Hello Friends,

Journalists, freelance writers, researchers and other staff that need to capture interview details in writing frequently look for an automated solution to transcribe their interviews. It can take hours to transcribe the text resulting from a brief phone interview, especially when an exact transcript is required.
Given its amazing recognition accuracy, many assume that Dragon speech recognition would be an ideal candidate for this task. The fact is that Dragon software was not designed to transcribe multiple voices. However, thousands of Dragon customers have successfully used the software to transcribe interviews using a basic technique called "voice writing" or "parroting." The key is how the interview content is delivered to Dragon.
Many people don’t realize that Dragon software is "speaker dependent" speech recognition. This means the software is designed to accurately transcribe the voice of a single speaker. When drafting an article or report, a user simply speaks out loud the text that he or she wants to appear on the screen. A user can complete articles and reports – either by dictating directly to the computer or by transcribing an electronic audio file of his/her voice from a digital recorder -- in a fraction of the time consumed by manual typing.
This changes when more than one speaker is introduced. Passing an audio file of an interview of two or more people directly to Dragon will not yield acceptable accuracy since the writer’s copy of Dragon is trained to recognize his/her own voice, not the voice of the person being interviewed. The “voice writing” approach delivers content to Dragon using only the voice of the trained speaker.

Thanks and Regards,
Umang Kumar

transcriptionhub.com