Thread: Promising Practices

Dragon Naturally Speaking: using dictation software to increase productivity

William Orr and Ann King – Barcelona, Spain

The principal motivation for investing in a speech-to-text application is to increase productivity. Regardless of the underlying reason for using voice rather than writing to produce text (poor typing skills, physical disabilities, repetitive strain injuries, or just plain laziness), the ultimate aim is to increase productivity while maintaining accuracy and quality.

Dragon Naturally Speaking (DNS) is currently the most accurate and reliable program of its kind but, though it does work “out of the box”, people starting to use DNS may encounter certain problems. The information in this presentation aims to be helpful in reducing any initial frustration and in flattening out the learning curve for newcomers.

The presentation will provide a brief demonstration of how the program works. It will also share some useful tips for getting your DNS setup properly trained as quickly as possible, streamlining the training process and overcoming some of the inevitable obstacles that arise when getting a piece of computer code, however ingenious, to correctly interpret idiosyncratic human speech.

The presentation will cover aspects such as getting the right equipment, optimum configuration of the program, difficulties and how they can be overcome, aspects of the work environment to bear in mind, and shortcuts to faster and more accurate speech recognition. It will also cover some of the other potential uses of the program, beyond merely dictating and editing text, that can speed up procedures regularly carried out by editors and translators.


William Orr is a freelance translator and interpreter.

Ann King is a freelance translator and software trainer.