard compression works by recording only the changes that occur between one frame and the next, which greatly reduces the amount of information that must be stored. Most typical video sources, such as an ordinary movie, involve only small visual changes from frame to frame, so storing only the differences is efficient. A complete key frame must be created whenever the visual change between frames is so large that encoding the difference would require more data than reproducing the whole image from scratch. Scene cuts and camera switches trigger these full-image key frames automatically. Because each non-key frame only records how it differs from the one before it, seeking to an arbitrary point in the video stream is not possible without a key frame nearby. To allow seeking at predictable intervals, encoders may insert a key frame once every ten seconds of video whether or not the image has changed enough to require one. This makes seeking possible at minimum ten-second intervals, though it increases the overall file size because extra full-image frames are stored. In the DVD MPEG-2 format, where the bitrate is already set high for quality, that size increase does not produce significant compression loss.