Whispered Interview Audio Transcription

I have several interview recordings where the participants drop their voices to a near-whisper right at the moments that matter most. I need those low-level passages isolated, enhanced and then transcribed accurately into clear, well-formatted text. Here is what the job involves: • Isolate and denoise the whispered portions without introducing artifacts or making the speech sound unnatural. • Transcribe the cleaned English dialogue with correct punctuation and speaker labels. • Flag any unintelligible words with time-stamps so I can review them quickly. I typically work with WAV and MP3 files; if you prefer a different format, let me know before we start so I can convert them for you. You’re free to use whichever tools you trust—iZotope RX, Adobe Audition, Audacity, or a machine-learning workflow such as Whisper or DeepSpeech—so long as the final audio and text meet these two acceptance criteria: 1. Whispered speech is audible at a normal listening level with background noise kept to a minimum. 2. Transcription accuracy is consistently above 95 %, with all inaudible spots clearly marked. When you respond, please include a short sample or case study from past work where you performed a similar extraction or transcription task. That will help me gauge whether your approach matches the sound profile of my recordings. Once I review your sample, we can agree on the number of files, milestones and turnaround. I’m ready to get started as soon as I find the right fit.