When the exact words matter
Political speeches, campaign rallies, parliamentary addresses, commencement speeches and keynote addresses are all quoted long after they are given. Official written versions are often the prepared text, and speakers depart from it: an added sentence, a changed figure, an ad-lib that becomes the headline. A transcript of the recording shows what was actually said.
Students analysing rhetoric, debate teams preparing rebuttals, journalists and campaign staff all work from that delivered version. The editor puts the player beside the text, so any line can be checked by clicking it.
Crowds, halls and loudspeakers
Speech audio is often harder than it looks. The microphone captures the speaker, but also chanting, applause that runs over the next sentence, and the echo of a public address system in a stadium or a square. Some things to expect:
- Lines spoken into applause or cheering are the ones most likely to be incomplete; listen to those again.
- Echo from big venues blurs word endings; audio repair helps with steady noise but not with echo.
- Crowd chants and call-and-response may come out as partial words or be skipped.
- A speech with an interpreter talking over it can produce a mix of both languages; set the language you want to follow.
Long speeches and other languages
Copy the video's address, press Import from link in the upload box above and paste it. The youtube.com/watch and youtu.be forms both work. RocketScribe shows the title and length first, then fetches only the audio track and transcribes it.
An hour-long address is split into chunks and processed in parallel, so it is ready in minutes. Speeches in any of more than 90 languages work; the transcript is written in the language spoken. There is no built-in translation, so for a speech in Spanish or Arabic, read it in English through the Google Translate link on the transcript or the AI prompt export, and quote the original wording alongside your translation.
Keeping a record
Export PDF for a fixed copy that can't be edited by accident, which suits archives and citations. DOCX with timestamps is the working version for an essay or article. CSV gives one row per segment with its start time if you are comparing several speeches, for example counting how often a theme comes up across a campaign.
If several speakers share the stage, speaker identification on paid plans or with a Minute Pack labels them. Only transcribe videos you have the right to use. A transcript of someone else's video is fine for your own notes and research; republishing it may need the owner's permission.