HyperFrames v0.8.221
By Steven Van ·
Snapshots read each video clip once, Ctrl+C stops preview at once, and the CLI gains a motion blur docs topic.
Video snapshots in HyperFrames now probe each clip once for all the times requested, so they take less time. Preview also stops straight away on Ctrl+C. Both are in v0.8.221, released 30 October 2026.
CLI changes
- Snapshot probes each video clip once for all the times it is asked for.
- Preview stops at once on Ctrl+C after Studio was open.
- add takes several names in one call, and adopted audio is labelled voice or sfx.
- check no longer flags a hidden scene child or a small animated sprite.
- A missing registry item now names the closest matches, and a blocked browser says how to allow it.
- A failed late-image snapshot reports why it failed.
- hyperframes docs motion-blur shows the motion blur guide.
Other fixes
- Core: long cut-list edits play every cut without dropped or late frames.
- Studio: pending video edits are no longer capped at 10.
- Registry: caption components keep the last phrase until its last word ends.
The motion blur docs topic
The linked commit adds motion-blur to the topics listed by npx hyperframes docs, alongside data-attributes, examples, rendering, gsap, troubleshooting and compositions. The guide describes motion blur as a shutter smear on fast-moving elements. It is opt-in, and nothing blurs until the composition root has data-motion-blur="on".
The guide lists these attributes:
- data-motion-blur: on enables blur, and off opts out on the root or on any subtree.
- data-motion-blur-shutter: shutter angle in degrees, from 0 to 720 (default 240).
- data-motion-blur-samples: samples per frame, from 2 to 64 (default at least 32, sized to the smear).
- data-motion-blur-events: on runs timeline callbacks while blur looks for movers, which is needed when motion is set from a render(t) callback.
- data-motion-blur-tier: auto (default), capture or filter.
The same attributes drive the browser editor preview and hyperframes render.
Capture and filter tiers
The capture tier redraws each mover along its path, so turns and zooms smear truthfully. It needs a browser with HTML in Canvas. The filter tier puts an SVG filter on each mover. It works in any Chrome and costs little, but turns and zooms are only softened. With auto, HyperFrames uses filter when every mover only translates and capture when anything turns or scales.
Checking and rendering
npx hyperframes check reports motion_blur_filter_fallback, motion_blur_no_movers and motion_blur_unverified. hyperframes render --page-motion-blur off renders without the composition's blur. Blur is off for HDR, distributed and sub-frame-blur renders.