Transcription Guidelines
Follow these guidelines to create high-quality, consistent transcripts that serve our DeafBlind community.
Transcript Styles
Each video can have multiple transcript styles. Choose the style that best fits your skills and the video's content.
Verbatim
Word-for-word transcription of what is signed, including:
- All glossed signs (typically in CAPS)
- Filler signs and false starts
- Non-manual markers in brackets [raised eyebrows]
- Fingerspelling noted with #WORD format
Best for: Linguistic analysis, ASL learners studying grammar
Cleaned
Verbatim style but with:
- False starts removed
- Filler signs removed
- Grammar smoothed for readability
- Key non-manual markers retained
Best for: General reading, understanding content
Summary
Brief overview capturing:
- Main topics and key points
- Important information
- General tone and purpose
- Typically 10-20% of verbatim length
Best for: Quick reference, deciding whether to read full transcript
Meaning-for-Meaning
Conceptual translation that:
- Captures the intent and emotion
- Reads naturally in English
- Preserves cultural context
- Conveys what the signer meant, not just what they signed
Best for: Understanding the message as the signer intended
Podcast Style
Narrative format that:
- Reads like a story or article
- Works well with text-to-speech
- Includes context and transitions
- May include brief descriptions of visual elements
Best for: Audio consumption, casual reading
Quality Standards
Accuracy
- Capture the full meaning of the content
- Don't skip sections or summarize when doing verbatim/cleaned styles
- Use [unclear] for portions you cannot understand
- Research unfamiliar name signs or references
Formatting
- Use paragraphs to break up content (roughly every 3-5 sentences)
- Use headings if the video has clear sections
- Maintain consistent formatting throughout
- Spell check your work
Cultural Sensitivity
- Preserve Deaf cultural references and context
- Don't "correct" ASL grammar to English grammar in verbatim styles
- Include cultural context notes when helpful
- Respect the signer's identity and perspective
Using AI Tools
AI-assisted transcription is welcome, with these requirements:
- Always review and edit AI-generated content
- Disclose the tools used in your submission
- Verify accuracy - AI often misses ASL nuances
- Add cultural context that AI may miss
Common AI tools:
- OpenAI Whisper (for voice interpretation audio)
- GPT-4 or Claude (for cleaning/formatting)
- Google's ASL recognition (experimental)
Submission Process
- Claim a video from the catalog
- Watch fully before starting
- Create transcript following style guidelines
- Self-review for errors and clarity
- Submit for steward review
- Respond to feedback if revisions requested
Common Issues to Avoid
- Incomplete transcripts - Include the entire video content
- Missing context - Note who is signing and relevant background
- Over-editing - In verbatim style, preserve original signing
- Under-editing - In meaning-for-meaning, make it readable English
- Ignoring visuals - Note important visual elements
← Back to Contributor Portal
Contributor Application
Loading application status...
Once your application is approved, you’ll have access to the task queue and transcription tools. Read the guidelines in the meantime.
Log In
Access your dashboard. By invitation only.
Authentication Required
Please log in to access the contributor portal.