
TuTranscript
@tutranscript
Diarization says who spoke, not who they are

Add a conversation and speaker label transcription divides it by voice, marking each line with whoever was talking at that point.
A transcript of a conversation is unusable until someone knows who was talking.
Six pieces of attribution that a plain transcript leaves to the reader.
Divide the recording by voice
Vocal character is tracked across the file, and each change of voice starts a new speaker turn.
Mark every stretch of speech
The label sits at the start of the passage it belongs to, not at the top of a section.
Rename a label once
Change speaker two to a real name and the whole page follows, so a speaker label transcription is corrected in one place.
Keep the label with its timing
Each attributed passage holds the moment it was spoken, so a disputed line can be replayed as said.
Admit where a voice is unclear
A passage that could belong to either of two similar voices is treated as uncertain rather than settled.
Hand back attributed text
The labelled page exports as it stands, ready for minutes, a write-up or a second reader.

TuTranscript
@tutranscript


TuTranscript
@tutranscript


TuTranscript
@tutranscript


TuTranscript
@tutranscript


TuTranscript
@tutranscript


TuTranscript
@tutranscript

Three steps from a recording to an attributed page.
A meeting, an interview or a panel goes in as it was recorded, in whatever format it already has.
The recording is transcribed and each change of speaker starts a new turn with its own label.
Replace the numbers with names or roles, then take the page away with the attribution still in place.
Four jobs where the name beside a line is the point.
Grace Whitfield
HR Advisor
Grievance meetings stand or fall on who said what. A speaker label transcription is the difference between a note that can be defended and a summary nobody can check.
Marcus Ibarra
Journalist
Two-hour interviews with an interpreter in the room. The labels keep answers apart from questions, which stops me guessing during the write-up.
Yuki Tanaka
Product Manager
Design reviews run with eight people in the room. Reading a speaker label transcription afterwards shows who pushed back, not only what was decided.
Ruth Osei
Focus Group Moderator
Six participants, one microphone. Renaming each label to a participant code takes a minute, and the rest of the analysis works from clean attributed text.
Accuracy, naming, speaker counts and what leaves the page with the labels.
Upload a conversation and a speaker label transcription separates the voices, so the write-up begins with who said what.
Label a Recording