Convert a transcript (TXT, SRT or VTT) to a Word document

What goes in

• Plain text. With blank lines between paragraphs, each block becomes a paragraph. Without them, every line is a paragraph. You can force either behaviour.
• SRT and VTT. The cue numbers, timing lines and formatting tags (such as italics) are dropped and the words are kept, one paragraph per cue or joined into paragraphs after a pause of two seconds. Tick the timestamp option and each cue starts with its time in bold, such as [1:05] or [1:02:03].

1
00:00:01,000 --> 00:00:03,500
Good morning everyone.

becomes, with timestamps:   [0:01] Good morning everyone.
becomes, without:           Good morning everyone.

What comes out

A standard Office Open XML document: a ZIP file holding the page, its styles and a little metadata, written in your browser without any library. Word, Google Docs, LibreOffice and Pages all read this format, and we check each file we build by opening it with a document library; open yours and look at the first page. The first paragraph can be a title. Everything else is plain paragraphs, so the formatting is yours to choose afterwards.

What it does not do

It does not transcribe anything: it converts text you already have. It does not add speaker names, headers, page numbers or tables. It writes only .docx, not .doc or PDF. And it does not turn TryScriptorium into a Word exporter: the app's own export formats stay TXT, timestamped TXT, SRT, VTT, Markdown, JSON and CSV, which is why this lives on its own page. If you are looking for the way from audio to a Word file, the route is: transcribe the file, export the TXT or SRT, and convert it here. The longer explanation is on the audio to Word page.

Editing it afterwards

A Word file is a good place for the last human pass. Use Find and Replace to correct a name the speech recogniser got wrong everywhere at once, run the spell checker over the text, and add headings or speaker names where they help the reader. If you kept the timestamps, they stay at the start of each paragraph as bold text, so you can search for them or delete them with a single wildcard replace. For long recordings, splitting the transcript into sections before you convert makes the document easier to navigate; the split subtitles tool does that for an SRT.

Check the result

Open the file and read the first page before you send it on. Fonts for Arabic and other non-Latin scripts come from your computer, not from the file, and a transcript made by software still needs a read against the audio for names, numbers and quotes. To count the words first, use the transcript word counter.