SUBTITLE GENERATION
How to generate subtitles from audio
Speech recognition uses the audio track, but the uploader accepts a video file. MP4, MOV, WebM, and MKV are the drop targets. A loose MP3, WAV, or M4A is rejected until you place that recording inside one of those containers.
You already have a finished picture? Upload that video directly. You do not need to strip the audio out. Generate subtitles from video
Do it in this order
- 01
Put a bare recording into a video file
If the source is only an MP3, WAV, or a voice memo, mux it into an MP4, MOV, WebM, or MKV. A still frame plus the audio is enough. Keep the same duration as the recording so the cue times match what you will publish.
- 02
Upload that file, not the loose audio
Drop the video container on the homepage or the Video to SRT page. The picture, even a still, stays in the browser. SubtitleGenerator extracts audio locally and sends that extract for transcription.
- 03
Set the spoken language, then generate
Choose the language on the recording. Podcasts with two languages need the language that carries the speech you want written down. Hinglish is the mixed Hindi-and-English option. There is no Auto mode.
- 04
Review names, then export the file
Fix Loop is where guest names and product words usually show up. Downloading the SRT requires Pro or Max. If you included a real picture and want captions burned in, export the video instead; Free can do that with a watermark at 720p.
What usually goes wrong
Dropping the MP3 and assuming it failed to transcribe
The file never enters the editor. The accept list is video/mp4, video/quicktime, video/webm, and .mkv. Wrap the audio, then generate.
Transcribing a stereo mix with music under the voice
Background music does not get its own track. Lyrics and speech can land in the same cues. Generate from the voice-forward mix when you have one.
Exporting SRT from a file that still has the slate
Cue times follow the file you uploaded. Trim the container to the episode before you generate, or every later cue will be late.
Questions this page should settle
Can I upload MP3 or WAV directly?
No. Put the audio in an MP4, MOV, WebM, or MKV, then upload that file. The transcription still runs on the audio track.
Does an audio-only container produce a captioned video?
You get timed cues either way. A burned-in export shows whatever picture is in the file. Use SRT when the destination is a podcast player or another editor.