keyframes · shared timing
AI keyframe animation where the keyframes leave the node.
Most AI video tools give you a clip and leave the timing to luck. In QueyFrames a keyframe is a first-class output: marks on a waveform, beats in a song, moments in a clip: every generator publishes its timing, and downstream nodes reuse it to drive their own.
Motion lands on the beat, cuts line up across layers, and a change upstream re-times everything that depends on it — without a single value being retyped.
A keyframe is a span, not a setting
The shape matters, because it is what lets a keyframe stand alone rather than only meaning something relative to the one before it.
start and end
A segment carries both, in seconds against the composition — so it is a stretch of time you can read on its own, not a point whose meaning depends on its neighbours.
a description
"The dolphin leaps toward the boat" is a keyframe too. It is what a generative video prompt can actually use, and what makes a plan readable when you come back to it.
properties with from → to
Position, scale, rotation, opacity, colour — each one names both of its ends, so a span renders correctly wherever it is dropped and holds its last value outside the spans that name it.
Where keyframes come from
Four different nodes publish timing, and they all speak the same clock: named milliseconds into the media, never a fraction of one node’s own length. That is what makes a mark mean the same instant everywhere it is read.
the waveform analyzer
mark by earPlay a clip back, mark its section bounds on the wave, and send those timings on. The obvious way to cut picture to music you already have.
the music node
automaticIn composition-plan mode it publishes the end of every section, named for it — so the structure of the track is available as timing without anyone marking anything.
the clip timelines
mark by eyeThe video node's own ruler and the composite's marker lane both take marks you drop at the playhead, drag to move, and send downstream as guides across every track.
generated plans
describedDescribe a shot and get the composition, the title and a full set of keyframe spans back — as ordinary editable rows, not a locked result.
What they drive
The same track can feed several consumers at once, and each one reads it in the way that makes sense for what it does.
programmatic renders
Segments become interpolated per-frame styles in a real 1920×1080 Remotion render. Wire markers instead and n marks re-cut the node's own segments into n−1 spans, keeping their descriptions.
AI video prompts
Keyframes are numbered across every input slot and written into the generation prompt individually, so the model is given the pacing rather than guessing it.
music structure
Wire a keyframe track into the music node and its marks become the section boundaries of the composition plan — a song built to the cut, instead of a cut fitted to the song.
Keyframes — common questions
- What is a keyframe in QueyFrames?
- Two things, on one clock. A marker is a named millisecond into the media — the end of a chorus, a cut point, a beat. A segment is a span with a start, an end, a description and the properties it animates. Both travel on the same ports, so any node can read either.
- How is this different from keyframes in a normal video editor?
- In a normal editor a keyframe belongs to one property on one clip. Here it leaves the node on a port, so several downstream nodes read the same timing at once — and when the source changes, every one of them re-times instead of being re-cut by hand.
- Where do keyframes come from?
- Mark them by hand on a waveform, a generated clip or the timeline ruler; let the music node publish the end of every composition-plan section automatically; or describe a shot and get a whole plan of keyframe spans back to edit.
- Can keyframes drive an AI video generation prompt?
- Yes. Wired keyframes are numbered and placed into the video prompt individually, so the model is told when things happen rather than being left to invent the pacing. Anything past the clip's maximum length is dropped rather than crushed onto the last frame.
- Can one keyframe drive one node?
- Yes. The keyframes node fans its segments out one per port, and the split node does the same for an incoming track, so a single mark can feed a single downstream node instead of the whole track going everywhere.
- What happens if I move a keyframe after wiring it?
- Downstream nodes re-time to the new position. Where a timing would break a provider's own rules — music sections have a minimum length, for instance — the connection is refused at the port with the reason, and an edit that makes an existing wire illegal cuts it rather than silently stretching the result to fit.
Keep reading
QueyFrames — AI video generator on a node canvas
The AI video generator the whole canvas is built around — video, images, audio and text on one graph.
Gemini AI video generator — text to video
How QueyFrames uses Gemini AI for video and stills — text to video, image to video, and the prompt surface around them.
AI music, voice and sound effects generator
The audio half — AI music, AI voice generation and sound effects, all publishing timings the video nodes read.
AI video generation guides and workflows
Guides and workflow notes — prompting for pacing, image to video, and keeping audio and picture on the same beat.
Mark it once. Everything downstream follows.
QueyFrames is in early access. Create an account to join the list.