Frame rate

• 1. Introduction

  ›

  Frame rate is the invisible pulse beneath every moving image you have ever watched. It is the number of still pictures shown in sequence every second, and the gap between a crude silent flicker and cinema you believe is real. Thomas Edison once declared that 46 frames per second was the minimum the eye needed to perceive motion, adding: "Anything less will strain the eye." That threshold shaped debates that rippled through a century of film, television, and video games. How did engineers settle on the numbers that govern every screen today? Why does the electricity grid have anything to do with the movies? And what happens inside the human eye that makes any of this work at all?

• 2. What the Eye Actually Sees

  ›

  Human vision can process 10 to 12 images per second and recognize them as individual pictures. Above that rate, the brain stitches them together as motion. But the threshold for perceiving a flickering light as steady is a different matter entirely. Studies show that modulated light, the kind a computer display produces, reads as stable to most people when the rate exceeds 50 FPS. That point is called the flicker fusion threshold. When the light source carries an image rather than a plain glow, that threshold climbs far higher, into the hundreds of hertz.

  Persistence of vision adds another layer to this already strange picture. A single flash of light lasting just one millisecond can register in the mind as lasting somewhere between 100 and 400 milliseconds. Two very short flashes can merge into something new entirely: a 10-millisecond green flash followed immediately by a 10-millisecond red flash is perceived not as two events, but as a single yellow one. Human image recognition goes further still. People have been shown to identify a specific image within an unbroken sequence where each picture lasts as little as 13 milliseconds. The visual system, in short, is neither simple nor slow.

• 3. Silent Film and the Hand Crank

  ›

  Early silent films were filmed at rates anywhere from 16 to 24 frames per second, but those numbers were not fixed. Cameras were hand-cranked, and the operator could speed up or slow down mid-scene to match the desired mood. The projectionist in the theater had the same power: a rheostat controlling the voltage to the film mechanism could shift the playback rate up or down in real time. Film companies often sent their pictures to theaters expecting them to be shown faster than they were filmed.

  All of those rates were enough to create the sensation of motion, but the result looked jerky. Projector engineers tackled this with an elegant mechanical trick. By fitting projectors with double-bladed or triple-bladed shutters, each frame was flashed on screen two or three times before the next one arrived. A film shot at 24 FPS was therefore displayed at 48 or 72 flashes per second, reducing the eye strain that lower apparent rates produced. By the mid to late 1920s, the industry had pushed silent film frame rates to the range of 20 to 26 FPS as equipment improved.

• 4. The Moment Sound Changed Everything

  ›

  Sound film arrived in 1926, and it ended the era of flexible frame rates overnight. The human ear is more sensitive to frequency changes than the eye, and any wavering in film speed would warp the pitch of dialogue and music in ways audiences could immediately detect. A new standard had to be chosen and locked in. Many theaters had already been running silent films at 22 to 26 FPS, so 24 FPS emerged as a practical midpoint.

  Between 1927 and 1930, studios updated their equipment and 24 FPS became the industry standard for 35 mm sound film. At that rate, film moves through the projector at 456 millimeters per second. The familiar two-blade shutter then displayed each frame twice, producing 48 images per second on screen and satisfying Edison's earlier recommendation. Many modern 35 mm projectors go further, using three-blade shutters that flash each frame three times, reaching 72 images per second.

• 5. Animation on a Budget

  ›

  Drawn animation has its own relationship with frame rate, shaped as much by economics as by aesthetics. Most animated films run at 24 FPS, but animators discovered early on that drawing a new image for every single frame was enormously expensive. The solution was animating on twos: one drawing held for two consecutive frames, cutting the drawing count to 12 per second. For most movement, that is fluid enough that viewers do not notice.

  When a character needs to move quickly, the compromise breaks down. A fast punch or a spinning turn requires animating on ones, a new drawing for every frame, to keep the motion legible. Skilled animators blend the two approaches, switching between them as the action demands, managing both the visual result and the production budget at once. Saturday morning cartoons of the mid-1960s pushed even further in the economical direction, shooting on threes or fours. That means only 8 or 6 drawings per second on screen. Anime also typically runs on threes or twos.

• 6. Television, Color, and the Grid

  ›

  Analog television did not inherit its frame rates from film tradition. It inherited them from the electrical grid. Engineers synchronized broadcast standards to mains frequency because the power grid ran at an exceptionally stable frequency, making it a reliable timing reference. Regions running 50 Hz grids, most of the world, built television around 50 FPS. North America, Japan, South Korea, Mexico, Canada, and the Philippines, running on 60 Hz, developed a 60 FPS standard.

  Color television introduced a complication. Displaying color on legacy black-and-white sets produced an artifact called dot crawl, a crawling pattern visible on highly saturated surfaces. Engineers found that dropping the frame rate by 0.1 percent made the artifact disappear. That is why video transmission standards in North America, Japan, and South Korea settled on 60 divided by 1.001, approximately 59.94 images per second, rather than a clean 60. The confusion this created in early digital video software was widespread. Many developers wrote software on the incorrect assumption that only 29.97 images were expected each second, misunderstanding how interlaced formats work.

• 7. Games, Tearing, and Frame Time

  ›

  Computer games face a challenge film never does: the images are generated in real time, frame by frame, rather than recorded in advance. For a long time, 60 frames per second was considered the baseline for smooth gameplay. PAL markets, running on 50 Hz television output, caused fast-paced games such as racing or fighting titles to run noticeably slower before the sixth generation of consoles. Developers who did account for the difference altered game code to match pacing across regions, with varying results.

  Computer monitors built for competitive gaming can now reach 360, 500 FPS, or higher. The practical effect of those rates is reduced blur on fast motion, whether sprinting through an open-world game or tracking details in a fast multiplayer fight. Input latency also drops at higher frame rates. A related concept, frame time, measures the gap between individual frames rather than the average rate. A game averaging 60 FPS can still feel choppy if frame delivery is uneven, which is why some reviews measure the worst 1 percent of frame times, the 99th percentile, as a separate quality indicator.

  When a game's frame rate and a display's refresh rate fall out of sync, the screen shows two frames at once in a horizontal split called screen tearing. Vsync eliminates the tear but caps the frame rate to the display's refresh rate, adds input lag, and can introduce its own stuttering. Variable refresh rate displays resolve this by adjusting the display's rate dynamically to match whatever frame the game is currently delivering.

• 8. Synthesizing Frames

  ›

  Frame rate up-conversion, abbreviated FRC, is a process that generates new intermediate frames between frames that already exist. Low frame rates produce aliasing and abrupt motion artifacts that degrade perceived video quality, making temporal resolution one of the key factors in how good video looks. FRC algorithms now appear in visual quality enhancement, video compression, and slow-motion video generation.

  Two major families of algorithms handle the problem differently. Flow-based methods analyze the optical motion between two consecutive frames and predict what an intermediate frame should look like, adjusting for object depth through a flow projection layer. Pixel hallucination-based methods use a different tool called deformable convolution, replacing optical flow data with offset vectors. The trade-off between the two is practical: hallucination-based approaches can interpolate middle frames within the feature domain, but they tend to produce blurry results when fast-moving objects are present, unlike flow-based methods. As video formats push toward 120, 240, and 300 frames per second natively, the algorithmic challenge of bridging the gap between source and display frame rates will only grow more consequential.