Can I change the words and colors?

Yes. topic and terms are JSON inputs, with package palettes and overrides. The current cloud lettering is fixed neutral white over darkened footage; changing palette does not automatically recolor that field.

Do I need to supply footage?

Yes. Bring a full 6.8-second local take plus an aligned transparent cutout, or allow automatic local cutting. The foreground layer is required so all typography can sit behind the real subject.

Can I revise the motion?

Yes. Keep the persistent anchor, staggered term entrances and matched foreground above the type. Re-render after position, crop or timing edits and inspect both layering and collisions.

Are larger words more frequent or important?

Their sizes are authored for the composition and vary by array position. The effect does not calculate frequency or relevance; explain the selection yourself.

Will the words automatically stay clear of the presenter?

The subject can intentionally occlude the type, but the positions are fixed. Use suitable framing and short terms, then review the real moving composition; no tracking-based rearrangement is performed.

CreatorMotion · Topic Cloud Behind Presenter

Previews from the publisher

Build a field of related ideas behind a real presenter. A large topic anchor appears first, then up to eight terms enter at uneven sizes around it while the foreground subject stays above all lettering. The 6.8-second shot keeps the title, terms and person in place after their entrances. It creates an editorial topic cloud from supplied words rather than analyzing text to determine their importance.

What you get

A local generator, editable HyperFrames layered presenter shot and motion/review instructions for a 1920×1080,30 fps segment. Generate the composition with your continuous footage and aligned cutout, then render it.
topic and up to eight terms. The large white headline anchors the field while smaller uppercase terms use authored upper, side and lower positions with varied sizes. Term size comes from its slot, not frequency or a calculated relevance score.
Topic entrance from 0.25 to 0.58 seconds followed by sequential terms from 0.95 seconds at 0.37-second intervals. Each stays still once revealed. A darkened base video, typography field and matched foreground matte form the depth relationship.
Four styles keep the same data and choreography: White & Yellow (white-yellow), Charcoal & Lime (charcoal-lime), Paper & Orange (paper-orange), and Black & White (midnight-ice). Semantic style overrides are supported; check contrast after changing them. Dark canvases use a fine grid. Local fonts include Inter, IBM Plex Mono for measurements and ordinals, Smiley Sans and Noto Sans SC. Chinese text selects Smiley Sans automatically, including accompanying Latin text; --zh-font noto-sans-sc selects Noto Sans SC, while --font takes precedence for a supplied font.

Before you choose

Select a small, coherent set of terms and inspect how the person divides the type field. A topic cloud should introduce a range of ideas without implying that decorative word sizes are measured evidence.

Good fit

A creator introducing the range of ideas around one concise subject while keeping the person visible.
A teaching passage that needs an anchor concept and several associated terms rather than a complete definition.
A series with consistent presenter framing, type hierarchy and accumulating conceptual fields.

Not for

Automatic keyword extraction, frequency-based word clouds, topic modeling or substantiating a relationship between supplied terms.
More than eight displayed terms, long paragraphs, automatic collision avoidance or a promise that every word stays unobscured by a moving presenter.
Automatic long-form editing, generating absent presenter footage or music, or copying another creator’s identity and source material.

How to use

You say

/creatormotion-type-topic-cloud Turn my supplied explanation into a white-yellow B-roll scene using the Topic Cloud Behind Presenter effect. Use the text, data and media I provide, and ask for any required input that is missing.

Prepare a short topic and up to eight short terms. Choose the order intentionally: positions and size vary by index. Keep meaning independent of any implied statistical weight, and replace illustrative defaults with terms relevant to your explanation.
Supply a continuous local --media take at least 6.8 seconds long and its same-frame transparent presenter-matte through --matte FILE or the image map. Base and foreground share the 135% crop and left translation. Keep them aligned through every frame; no still portrait fallback is supported.
If no cutout exists, default --matte auto invokes pinned local HyperFrames remove-background, with first-use model download and checksum caching. Alpha-bearing footage can be used as its own cutout. --matte off disables automatic cutting but cannot make this required-matte effect work without an alpha source.
From the extracted package, prepare Node.js 22+, FFmpeg and HyperFrames CLI 0.8.30. Run node scripts/create.mjs --out PROJECT --data CONTENT.json --style white-yellow, substituting another documented palette if wanted. Read the motion specification and image policy before adapting the scene; use semantic --overrides for deliberate style changes. Add --media /absolute/PRESENTER.mp 4 and an available --matte /absolute/MATTE.webm. Videos remain muted; do not loop a shorter take. Narration is separate. For 4K delivery, bring 4K source and use --resolution landscape-4k on rendering.
Inspect the editable project at entry, each motion handoff and final hold, then render with the command printed by the generator. Check the resulting frames for text collisions, correct numbers and readable contrast. Keep at least 1.2 seconds for the final reading when editing, and compare all four palettes with identical content after structural changes. Narration, music and assembly into a longer film remain separate work. Inspect matte edges, actual face/letter overlap and every smaller term. Only the intended foreground-over-type relationship permits occlusion; term-to-term collisions remain failures. Keep earlier words stable and review cutout quality rather than treating recorded checksums as visual proof.

Requirements and costs

You bring: topic: provide your own content or retain the illustrative default. · terms: provide your own content or retain the illustrative default.
Dependencies: HyperFrames CLI 0.8.30 or compatible — required · Node.js 22 or newer — required · FFmpeg — required
Outside costs: Not collected yet
Tools used: HyperFrames
Publisher testing: v1.0.2 — Claude Code / Claude Desktop / Cowork / Codex / ChatGPT