Post written by Jane Duncan.
A comprehensive oral history interview can sometimes run up to 3 hours long. If you’ve ever transcribed an oral history, you know it can take at least 4 hours of typing for every hour of audio, if the speaker is easily understood and the audio is of high quality. Once this first phase of transcription is complete, the finer points of editing are tackled: formatting, proper punctuation, and research of key terminology is done, adding significantly to the overall timeline for the completed transcript.
Trint transcription service (https://trint.com) is a helpful ally in your transcription workflow. The service uses artificial intelligence to “learn” the unique qualities of your speaker’s voice and convert speech to text. Once you upload the audio (or video) file to the easy-to-use interface, Trint will take about 10-15 minutes to convert a 45- minute interview to about 12 pages of text. There will be some work to be done correcting misspellings, but the majority of the interview will be transcribed accurately enough to be well worth your subscription to the service. Trint can also detect vocal changes, formatting the interview in questions and answers, further saving transcription effort.
There are several features that add value to the Trint interface. If your interview contains technical terms, you can upload a list that will add vocabulary to the Trint library improving the accuracy of the speech-to-text conversion. Trint AI algorithm transcribes in 31 languages. Trint offers a variety of pricing options, by annual or monthly subscription, and can be customized for use by individuals or teams.
Photo by TheAngryTeddy (Pixabay.com, licensed under either Creative Commons Zero License for Public Domain or the Pixabay license)