VFX Atlas
H 248 / 3212000s

Datamoshing

Experimental background. A significant wave of music videos and online activity occurred in 2009.

Typeface · Rubik Glitch

Lee Brenton: My first datamosh, November 15, 2009: Original video still. Temporal effect: still image or image sequence.
Lee Brenton: My first datamosh, November 15, 2009: Original video still. Temporal effect: still image or image sequence.Lee Brenton. CC BY 3.0View source ↗

Prompt

Apply a datamoshing effect with blocky compression fragments, smeared color patches, and displaced textures that appear dragged along implied motion paths. Let selected edges dissolve into fluid rectangular trails and mismatched image residues while retaining recognizable central features. Use uneven digital breakup rather than uniform noise. Keep the original subject and composition.

Paste it into LetzAI with your own image to apply the effect ↗

Identify

Pixel blocks become fluid. Foreign textures follow body movements. One scene seems to melt into the next.

Principle

Video compression1 stores motion data. Modified reference images2 cause this data to be applied to unsuitable image content. For example, Welcome to Heartbreak and Evident Utensil spread the look in 2009.

Try it out

Compare blending, morphing, and Datamosh. Do the transparency, shape, or the relationship between movement and texture change?

Background

Experimental prehistory; significant music video/online wave in 2009.

Temporal video compression doesn't save every frame completely, but also captures change and motion information, among other things. If reference frames are missing or motion data is used inappropriately, the texture of one scene can bleed into the motion of another.

Kanye West, Welcome to Heartbreak (2009, directed by Nabil Elderkin), and Chairlift, Evident Utensil (2009, directed by Ray Tintori). Earlier art reference: Takashi Murata, Monster Movie (2005).

Image: Historical original / source image. Color and macroblock areas smear across the subject. Still image from a real datamoshing sequence; the video shows how motion data further distorts the old image.

Distinction

Datamoshing is a specific subgenre of glitch aesthetics, not a name for every RGB shift or screen noise. Compression errors predate the term. The 2000s are marked by their deliberate artistic use and popularization.

Aftermath

The normally invisible codec becomes an image engine. A technical defect appears as a painterly transition.

Further examples · 7 images

Decodierfehler Videowiedergabe 4K-Video 20230824
02 Decodierfehler Videowiedergabe 4K-Video 20230824A complex field of multicolored pixel blocks and digital fragments is visible. The image illustrates a decoding error where motion data and textures have merged into a fragmented, non-representational grid characteristic of datamoshing.PantheraLeo1359531 · Public domain · Wikimedia Commons
Macroblocking error
03 Macroblocking errorCloud textures are distorted by macroblocking errors where motion vectors are applied without new image data. The resulting blocky, melting clouds are a clear example of compression-based datamoshing.LikakiPhotos · CC BY-SA 3.0 · Wikimedia Commons
Tramas mpeg
04 Tramas mpegThis diagram illustrates the structure of MPEG I, P, and B frames. Datamoshing manipulates these specific compression units by removing I-frames to force motion vectors onto the wrong image data.Jennifer Galindo · Public domain · Wikimedia Commons
H.264 block diagram with quality score
05 H.264 block diagram with quality scoreThis diagram outlines the H.264 compression process, including motion estimation and intra prediction. Datamoshing specifically exploits these technical stages by manipulating motion vectors and removing keyframes to create intentional visual errors.Karthikeyan R, Sainarayanan G, Deepa SN · CC BY-SA 4.0 · Wikimedia Commons
AI illustration: Pixel blocks from the walking figure drag across the grass background
06 AI illustration: Pixel blocks from the walking figure drag across the grass backgroundPixel blocks from the walking figure drag across the grass background. The trail illustrates P-frame motion data being applied to a static area without a refreshing I-frame.Illustration generated for the Atlas (AI)
AI illustration: A street scene exhibits vertical smearing and rainbow-colored pixel artif
07 AI illustration: A street scene exhibits vertical smearing and rainbow-colored pixel artifA street scene exhibits vertical smearing and rainbow-colored pixel artifacts. The distorted right panel demonstrates how movement data causes textures to melt and bleed into adjacent pixels.Illustration generated for the Atlas (AI)

Sources