I was using whisper-1 model to transcribe Chinese sermon subtitle weekly, and it was just fine until just yesterday. I found the generated srt file with totally off timing. The subtitles are separated in chucks with long pauses just as the sentence was still ongoing. The text was okay, just the timing issues. I tried with or without language parameter (language =“zh”) and it did not help. Anyone else experiencing the same?
Related topics
Topic | Replies | Views | Activity | |
---|---|---|---|---|
Whisper API - subtitle timecodes out of sync | 2 | 1798 | January 17, 2024 | |
Timings offset for long audio in whisperX | 0 | 23 | April 18, 2025 | |
Whisper skipping some parts of the audio | 1 | 937 | July 29, 2024 | |
Whisper API seems very slow or doesn't always work | 3 | 3398 | February 8, 2024 | |
Whisper AI doesn't transcribe | 4 | 2127 | December 25, 2024 |