Transcript timing ยท 5 min read
Transcription with Timecode: Examples and Review
Compare transcript timestamps, SRT cues and VTT cues with a worked example. Keep transcription time codes aligned with the recording during review.
What is transcription with timecode?
A transcript with timecode pairs words with positions in an audio or video recording. Those positions help you find a quotation, check an unclear phrase or prepare captions. A reading transcript may show only a segment start; a subtitle file needs both the start and end of each cue.
Transcription time codes usually describe elapsed time from the beginning of the file. This is different from production timecode embedded in a camera recording, which may use hours, minutes, seconds and frames. Confirm which system your editor expects before delivering a file.
A timecode transcription example
The following invented example illustrates formatting; it is not a measured transcription result. Suppose a phrase starts at 12 seconds and ends at 15.5 seconds: We still need to check the delivery date.
A reading copy can show [00:12] We still need to check the delivery date. That start marker is enough to locate the passage, but it does not specify how long a caption should remain visible.
In SRT, cue 1 contains a timing line of 00:00:12,000 --> 00:00:15,500 followed by the phrase. A blank line separates that cue from the next. SRT uses a comma before milliseconds.
In WebVTT, the file starts with WEBVTT and a blank line. The corresponding cue timing is 00:00:12.000 --> 00:00:15.500 followed by the same phrase. WebVTT uses a period before milliseconds. An SRT file cannot be turned into a valid VTT file simply by renaming its extension.
Review words and timing separately
First compare the words with the recording. Verify names, numbers, negation and uncertainty before polishing the reading copy. Timestamps help you locate evidence, but a plausible sentence is not evidence that the transcription is correct.
Then play the captions with the exact file they will accompany. Check that each cue begins near the spoken phrase and remains visible long enough to read. Inspect the beginning, middle and end, plus rapid speech or speaker changes.
Do not promise frame-accurate captions from an automatically generated draft. Alignment quality depends on the source recording and the processing result. Review it in the destination player before publication.
Recognize offsets and drift
If every cue is late by the same amount, inspect whether an intro was added or removed. A constant offset can often be addressed by shifting all cue times together after confirming the amount.
If the delay grows during playback, a single shift will not solve it. Check that the transcript and video refer to the same edit and duration. If the delay begins after a cut, inspect that edit boundary before changing the whole file.
Keep a copy of the original captions. Verify any timing repair in the actual player, including the final minute, rather than judging it from the first cue alone.
Choose the file for its destination
Use a reading transcript for notes and quote review. Use SRT when your editing or publishing tool requests it, and VTT when your web video workflow requests WebVTT. Check the destination documentation rather than assuming every player accepts both.
Keep the reviewed text connected to the source until your quotations and captions are approved. If the video is trimmed or rearranged, review the timing again; a transcript aligned to the old edit will no longer match every passage.
Use Wordtake for a timestamped draft
Upload a recording or import a supported public media link in Wordtake, then review the transcript and export TXT, SRT or VTT. An account and credits are required for processing. Check the recording before publishing quotes or captions.
Wordtake processes the uploaded media; it does not preserve or claim to read embedded camera frame timecode. The examples above explain elapsed transcript timing and caption cues.