Creating Stock Footage Metadata for Blackbox Using AI Assistance
Preparing accurate and comprehensive metadata is crucial for getting your stock footage seen and licensed. Blackbox requires metadata in a specific CSV format. Here’s a breakdown of the workflow I use, leveraging AI tools like Google Gemini to streamline the process, especially for descriptions and keywords.
Tools Used:
- Video Files edited with DaVinci Resolve (e.g., .mov)
- Formalizer
- Google Sheets (or any spreadsheet software that can export CSV)
- Google AI Studio (featuring Gemini) or a similar Video-to-Text AI tool
- Blackbox Uploader
Step 1: Prepare Your Spreadsheet Template
- Open Blackbox Metadata Template document: Download and open Blackbox Metadata Template document, save a copy and delete any example data. Keep top row with the essential columns:
- File Name
- Description (Max 200 characters, min 3 words – though my prompt aims for 150)
- Keywords (Min 5, Max 49, separated by commas, no repetition)
- Category (e.g., Location & Buildings)
- (Potentially others like Batch Name, Editorial Use)
- List Your Files: Populate the File Name column with the exact filenames of the video clips you intend to upload.
(See image bellow for an example of the final sheet structure)
Step 2: Define Your AI Prompt
Before generating metadata, craft a precise prompt for the AI. This ensures consistency and meets specific formatting requirements. After many attaempts, my standard prompt is:
“Please provide a brief, professional description of this drone footage (up to 150 characters), including camera movements. Do not end with a period. Additionally, list up to 30 relevant, specific, single-word keywords in lowercase, separated by commas.”
(See images for prompt refinement ideas and for the prompt in action)
Step 3: Generate Metadata Using AI (Repeat for Each Video)
- Export Full and Standard Resolution: In Da Vinci Resolve/Premiere export two versions, one in high resolution an another version in SD resolution.
- Select SD Video: In your file manager (like Finder on macOS), identify the next video file you need metadata for an choose it’s Standard Definition (SD) version.
- Upload to AI Tool: Go to your chosen AI tool (e.g., Google AI Studio with Gemini). Upload the selected SD video file.
- Enter Prompt: Paste your pre-defined prompt into the AI interface.
- Run AI: Execute the prompt. The AI will analyse the video and generate the description and keywords.
- Review & Copy: Check the AI’s output for accuracy and relevance. Copy the generated Description and Keywords. (See bellow images)
Step 4: Populate the Spreadsheet
- Paste Metadata: Go back to your Google Sheet. Paste the generated Description and Keywords into the correct row corresponding to the video file name.
- Add Other Info: Fill in the remaining columns like Category for that row.
- Repeat: Continue Steps 3 and 4 for all video files listed in your spreadsheet.
(See image bellow showing the sheet being populated)
Step 5: Final Review and Spell Check
- Review Sheet: Once all rows are filled, quickly scan the descriptions and keywords for consistency and obvious errors.
- Spell Check: Use the built-in Google Sheets spell checker (Tools > Spelling > Spell check) to catch any typos. (See image bellow)
Step 6: Export as CSV
- Download: In Google Sheets, go to File > Download > Comma-separated values (.csv).
- Save: Save the CSV file with a recognizable name. This file is now ready to be uploaded to Blackbox alongside your video files. (See image bellow)
Next Steps:
You would now typically proceed to the Blackbox website or application, upload your batch of video files (similar to the selection shown in image bellow), and then upload the .csv metadata file you just created.
This AI-assisted workflow significantly speeds up the often tedious process of writing descriptions and brainstorming keywords, while maintaining a good level of quality and consistency for Blackbox submissions.
You must be logged in to post a comment.
Leave a comment