Great to be able to export text generated from the OpenVino Whisper. However, the export only includes fractional seconds. When doing human interaction it requires HH:MM:SS format.
Could the HH::MM:SS be added as an option? If so, as a new column in the text generated from the Edit > Label > Label Editor > Export.
You don’t say what you mean by “doing human interaction”.
The major subtitle/caption formats all require milliseconds.
A text editor that allows you to make replacements using regular expressions will remove the milliseconds by replacing [.]\d{3}
with nothing.
You may want to export to SRT (use the dropdown that says “txt” in the export label dialog to change the format) and use any subtitling tool to refine your label placement
Thanks. Great to see it does support .SRT.
I was not aware of that. So the HH:MM:SS is supported.
What is preferred in this case is a single line with:
Start Seconds End Seconds Start HH:MMSS Text. or if possible simply:
Start HH:MMSS End HH:MM:SS Text.
Thanks for the reply.
The background is we have a team working on transcribing 60+ minute lectures.
With the transcription accuracy being 80-90% they do need an occasional time reference to correlate the document text with the time on the audio. At present I am importing label.txt to a spreadsheet. For the seconds time column we apply: =TEXT(A1/(246060),“hh:mm:ss”) to convert the column to time then copy sections to a document. So, in this case the single line label text output is the preferred data to use.
Thanks again to you both.
D.
the asterix in the formula was missed, but it is: =TEXT(A1/(86400),“hh:mm:ss”)