Why an SRT upload disappoints
Google Docs is a document editor. An SRT is a cue list: a number, a line with -->, then a
fragment of a sentence, then a blank line. If you upload interview.srt to Drive and open it
as a Doc, Docs does what any text importer does. It shows the characters. You still have to strip cue
numbers and stitch lines before anyone can comment on a thought instead of on cue 214.
The path that matches how Docs actually works is a real Word file. Convert on the
SRT to Word page, download .docx, upload that file, and choose Google Docs
when Drive asks how to open it. You get paragraphs, optional time ranges, and bold speaker labels when the
caption file had them. Reviewers use comments and suggestions on those paragraphs. That is the whole job.
Nothing in the conversion step sends the transcript to a server. The upload to Drive is a separate choice
you make after the file is on disk. If the recording is confidential, keep the .docx local
and skip Drive. The tool does not require an account to convert.
Upload the .docx
- Paste or drop the SRT, VTT, ASS, or Zoom transcript into the converter. Confirm the preview is paragraphs, not raw cues.
- Set Show timestamps and Merge same speaker for the review you want. Change your mind and download again — do not plan to strip hundreds of clocks inside Docs.
- Click Download .docx. Open it once in Word or LibreOffice if you want to confirm speakers before sharing.
- In Google Drive, click New → File upload and select the
.docx. - Open the uploaded file. If it opens in a preview that is still Word-shaped, use Open with → Google Docs so comments and suggest mode are available.
Drive’s conversion is usually faithful for this kind of file: headings stay headings, bold speaker names
stay bold, italic time ranges stay italic. The font may change from Calibri to Arial or another face Docs
substitutes. Line spacing may loosen. Neither change reintroduces cue numbers. If a speaker line lost its
bold, it was almost always because the SRT never marked a speaker and the preview was plain text — fix the
labels and export again rather than bolding every turn by hand. How those labels get into the file is
covered in interview SRT to Word.
Very long recordings still fit. A two-hour talking-head transcript is small compared with a Docs file that
has images. If the converter refused the paste, the transcript was over the size limit in the browser, and
uploading to Drive would not have parsed the SRT either.
Paste into a Doc you already have
Copy formatted on the converter writes the live preview as HTML plus plain text on the
clipboard. Paste into an existing Google Doc when the transcript is a section of a brief, not the whole
file. Use Paste without formatting only when you explicitly want to inherit the Doc’s styles and lose the
bold speakers. The default paste keeps the structure you saw in the preview.
Clipboard paste is the better move for a quote of a few minutes. File upload is the better move for a full
episode, because you can replace the Doc later by uploading a corrected .docx without
scrolling through a 40-page paste looking for the old version. The walkthrough of both download and copy
is the same one as how to convert SRT to Word.
Comments, suggestions, and what not to comment on
Turn on suggest mode when a translator or a producer should propose wording without destroying the
paragraph. Comments belong on a sentence (“this name is spelled Minh”) or a time range (“this is early
against picture”). They do not belong on a cue number, because the Doc should not have cue numbers.
Timestamps in the Doc are text. Editing 12:04–12:09 to 12:10–12:15 does not move
the subtitle on YouTube or in the player. If the review is about sync, say so in the comment and change
the caption file. If the review is about wording, edit the Doc, then decide whether the caption file must
be updated to match. This converter will not write the Doc back to SRT.
For a bilingual review, build the two-column table before you share the link, or share two Docs and say
which is source. A single merged paragraph of alternating languages is hard to comment on.
Bilingual subtitles to Word explains when to turn merge off so
each pair stays a block you can drop into a table.
A clean reading copy versus a QC copy
Make two downloads if two people will open the Doc. The QC copy keeps timestamps so a comment can cite
18:42. The client copy hides timestamps so the page reads as notes or as a draft article.
Upload them as two files with dull, obvious names (interview-qc.docx,
interview-clean.docx) instead of stripping clocks inside the shared Doc and losing the only
indexed copy. The reasoning is spelled out in
SRT without timestamps.
Share the Doc, not the original video, when the reader only needs the words. Keep the caption file in the
same Drive folder if those readers are also allowed to see timing. Write the source filename in the first
line of the Doc so a later export does not get mistaken for this one.
What tends to go wrong on the first upload
- Uploading the
.srt and then asking reviewers to ignore the arrows. They will not. Convert first. - Pasting the raw cue list into Docs and running a pile of Find-and-Replace rules. One missed arrow line looks like a sentence.
- Leaving merge on for a dual-language SRT, then commenting on a paragraph that mixes both languages.
- Putting a confidential interview in a personal Drive because the converter was local and the next step was automatic. The next step is not automatic. Upload only if that is the system your team uses.
Related workflows
Download the .docx
Set Show timestamps and Merge same speaker before you download — Docs will not strip clocks you left in the file. Then Drive → New → File upload → Open with Google Docs.