Loading...
Loading...
Draw one rectangle over the detail you want to match, then add trusted DAM images as references.
or click anywhere to browse
MP4, MOV, WebM, M4V
Used for the filename and as context for AI description
If left blank and AI description is enabled, this will be auto-generated
AI will watch the video and generate a detailed prompt/description (uses Gemini 2.5 Flash)
or click anywhere to browse
MP4, MOV, WebM, M4V
Rendering preview...
No presets match
Apply/Export: full resolution, v4.25
Click on the video to place tracking target
Select a clip in the timeline to adjust color
No titles yet
Click "Add Title" to create text overlays
Use {WORD} for hero styling.
Modifiers: {WORD|size:large}, {WORD|size:small}, {WORD|italic}
Select a clip in the timeline to add transitions
Select a clip in the timeline to add effects
Optional imperfect handheld zoom-in curve.
Stretch Edges extends border pixels instead of mirroring nearby detail.
Clamp edge fill inward to hide seam rows/columns.
Shake-derived blur (web preview + web GPU export)
No zoom effect applied
No blur effect applied
No dim effect applied
No subject split applied
No subject split report yet.
No gradient overlay applied
No denoise effect applied
Motion blur uses optical-flow processing (Fast or RSMB engine). Results are cached.
No motion blur effect applied
No overlays yet
Use Actions → Add Logo/Alpha Overlay
Select a logo overlay to adjust scale and position.
Loading...
MP3, WAV, M4A, AAC • Max 100MB
Create custom music with ElevenLabs Eleven Music
10–120 seconds
Tip: Put one sentence per line. Each line can be generated and moved as its own clip.
Uses voices from your ElevenLabs account.
Generate at least two VO clips to adjust spacing.
Applies fixed gain and compression for clarity over music.
Scales all VO while preserving each line's own volume setting.
Matches current timeline VO clips based on the exact segment used for each line.
Lowers music while voiceover clips are active.
Balanced keeps broader phrases. Tighter Matching makes smaller VO segments for line-by-line replacement.
No voiceover clips
Replaces audio on the selected VO clip and keeps its position/pin.
AV1 is slower but higher quality for multi-pass editing
Fill color for letterbox/pillarbox areas
Dims UI and expands video when playing
Shows true film cadence for titles and animations
Displays current playback speed in preview window
Reduces audio sync corrections for wireless headphones
Choose a worker to pin this project to one render server, or leave as Any available worker. Worker limit is per project, from 1 to 48, and empty uses the server default.
Round trip: the worker stages sparse inputs, renders remotely, uploads the final edit directly back to this laptop, and backfills intermediate artifacts afterward.
Loading remote render status…
Proxies can be regenerated from source files
Reference files for AI chat context
Editing global settings for guest.
Performance settings for development.
These developer settings are saved per user on the server.
Local media routing and source priority are intentionally not wired in normal runtime.
All app-managed media currently loads from remote CDN/Cloudflare-backed URLs.
Future re-hookup is deferred. The old implementation remains in the repo as dormant dev scaffolding only.
Uses stale cached thumbnails, skips video frame extraction
Assumes proxies are already built
Polls every 30s instead of 5s
Restore cached thumbnails on page refresh
Log zone data at each step (open, save, trim)
When off, Vidgen will not scan the full Natron OFX plugin list. RSMB still works.
When off, the DAM page will skip POI status checks and hide the DAM POI detection UI.
Local renders in the browser on the machine viewing the cut. Server runs headless on the backend host serving this page.
Use Reframe POI to set crop positions. This selector is single-choice for "Bake Titles" and "Export Now". Batch actions run all standard delivery aspects plus 2:3 when enabled.
Scale video slightly to hide edge artifacts
Resolution and frame rate are set in the Project tab
No exports yet
This cut is not in Linked Mode
Select a clip in the timeline to apply OFX effects
Loading finals…
Drop video file here or click to upload
MP4, MOV, WebM, M4V • Max 100MB
AI will validate each piece of feedback, add timestamps, and prioritize fixes
No evaluations yet
Gemini 2.5 Flash will break down your video shot-by-shot
Plan deliverables by status or by date.
No canvas selected
Select a canvas from the list or create one via chat
The Cake Matrix is optimized for desktop viewing. Please use a larger screen to access this feature.
Upload product photography references for AI image generation
Non-product references for AI generation (bakers, ingredients, environments)
Generate music prompts with AI, then create audio with ElevenLabs
No music prompts yet. Generate some above!
No audio files yet. Generate music from your prompts!
ENVATO_TOKEN in .env
Get your token from build.envato.com
Your favorites collection ID from AudioJungle
Set ENVATO_TOKEN in .env to connect
Loading...
Loading assets...
Batches are staging areas for files before ingestion:
jan2025_valentines)dam/batches/{batch_id}/After uploading, go to Ingest to process the batch into the catalog.
Drag & drop files here, or click to select
Import files from a batch into the catalog
Ingestion processes files from a batch and creates catalog assets:
Original files are NOT copied - only thumbnails and frames are stored. After ingestion, run Extract to add AI metadata.
Import directly from a folder (recursive, resumable)
Import from large network drives or local folders:
Paste or type the full path to your network drive folder.
Click Browse to select a folder, or paste/type the full path. Server will scan recursively.
Re-import PNGs to fix transparency artifacts
Some PNG images have transparency that wasn't handled properly during initial import:
Point to the folder containing original PNG source files.
Checking... PNG-sourced assets in catalog
Detect point of interest for smart cropping
POI detection uses Gemini AI to identify the visual center of each image:
Requires GEMINI_API_KEY and google-generativeai package.
Use Gemini AI to extract metadata from assets
Extraction uses Gemini AI to analyze assets and generate rich metadata:
Pending processes new assets. Failed retries assets that encountered errors. Requires GEMINI_API_KEY.
View aggregated statistics from your catalog:
Stats are computed from the manifest. Run Extract to populate metadata for accurate stats.
Last updated: -
Select references for multi-image generation. Primary subject + optional style and context references.
Authoritative server backups, git snapshots, schedules, roots, and coverage.
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Loading managed runtime config…
Loading runtime config
Select a cut to view its pipeline status
Select a family to load reference images
| # | Time | Duration | Type | Subject | Action | Camera | Mood |
|---|
What needs to change? This helps when recreating the image.
What needs to change? This helps when regenerating the video.
Browse approved, slots, matrix, used, and DAM references in a centered overlay.
Loading...
Loading...