Open vs closed captions: choose before you export

Start with the viewer’s controls

Use closed captions when your destination supports a selectable text track and viewers need to turn it on, change language or adjust its appearance. Choose open captions when the intended presentation requires text to remain visible. Burning subtitles into a video is one way to create open captions: the letters become part of the picture. Keep a clean video master and the editable timed text whichever delivery you choose.

Begin with media you own or have permission to edit and distribute. Ask the recipient what their actual player accepts, whether viewers can reach its caption controls, and whether the same video will be reused elsewhere. A file that looks right on your Mac can behave differently after upload, embedding or a change of player. Decide for that destination, then check the delivered result.

Separate visibility, storage and language

W3C distinguishes closed captions, which viewers can show or hide, from open captions, which are always displayed. That describes the viewing experience. Storage is another question: selectable text may arrive as a separate SRT or WebVTT file, or as an embedded subtitle track inside a compatible video container. An embedded track is not necessarily burned into the picture. Renaming an SRT file to MP4 does not combine it with the video.

The words captions and subtitles also vary by region. Here, captions means text that conveys speech and meaningful audio information; subtitles may also mean translated dialogue. Neither label tells you whether the text is selectable. A translated track can be closed, and an English caption can be burned in. Specify language, content and delivery separately when asking someone for a file.

Choose selectable tracks for flexibility

Selectable tracks are a useful default for a player that supports them. One clean video can be paired with reviewed language files. A wording correction can be delivered as a replacement track instead of rebuilding the picture. Some players let viewers choose text size, color or position, although W3C warns that support varies. Check those controls instead of assuming every player exposes them.

An SRT sitting beside the video is only an input to this workflow. It must be loaded by the player or uploaded and associated with the right video and language. YouTube’s own instructions, for example, include a timed-file upload path in its subtitle workflow. That is a separate publishing step; exporting a file from Polyglot does not upload or enable it on YouTube. For another destination, use that service’s current instructions.

Choose burn-in for a deliberate fixed presentation

A burned-in version can be appropriate when a display has no usable caption controls, the recipient explicitly requests always-visible wording, or your intended presentation needs a fixed design. HandBrake’s documentation describes hard burn as permanently writing subtitles over the image. Its selectable-track alternative still requires compatible playback software. These are different outputs, not two names for the same export.

The permanence creates practical tradeoffs. Viewers cannot use a caption toggle to hide text that is already in the picture, and the player cannot restyle those letters as a text track. A correction needs a new rendered video. A different language usually needs another derivative. Resizing, compression and cropping also affect the caption pixels, so inspect the final encoded and uploaded version at its smallest intended viewing size.

Plan one master and clearly named derivatives

Imagine a fictional, permission-cleared sixty-second studio tour. Its website player supports selectable captions in English and Dutch. A separate unattended display needs English text to remain visible. Keep the same approved edit as the starting point, deliver the website’s language tracks separately, and create a distinct burned-in copy for the display. A naming plan like this is a manual handoff convention, not an automatic Polyglot feature.

Record which video edit each text file matches, the language, who still needs to review it and where it will play. Do not silently substitute a new cut while reusing the old timestamps. If the display needs a different crop or duration, treat it as another edit and review its matching captions. Give the recipient an explicit choice of files instead of a folder full of ambiguous final-final revisions.

tour-v3-clean.mp4
tour-v3-en.srt
tour-v3-nl.srt
tour-v3-open-en.mp4

Website: clean video + tracks
Display: open-en video

Prepare reviewed text before rendering video

In Polyglot, start from authorized media or an existing subtitle file, review the timed text, then export the SRT or WebVTT required by the destination. Check names, numbers, missing words and translated meaning against the recording. For captions intended to convey the audio, also review meaningful sounds and speaker context; speech recognition alone is not that complete editorial check.

Polyglot exports separate subtitle files; it does not burn captions into video. If you need that output, take the reviewed SRT and clean video into a separate encoder that supports burn-in, such as HandBrake. Select the intended language and save a new video copy. This guide describes the handoff, not a tested encoder preset or a built-in integration. Keep the original master and subtitle project available for corrections.

Test the actual destination, including double captions

For selectable captions, load the intended video and track in the receiving player. Turn captions off and on, choose each language, and check an early, middle and late passage. Confirm that the caption menu’s language label matches the text. Where appearance controls exist, try a larger text setting and check that captions remain readable. Test on a phone-sized view as well as the desktop layout.

For burned-in delivery, watch the final derivative after any upload processing. Check the smallest display, the final crop, bright and busy backgrounds, and important objects near the caption area. If the destination also enables a selectable or automatic track, it may show a second layer of words over the burned-in text. Check both on/off states and resolve that duplication deliberately; the clean master with a reviewed selectable track may be the better deliverable.

Watch once with sound off. Follow the message through speaker changes and meaningful sounds, and note anything the text leaves unclear. Neither always-visible words nor a working CC button proves that the captions are accurate or that the complete video meets an accessibility standard. Keep unresolved passages in the review notes before approving delivery.

Make the next correction from the clean master

If a name changes after delivery, correct the editable subtitle source first. Replace and verify the selectable track where supported. For the burned-in copy, render again from the clean master with the corrected text, then repeat the destination checks. Painting new words over old captions is a poor substitute for retaining a clean source.

If the only surviving file has burned-in text, a caption toggle cannot recover the original picture behind it. Ask for the clean source before promising a removable or restylable version. Keeping the master, reviewed timed text and delivery notes together makes that limitation avoidable on the next project.

Keep exploring.