For all the technical toys available to assist transcribers in an accurate transcription, there is one tool that many transcribers take for granted… the human brain.
In terms of voice recognition software, the brain is the only tool that transcribers can rely upon completely. The brain cannot be used as a last resort. It must be used first and foremost. When we cannot discern one phrase from another, we may enlist the help of voice-to-text software, where the computer displays what it thinks it hears in text form.
Problems with Voice-to-Text
The problem with this is that it can be quite inaccurate. For example, the word ‘cap’ may be clear in the audio file, and in the absence of other syllables, may be displayed correctly through text. But if another person is speaking simultaneously and ends their statement or word with a ‘S’, the computer may display ‘scalp,’ depending on whether or not the ‘L’ is present in the background. Even if the ‘L’ is not present, it’s ultimately the computer’s best guess.
The brain enables us to make our best guess based on computer models. An attribute that computer models do not possess is continuity.
Voice Recognition Software
A computer is only as smart as we tell it to be. Voice recognition software compartmentalizes pitch and volume, among other things to identify whose voice is whose. When voices have a similar pitch and volume, the computer may get confused and blurt out words or phrases that don’t make sense. On the other hand, the human brain is trained in audio continuity. Using continuity, we can listen for one voice to the exclusion of the next, or even multiple voices.
Transcribers may not have all the information about what they are transcribing before they sit down. Many events that would need transcribing are televised. If a transcriber is confused about who is speaking, they can go online and search for the corresponding video. Transcribers are well-trained in reading lips. Transcription software does not have the capacity to remind us of our resources. Only the brain can do that.
4,896 total views, 6 views today
Beth Worthy is the Director of Operations for GMR Transcription Services, Inc an Orange County, California based company that has been providing accurate and affordable transcription services since 2004. GMR Transcription has worked with over 8,000 clients spanning myriad industries and prides itself on its customer service and quick turnaround time. Their services include audio transcription, video transcription, and digital transcription, as well as Spanish and Mandarin translation. Google +