Amharic Field Recordings Transcription

I have a collection of Amharic field recordings gathered for ongoing academic research, and I need every minute rendered into clear, accurate text. Because these are natural, on-site captures, you can expect background noise, occasional cross-talk and a variety of regional accents, so close attention and native-level fluency are essential.

The final text should be delivered in a Word document, written in standard Amharic Unicode (e.g., Abyssinica SIL) so it remains fully searchable and copy-safe for later linguistic analysis. Clean-verbatim style is preferred: include filler words when they add meaning, but leave out repeated false starts and obvious hesitations. No translation—just straight transcription.

Deliverables
• One .docx file for each audio file, named to match the original filename
• Speaker labels and time-stamps every 30 seconds
• A short log noting any inaudible or uncertain segments

Acceptance criteria: 98 % or higher word-level accuracy, consistent speaker labeling, and no unmarked gaps longer than three seconds.

If you regularly work with tools like Express Scribe, ELAN, oTranscribe or similar, mention that along with how many minutes you can reliably handle per day and your expected turnaround for the first batch. I review the first sample quickly so we can align on formatting before the rest of the recordings move forward.