Import your footage to begin editing
00:00:00◖━━●━━◗ ⛶
Media stays in this browser session.
00:0000:3001:0001:3002:0002:30
▸ Video 1
Import media, then add clips here.
▸ Video 2
▸ Video 3
♫ Audio 1
♫ Audio 2
♫ Music
♫ SFX
Save projects in this browser and reopen them. Reimport media to restore previews.
Real metadata inspection and transparent quality checks. No fabricated AI analysis.
Organize your editorial vision and keep decisions with the project.
Clip-level edits are editable in the browser. Multi-track compositing and effects require later development.
Import audio and add it to the audio lane. Listen to your music while reviewing clips.
Export an experimental FCPXML timeline and a human-readable edit decision list. Source paths are placeholders and require relinking.
Analyze actual video frames on your device to identify scene changes, exposure, and visual detail. These are explainable computer-vision measurements, not emotional understanding or cloud AI. Results stay in this browser session.
Choose an imported video. We sample frames at regular intervals and flag large visual differences. Long footage may take time on an iPad.
Waiting for footage.
Import an existing SRT subtitle file from a transcription tool, then search the actual dialogue and generate timecoded story markers. This version does not transcribe audio itself.
Compare measured brightness, detail and detected visual changes. These are starting points for your editorial judgment, not guarantees of shot quality.
Run scene detection to see recommendations.
Keep the original editing workspace above. Use this analysis alongside your manual timeline, creative directions and experimental Resolve export.
Turn your footage notes and imported subtitles into an editable shot plan. This version builds a real, reviewable cut list from timecodes and text matches; it does not pretend to understand emotion or generate a rendered film.
No footage or subtitles have been processed in this workspace.
Add specific source ranges from any video. These become part of the same rough-cut plan.
Change the order or remove selects before exporting. The exports reference source filenames and time ranges; relinking and compatibility must be checked in Resolve.
For true automatic transcription, semantic shot recognition, and cloud projects, a processing backend is still required. EDL supports one simplified video track; frame rates and source reel mapping are approximate.
An expanded local-first creative suite. Analyze real frames, build scored selects, prepare editable shot plans, tap music tempo, annotate decisions, and carry projects between sessions. Processing happens in your browser; no footage is uploaded by this module.
Sample video frames and score measurable brightness, edge detail, and changes between frames. These are technical heuristics—not emotional recognition or a trained AI model.
Review sampled frames and approve candidates. Each approved range can be sent into the existing Story & Rough-Cut Lab.
Tap in time with a track to estimate BPM. Create a cut-spacing guide and export beat markers. No automatic beat detection is claimed.
Create an explicit edit blueprint for a future AI service or your own editing session. The guidance is rule-based and fully inspectable.
Save shot selections and creative notes in this browser or transfer them as JSON. Video files are not embedded; relink the originals after reopening.
Plan output dimensions and generate a handoff checklist. This does not render or reframe video.
Adjust in/out times, notes, and order. Export a clean CSV or JSON, or send the sequence into V6's manual-select timeline planner. Files must be relinked on the editing computer.
A shared editing decision timeline with real source-video playback, trim and split controls, V8 approved-shot import, local project backup, and experimental Resolve handoff. Your original footage stays on your device.
Import original videos here to relink your timeline. Media isn't uploaded to a server or saved in the JSON.
For cross-device use, export JSON and import it on the other device, then relink original videos. This is not cloud sync.
One shared V9 timeline. Build from timed transcripts, select story moments, and apply transparent editing instructions. This is a local, rules-based assistant—not cloud AI transcription or emotional recognition.
Import an SRT subtitle file, or paste SRT text. Select passages and send their actual source time ranges into the unified timeline.
Apply deterministic commands to the current unified timeline. Every change is undoable using V9's Undo button.
V9 remains the authoritative timeline and export workspace. Original files stay local; subtitles may contain private dialogue, so don't share exported project files without reviewing them.
One connected workflow: configure a private transcription service, produce timecoded dialogue, send it to Director Lab, and prepare a Resolve handoff. No browser API keys. Your original video files stay local unless you explicitly upload a file for transcription.
A Worker is required for transcription. The included Worker uses an OpenAI API key stored as a Cloudflare secret. AI transcription incurs provider charges. Do not put API keys into this page.
Review names and dialogue before using them. Director Lab will let you select lines and add them to the unified timeline.
After sending SRT to Director Lab, choose the exact source filename, select the lines, and add them to the V9 timeline. Export FCPXML or EDL from Studio IX, then relink and verify the media in DaVinci Resolve. This is an interchange workflow, not a live Resolve plugin.