Choose a transcript or timed subtitles
Transcribing video turns its spoken audio into written words. A plain transcript is useful for notes and quotations. Timed subtitles also tell a player when to show each line. Neither automatically adds styled captions to the pixels of your video.
| What you need | Download | What to do next |
|---|---|---|
| Readable text for notes, a document or a draft | TXT | Open it in a text editor and correct names, numbers and punctuation. |
| A subtitle track for an editor or upload | SRT | Import it into a compatible editor or platform and check its timing against the same video. |
| Timed captions for a web player | VTT | Use a player that supports WebVTT and verify the displayed cues. |
| Segment data for a custom workflow | JSON | Inspect each segment’s text, start and end values before integrating it. |
Transcribe a video with BlitzReels
- Open the subtitle generator. Sign in and choose a workspace if prompted. Transcription uploads the recording to your workspace for processing; it is not a local-only browser conversion.
- Choose your video. The public upload accepts a source up to 1 GB. Use a recording with audible speech, such as an MP4, MOV or WebM. A container extension alone does not guarantee that every codec inside it is supported.
- Select Upload and generate subtitles. Wait for the upload to finish. Once the job has started, it can continue if you leave the page. Avoid starting another copy because processing is taking longer than expected.
- Read the returned transcript. The result shows the generated text with segment start times. Play the original recording and check names, technical terms, numbers and overlapping speakers.
- Choose your download. Select Download TXT for text, Download SRT or Download VTT for timed cues, or Download JSON for structured segments. Edit the downloaded file in a suitable editor when corrections are needed.
- Check the final use. For a subtitle track, import it beside the same video and play it from beginning to end. For notes or a quotation, compare the final wording with the source.
You need an account and a workspace with available transcription allowance. The tool applies the workspace’s limits; this guide does not promise unlimited free transcription or a fixed processing time. Check current plan terms before uploading a large batch.
How to extract text from an MP4 video
For spoken words, upload the MP4 to the subtitle generator, generate its transcript and choose Download TXT. Check the audio first: an MP4 containing only music or silent slides has no speech to transcribe. Text shown only in the image needs optical character recognition (OCR), which this speech-transcription workflow does not provide.
Use a webinar transcript to prepare short clips
Search the reviewed transcript for a customer question, a product explanation or an objection the speaker answers. Keep enough surrounding context to preserve the meaning. Note the segment start time and verify the excerpt against the recording before preparing webinar clips. Review names, claims and caption timing again after editing.
Review words and timing separately
A correctly recognized sentence can still be hard to read if it appears too early, stays too briefly or breaks in the middle of a name. Check the words first, then play the timed captions with the source.
- Names and terminology: listen again instead of replacing an uncertain word with a plausible guess.
- Numbers and negatives: verify prices, dates and words such as “not” that can reverse the meaning.
- Speaker changes: check who said each line. Do not infer a speaker’s identity from the transcript alone.
- Timing: make sure a line appears with the matching speech and clears before unrelated dialogue.
- Reading on a phone: shorten awkward line breaks and check the final caption size and placement in your editor.
Review the whole short clip before publication. Sampling a few lines can reveal a problem in a long recording, but it does not verify every quotation or caption.
Recover when transcription does not finish
| What you see | Next action |
|---|---|
| The upload is still in progress | Keep the page open until the upload completes. If it fails, check your connection and the 1 GB source limit before retrying. |
| A saved video is still processing | Return with the same account, workspace and browser, or open the video in your media library. A running job does not need another upload. |
| The transcription failed | Read the error. The tool can retry transcription from the saved video, so a new upload is not always necessary. |
| No spoken words were found | Play the source with sound. Check that it contains speech rather than only music, silence or text already on the screen. |
| Words are inaccurate | Compare with the recording, correct the downloaded text and consider a cleaner source with less background noise or overlapping speech. |
| Subtitles drift after editing | Check that the subtitles belong to the exact video version being played. Trimming, removing pauses or changing speed can invalidate the earlier timing. |
Turn the transcript into a short video
Use the transcript to find a complete answer with enough context. Then use the clipping workflow to prepare shorter excerpts and review their captions and framing. A transcript download by itself does not cut your video.
If you already have the finished excerpt, resize it for Shorts or prepare a Reel. A fixed crop can cut off existing captions, so prefer a clean source and check the caption safe-zone guide before export.
Keep transcription, translation and text removal separate
Transcription writes down spoken words. Translation changes the language. Caption styling controls how words look in a video. Removing text already burned into a recording is a separate image-cleanup task. Choose the tool that matches the output you need.
This workflow does not read text from silent slides. If the words exist only as pixels on screen, speech transcription cannot recover them. If your goal is to remove visible subtitles, inspect the separate paid caption-removal tool and review its limits.
Before uploading a recording
Check that you can upload and process the recording, especially when it contains private client material. This transcription workflow saves media in a workspace and uses server processing. Local-only conversion tools on other pages have a different processing path.
Keep the original file and the reviewed transcript together. When the video changes, check whether the subtitle timing needs to be regenerated or adjusted. Do not overwrite a reviewed source with an unchecked AI draft.
