Explore Polyglot

Download now

For Apple Silicon · macOS 14.4+

Local AI. Your Mac.

The Polyglot journal

How to create an SRT subtitle file on Mac

A subtitle file is small. The details inside it matter. Here is a practical workflow for turning your recording into a reviewed SRT file with Polyglot.

Sculptural chrome subtitle bars and a play symbol with cobalt reflections on black.
AI-generated editorial illustration · Polyglot

Start with the recording you can use

To generate new subtitles, start from audio or video. If you already have an SRT, open Translate to translate it without transcribing the recording again. Choose a recording you made or have permission to process. A locally saved file is usually the simplest starting point: the transcription and translation stay on your Mac, and you do not need a video-link download to begin.

Before a longer job, play the file and listen to a section with speech. Check that the intended audio track is audible and that the file is the final edit. A transcript timed to a rough cut will not automatically match a shorter version with a different opening. Keep the original recording and give the project a name that identifies the edit.

Generate the first draft locally

Install Polyglot on an Apple Silicon Mac running macOS 14.4 or later. The subtitle workflow needs at least 8 GB of memory and approximately 2 GB for the speech model, plus 5 GB if you add translation. Start in Create to generate the original subtitles; translation is optional. Download the required models while you are online. After setup, you can process local media without uploading it for transcription or translation.

Import your file and choose whether you need the original-language subtitles alone or translations as well. Automatic speech recognition supports 25 European languages. Translation targets are a separate selection, so a translation option does not mean Polyglot can recognize speech in that language. For a first project, a short excerpt makes it easier to inspect the result before processing a long recording.

Review against the sound, not just the text

Play the recording alongside the generated cues. Check names, numbers, acronyms and any sentence that seems unexpectedly fluent but changes the speaker’s meaning. Correct the original text before reviewing translations. Polyglot keeps source timing and flags affected translations for review when the original text changes; it does not silently replace completed translations with a new draft.

Read each cue at playback speed. A sentence that looks fine on a page may be too long to follow on screen. Translation can also expand the text. Shorten wording only where it preserves meaning. Polyglot preserves source timing during text editing and translation. If every cue needs the same offset, use its explicit timing adjustment and verify the result. For one cue, adjust its Start and End times or drag it in the timeline; timing belongs to the selected language. Automatic drift correction and cue splitting still need suitable tools. The sync guide explains how to distinguish those problems.

Listen for useful sound information as well as speech. Automatic speech transcription is not a complete accessibility review. Where the audience needs a speaker label, music description or meaningful sound cue, add and check that information in your finishing workflow. Do not assume that an unreviewed AI transcript satisfies a captioning standard.

Export SRT and keep the pair together

Export the original-language subtitles, one completed translation, or all completed languages. Polyglot writes UTF-8 SRT files. SRT is a text-based sidecar format containing numbered cues, start and end timestamps, and subtitle text. It is separate from the video; exporting it does not permanently place visible captions into the image.

Keep each SRT with the exact media edit it belongs to. A useful naming pattern is launch-demo-v3.en.srt alongside launch-demo-v3.mp4, with a separate language suffix for each translation. Save the Polyglot project too if you want to continue editing later. Removing a local project does not remove your original media or the subtitle files you exported.

Check the file where people will watch

Open the video and its SRT in the compatible player, editor or publishing service you plan to use. Check the first spoken line, a moment near the middle and the final exchange. Confirm that accented characters display properly, line breaks look sensible and the selected subtitle language matches the file you supplied.

If you cannot see captions, first check whether the player has loaded the sidecar and whether subtitle display is enabled. If the captions drift or start at the wrong moment, confirm that the SRT belongs to the same edit. If you need captions that remain visible everywhere the video plays, use a compatible video editor to burn them in. That final rendering step is outside Polyglot’s SRT-export workflow.

If your destination asks for WebVTT, choose WebVTT subtitles in Polyglot’s Export dialog. The app writes the correct header and timestamp syntax; changing a filename extension alone is not conversion. Keep your reviewed source and check the exported captions in the receiving player. Our SRT versus VTT guide shows matching examples and delivery checks.

Return to the right subtitle track

In Polyglot, Files lists saved original and translated tracks, grouped by project. Open a track to edit it, or open its project to review the matching media and other languages. Use Translate to import an existing SRT or VTT, and Create when you need to generate text from a recording.

Sources & further reading

Polyglot: subtitle generation and editing

Install Polyglot and download models

W3C: understanding prerecorded captions

Published by Polyglot using an AI-assisted editorial workflow. How we prepare and update our guides.

Keep exploring.

Try it on your Mac.

Find the right tools for your words and video, with processing on your Mac.

Download now

Apple Silicon · macOS 14.4 or later