A Caption Glossary vs Automatic Spelling: Fix Names Consistently

A small approved glossary prevents repeated mistakes in product names, people, acronyms and technical terms. Automatic transcription may produce a plausible common word instead of the intended name. Use the glossary as a review aid or supported tool input, but verify each occurrence against the audio rather than applying blind replacements across the project.

Documentation basis: Subtitle Edit official repository; W3C caption guidance. The recommendations below are editorial analysis and original teaching exercises, not a hands-on product benchmark or a forecast of search performance.

What belongs in a useful caption glossary?

Include the approved spelling, common misrecognitions, pronunciation notes and source of approval. Keep it limited to recurring or consequential terms. A concise list is easier to use than a large unmaintained dictionary containing old campaigns and conflicting spellings.

Who should approve names and claims?

Use the client brief, official product source or the named person's confirmed preference where appropriate. Do not infer brand capitalization from a model's output. Record the approval source so another editor can resolve a disagreement without inventing a new convention.

Can a prompt guarantee correct spelling?

No. Some workflows support vocabulary hints or related controls, but the output still needs checking. Verify the actual feature rather than assuming a general instruction changes recognition behavior. A hint can help while still being ignored or applied in the wrong context.

Why are global replacements risky?

The same text may be correct in one sentence and a misrecognition in another. Review matches with surrounding audio before replacing them. Keep an undoable copy and check that punctuation, timing and line wrapping remain sensible after the correction.

How should the glossary evolve?

Add recurring errors after review and remove obsolete entries when the project changes. Keep language-specific spellings distinct. Use the glossary in the final quality check so a corrected name is not accidentally changed back by a later transcript regeneration.

A practical exercise

List five names from an upcoming video and their approved spellings. Transcribe a short sample containing them, then mark every occurrence in context. Correct the errors and check wrapping. Save the glossary with the project so the next editor does not repeat the same research.

Before you decide

• Use approved spellings.
• Review replacements in context.
• Retain the glossary through revisions.

Sources and scope

Sources checked October 24, 2026. Follow the current documentation for your exact model, application, account and destination.

• Subtitle Edit official repository
• W3C caption guidance
• OpenAI Whisper repository

A CAPTION GLOSSARY VS AUTOMATIC SPELLING: FIX NAMES CONSISTENTLY
Original Creator Lane decision worksheet | 2026-10-24

TASK
Audience / project:
Required outcome:
Current workflow and its specific limitation:
Options actually available without a new purchase:
Source material and permission:

DECISION QUESTIONS
What belongs in a useful caption glossary?
Your evidence / answer:
Unresolved detail / owner:

Who should approve names and claims?
Your evidence / answer:
Unresolved detail / owner:

Can a prompt guarantee correct spelling?
Your evidence / answer:
Unresolved detail / owner:

Why are global replacements risky?
Your evidence / answer:
Unresolved detail / owner:

How should the glossary evolve?
Your evidence / answer:
Unresolved detail / owner:


PRACTICE BRIEF
List five names from an upcoming video and their approved spellings. Transcribe a short sample containing them, then mark every occurrence in context. Correct the errors and check wrapping. Save the glossary with the project so the next editor does not repeat the same research.

ACCEPTANCE CHECKS
[ ] Use approved spellings
Evidence:
[ ] Review replacements in context
Evidence:
[ ] Retain the glossary through revisions
Evidence:

RESULT LOG
Option / exact version / account or interface:
Date / source URLs checked:
Input / settings / output version:
What was actually tested (leave blank if not tested):
Accepted results / failures / correction time:
Actual charges, if any (do not estimate as measured):
Privacy / permissions / delivery checks:
Decision and main limitation:
Next action / owner:
Recheck trigger:

REFERENCES
Subtitle Edit official repository: https://github.com/SubtitleEdit/subtitleedit
W3C caption guidance: https://www.w3.org/WAI/media/av/captions/
OpenAI Whisper repository: https://github.com/openai/whisper