Create SRT subtitles from audio or video on your Mac

Check that the media contains usable speech

Interviews, courses and screen recordings can all start with the audio track. Sonano transcribes the sound in a video; importing silent footage does not produce a reliable transcript. Play a short passage first to check the level, language, music and overlapping speech.

Download a transcription model and leave storage for the source and exports. Once prepared, local recognition does not require uploading the media to a transcription website. Free has daily usage allowances; see the current plan details.

Create the first transcript

1. Open Sonano’s transcription workflow and import audio or video.
2. Select a prepared Qwen3-ASR or Whisper model and the spoken language.
3. Submit the job and wait for completion. Inspect the error and model status if it fails.
4. Replay the result against the source and correct names, numbers and terminology.

If you are unsure which model to choose, compare a short excerpt using the Qwen3-ASR and Whisper guide. Processing time depends on the model, recording and Mac.

Review each caption after checking the words

An SRT cue contains a number, a start and end time, and the text. The following is an illustrative example with preset text and timing:

1
00:00:02,000 --> 00:00:05,200
Check this part of the recording.

2
00:00:05,600 --> 00:00:08,400
Then review when the captions appear.

Watch for captions that appear late, disappear before the sentence ends, or contain too much text for one screen. Pay particular attention to speaker changes and long pauses. A readable full transcript alone does not confirm good subtitle timing.

Match the subtitles to your media edit

Suppose a sentence starts at ten seconds in the original. Remove a two-second pause before it, and the same sentence begins earlier in the edited version. The original captions will now drift from the speech.

Sonano can export reviewed media edits. Use the corresponding edited subtitle timeline for that version, and the source timeline for the original. Save matching media and subtitles under the same version name to avoid pairing a new video with an old SRT.

Check the file in its destination

Export SRT and import it into the editor or player you will use. For a web text track, choose VTT if the destination requires it; renaming the extension does not convert the format. Check playback near the beginning, middle and end.

A separate subtitle file does not burn text into the picture or translate the original speech. Finish that rendering in your target video editor when needed. See the transcription documentation for the app workflow.