Communitygithub.com

reel-speed-ramps

Turn slow AI-generated scenes (Veo, Flow, Kling, Seedance, Codex…) with an avatar voice-over into a fast vertical reel. Speed ramps that never touch the voice, every cut landed on the beat of a song, and the song's sung lines moved into the gaps between voice-over lines so lyrics and voice never overlap. Use when someone asks to speed up / tighten / add speed ramps to a reel or ad made of AI clips, to fit an edit to a song's beat, or to use a song with lyrics under a voice-over.

reel-speed-ramps とは?

reel-speed-ramps is a Codex agent skill that turn slow AI-generated scenes (Veo, Flow, Kling, Seedance, Codex…) with an avatar voice-over into a fast vertical reel. Speed ramps that never touch the voice, every cut landed on the beat of a song, and the song's sung lines moved into the gaps between voice-over lines so lyrics and voice never overlap. Use when someone asks to speed up / tighten / add speed ramps to a reel or ad made of AI clips, to fit an edit to a song's beat, or to use a song with lyrics under a voice-over.

対応~Claude Code✓Codex CLI~Cursor
npx skills add https://github.com/synthlabcreator/reel-speed-ramps/tree/HEAD

お気に入りのAIに質問する

このエージェントスキルを事前に読み込んだ状態で新しいチャットを開きます。

ドキュメント

Reel speed ramps

One project file, five commands, all local (ffmpeg + Python). The script is scripts/reel.py; the example is examples/project.json.

Rules that make it work

  1. The voice-over is sacred: always 1x. Measure where the voice is on the Demucs vocal stem of each scene. The ramp blends back to 1x 0.35 s around it.
  2. Ramp the shots without voice, never the whole clip at once. Use one mode per shot:
    • c: fast in the middle (default, 2–3.2x peak).
    • k: fast at the edges and 1x at the key moment (jumps, tricks, impacts).
    • n: real speed. Use it for people walking: walking looks broken when sped up.
  3. The song decides the cuts. Pick cut times on the beat grid, put target_frames on each scene and let the script scale the ramp intensity until the scene lands on that exact frame.
  4. Lyrics never on top of the voice. AI music tools (Flow Music, Suno, Dreamina) ignore requested timing and "no vocals here" sections. So:
    1. Split the song with Demucs.
    2. Get word timings with Whisper.
    3. Move each sung line by a whole number of beats into a gap between voice lines. Whole beats keep it on time.
  5. Duck the music 14 dB under the voice-over (6 dB is not enough on a phone). Mix to −14 LUFS and add a limiter, because AAC encoding pushes peaks over 0 dBFS.

Workflow

  1. python scripts/reel.py measure scene1.mp4: prints the voice windows and the detected shots (with a suggested peak per shot) as JSON for the project file. Look at a contact sheet and fix the modes (k for jumps, n for walking). Check the voice windows by ear, because sound effects can sneak in.

  2. python scripts/reel.py ramps project.json: first pass without target_frames, to see the new length. Show it to the user without music before going on.

  3. Write the song prompt from the new timeline: BPM chosen so 4 bars ≈ one scene, sections, the drop on the edit's hit, and lyrics in the language asked. Generate it (ask before spending credits).

  4. python scripts/reel.py song song.wav --lyrics: gives the tempo, the beat, drop candidates, the stems and each line with its time.

  5. Fill music in the project:

    • song_drop: the drop's second in the song.
    • beat.
    • lines, each with start, end, shift_beats.

    Set target_frames so every cut lands on a beat: cut time = drop − k × beat.

  6. ramps again, then music. It warns CLASHES WITH VOICE if a line overlaps a voice; change shift_beats until it doesn't. Then assemble project.json out.mp4.

  7. Check the result:

    • a contact sheet;
    • Whisper on the final mix (every voice line must come out clean);
    • −14 LUFS and the peak below −1 dBFS.

Tail pieces

After the ramped scenes, tail adds pieces in order:

  • part of a clip at 1x with its own sound (volume);
  • a clip slowed down with a push-in (slow, push_in);
  • a PNG sequence (an end card rendered in Remotion, After Effects…).

Requirements

Python 3.10+, ffmpeg on PATH, pip install -r requirements.txt (numpy, scipy, soundfile, opencv-python, librosa, demucs, openai-whisper).

Case: Nike spec reel (2026)

  • Scenes: four AI scenes of 10 s plus a sneaker burst; each scene ends on an avatar voice line.
  • Ramps: the scenes went from 40 s to 28.3 s with every voice line untouched, and the reel from 48 s to 36.3 s. Walking shots stayed at 1x after the client's note.
  • Song: generated in Flow Music, 150 BPM, 5 credits, English lyrics. It came back 67 s long and rapping non-stop. Five lines were moved by whole beats into the voice gaps. The chorus lands on the drop, which is the cut to the product burst.
  • Client verdict: «much better».

関連スキル