How to transcribe a video to SRT subtitles

To turn a video into an SRT subtitle file, upload the video to a speech-to-text tool, wait for the timestamped transcript, fix any mistakes, and download it as subtitles. In Voxmint you do this in the signed-in app: upload the audio or video, edit the transcript, then export it as text or subtitles.

This is how the finished transcript looks in the editor, with each line tied to a time:

The rest of this guide covers what to check before you upload, how to edit the transcript so the subtitles read well, and what to do if you need captions for a language the transcriber does not cover.

What an SRT file is

An SRT file is plain text. It is a numbered list of subtitle cues, and each cue has a start time, an end time and one or two lines of text. Most video editors and video sites accept it.

1
00:00:01,000 --> 00:00:03,500
Welcome to the channel.

2
00:00:03,800 --> 00:00:06,200
Today we are fixing a leaky tap.

VTT is a close cousin used on the web. It has the same cues with a small header and a dot instead of a comma in the times. If a site asks for VTT, export that instead.

Step 1: Check which language you have

Speech to text in Voxmint covers Vietnamese, English, most European languages, Indonesian, Malay, Filipino, Korean, Hindi and Turkish. It does not cover Chinese, Japanese or Arabic, so for those you will need another route for the transcript.

Pick the language that is actually spoken. A transcript of mixed-language speech is harder to get right, so split a mixed video into parts if you can.

Step 2: Get clean audio

The transcript is only as good as the sound it hears. Before you upload:

• Remove long music beds or lower them under the voice.
• Use the original video file where possible rather than a screen recording of a screen recording.
• If one speaker is far from the microphone, expect more corrections.

Step 3: Upload and wait

Sign in, open speech to text, and upload the audio or video file. The transcript appears with timestamps when it is ready. You can start reviewing as soon as it opens.

Step 4: Edit the transcript

Do this step slowly. Automatic transcripts are a strong first draft, not a finished product.

1. Play the video and read along. Fix wrong words, names and technical terms.
2. Add punctuation where a sentence needs it. Punctuation tells viewers where to breathe.
3. Check numbers and brand names, which are the usual trouble spots.
4. Remove false starts such as "um, I mean, so" if you want cleaner captions.

Step 5: Make the subtitles easy to read

Good subtitles are shorter than the spoken sentence. These are common guidelines, not strict rules:

• Keep each cue to one or two lines.
• Break lines at natural pauses, not in the middle of a phrase.
• Leave each cue on screen long enough to read it.
• Use plain words and avoid long cues that fill the screen.

Editing the transcript into shorter lines is faster than fixing them afterwards in a video editor.

Step 6: Download SRT or VTT

When the transcript reads correctly, download it as subtitles. Choose SRT for most video editors and platforms, and VTT if a website player asks for it. Keep the plain-text transcript as well. It is useful for show notes, blog posts and search.

Then add the file to your video editor or upload it alongside the video on the platform where you publish.

Subtitles for speech you generate yourself

If the speech in your video was generated rather than recorded, you do not need to transcribe it at all. Voxmint can produce SRT or VTT captions timed to the speech it generates, because it already knows what each word says. Listen to a short script and hear how the timing follows the voice:

Welcome back. Today we are fixing a leaky tap in three simple steps.

Try the same line with your own script:

Try this line in Harper

For voiceover you generate in Voxmint, see text to speech with subtitles.

Common problems

Problem	What to try
Wrong names or terms	Correct them in the transcript before exporting
Cues that are too long	Shorten the lines in the editor
Language is not covered	Write or source the transcript another way, then time it in your video editor
Background music hides speech	Upload a cleaner mix

Can I edit the transcript before downloading?

Yes. The transcript opens in an editor with timestamps, and you can fix text before you export.

Can I get plain text instead of subtitles?

Yes. You can download the transcript as text or as subtitles.

Does it work for Chinese, Japanese or Arabic?

No. Those languages are not covered by speech to text today.

Is the transcript perfect?

No automatic transcript is. Plan a review pass, especially for names and numbers.

Sign in to Voxmint, upload your next video and have an editable transcript ready to export as subtitles.