A finished film or music video can still fail a streaming delivery review because the accessibility files were treated as an afterthought. Captions arrive out of sync. Subtitle languages are mislabeled. Audio description is missing from the asset list. A version that worked in the edit suite breaks in the actual player.
Accessibility is part of delivery, not an optional layer added after release. A clean package helps more viewers understand the work, gives festivals and streaming partners fewer files to chase, and makes every future placement easier to manage. This checklist gives independent creators a practical way to prepare captions, subtitles, and audio description before submitting a project.
Know what each accessibility asset does
Captions, subtitles, and audio description solve different viewing needs. Using the terms accurately prevents confusion with vendors, festivals, and distribution partners.
- Closed captions: Text that includes dialogue plus useful sound information, such as music cues, speaker identification, and important off-screen sounds. Viewers can usually turn closed captions on or off.
- Open captions: Text rendered directly into the picture. It is always visible and cannot be disabled.
- Subtitles: Text that translates or transcribes dialogue for viewers who can hear the program audio. Subtitle requirements vary by platform and territory.
- Audio description: A separate narration track that describes important visual information during natural gaps in dialogue.
Do not assume that an English subtitle file is automatically a complete English caption file. If it omits speaker names and relevant sounds, it may not serve the same purpose.
Build an accessibility inventory before export
Create one row for every version of the project. Track the picture master, runtime, frame rate, audio configuration, caption language, subtitle language, and audio-description availability. Include the exact version name used by your editor or post-production team.
This small inventory prevents a common problem: a caption file is timed to a festival cut while the streaming master contains a new logo, a revised credit, or several seconds of black at the beginning. The text may be accurate, but every cue is now offset.
At minimum, record:
- project title and version date
- exact runtime and frame rate
- start timecode or expected first-frame timing
- language and regional variant
- caption or subtitle file format
- whether the file is complete, reviewed, or still in progress
- audio track layout and channel count
- contact person for corrections
Prepare captions from the final picture master
Caption timing should be based on the same master you plan to deliver. If the picture changes, even slightly, repeat the timing review. Do not rely only on a transcript. A transcript captures words, while captions also need timing, speaker context, and meaningful audio cues.
Review captions for these practical issues:
- Accuracy: Names, places, slang, lyrics, and technical terms match the program.
- Synchronization: Cues appear with the speech or sound they describe and disappear at a readable point.
- Speaker clarity: Changes in speaker are understandable, especially when the speaker is off screen.
- Useful sound cues: Relevant sounds are identified without describing every minor noise.
- Readable breaks: Lines are divided at natural phrases rather than in the middle of a name or idea.
- Music handling: Lyrics and important musical information are presented consistently when needed.
Automated transcription can create a useful first pass, but a person familiar with the project should review the final file while watching the finished master. Proper names and overlapping dialogue are frequent sources of errors.
Create subtitles as separate language assets
Keep each subtitle language in its own clearly labeled file. Use consistent language codes or names across the folder, metadata sheet, and delivery email. If a translation was created from an older script rather than the locked film, compare it against the final edit.
A translator needs context, not only lines of dialogue. Share character names, spelling references, the final video, and notes about jokes, cultural references, or on-screen information that affects meaning. Ask the translator to flag lines that cannot be translated literally without losing the intent.
For music videos, confirm how lyrics, repeated choruses, spoken introductions, and non-lyrical vocals should be handled. Also confirm that you have the rights and approvals needed for any translated lyric presentation before making it part of a commercial release.
Plan audio description early enough to do it well
Audio description requires more than reading the screenplay aloud. The writer must decide which visual details matter, then fit concise narration into spaces where it will not cover essential dialogue or sound.
Before recording, provide the describer with the final master, character and location references, the approved spelling of names, and any visual information that should not be revealed too early. Review the script for clarity and spoilers. After recording, mix the narration with the original program audio and listen to the full result at normal playback volume.
Deliver audio description as a clearly identified alternate audio asset unless a partner requests a different configuration. Keep the original mix and described version tied to the same locked picture.
Use file names that survive handoffs
A file called captions-final-final2.srt creates risk. Choose a naming structure that a festival programmer or distribution coordinator can understand without opening the file.
A useful pattern is:
ProjectTitle_Version_Language_AssetType_Date.ext
For example, the asset type might identify English captions, Spanish subtitles, or an English audio-description mix. Avoid special characters that can cause problems during upload. When a correction is made, update the date or version identifier and remove the replaced file from the delivery folder.
Test in a real player, not only the editing software
A file that previews correctly in a caption editor may behave differently after upload. Test the final video and accessibility assets together in a player that supports the intended format.
- Start playback from the beginning and confirm the first cue appears at the right moment.
- Jump to several points in the middle and near the end to check for drift.
- Turn captions or subtitles off and back on.
- Check overlapping dialogue, songs, quiet scenes, and fast exchanges.
- Confirm accented characters and punctuation render correctly.
- Switch to the audio-description track and confirm it remains synchronized.
- Test on more than one screen size when the viewing platform allows it.
Watch the full program at least once with each final accessibility asset active. Spot checks catch obvious problems, but a complete review catches missing sections, duplicated cues, and late changes.
Package the delivery so another person can verify it
Organize the final folder around the locked master. Add separate folders for captions, subtitles, audio, artwork, and metadata only when that structure makes the package easier to understand. Include a short manifest listing every file, its language, its purpose, and its version.
If you are sending the project to more than one destination, keep a source package under your control and create a delivery copy for each partner. Requirements can differ. Do not overwrite the source package to satisfy one platform.
Creators preparing for broader distribution can review the VersusMedia Artist Services page and the VersusMedia FAQ before submitting. Festival teams building an online program can also use the Digital Festival Package overview to plan their showcase workflow.
Run a final accessibility delivery check
- The picture master is locked and matches every timed asset.
- Caption and subtitle languages are labeled consistently.
- Captions include speaker and sound information where needed.
- Translations have been reviewed against the final video.
- The audio-description mix matches the locked runtime.
- All files use clear, versioned names.
- Each asset has been tested in a real playback environment.
- The manifest identifies every included file and contact person.
- The source package is preserved separately from partner-specific copies.
Frequently asked questions
Should captions be burned into the video?
Not by default. Closed captions give viewers control and are commonly delivered as a separate file or track. Open-caption versions can be useful for specific screenings or social clips, but keep an uncaptioned master and separate caption file unless the recipient asks for another setup.
Can automated captions be delivered without review?
They should be treated as a draft. Review names, timing, speaker changes, sound cues, punctuation, lyrics, and overlapping dialogue against the final master before delivery.
Do I need every subtitle language before I begin distribution?
No. Start with the languages required by the platforms, festivals, audiences, or territories in your release plan. Preserve clean source files so additional translations can be created without rebuilding the package.
When should audio description be created?
After picture lock, but early enough for writing, recording, mixing, and a complete quality-control pass. Waiting until the delivery deadline leaves little room to solve timing or clarity problems.
Accessibility work is easiest to manage when it is part of the release plan from the start. Build the inventory, use the locked master, test every asset, and give partners a package they can understand. When your film or music video is ready, create a free VersusMedia account to begin the submission process.