Dictation Tips

Dictation is not the quick fix solution to writing that most people assume. Dictation is only successful under the following criteria.

First, the environment must not be too overwhelming that the microphone doesn't pick up competing noises. While the dictation software is surprisingly resilient in isolating the speaker's voice, any unnecessary sounds such as music or background talking will strain the system and potentially affect the outcome of the tool. This is not to say that dictation needs to be conducted in a soundproof environment but rather that results will be best when background noise is reduced.

The second criteria is clarity of voice. The speaker must not mumble or present speech patterns that deviate too much outside of the norm. Speech difficulties such as lisps, stuttering, or even accents may present a barrier to dictation working properly.

Finally, perhaps the greatest barrier is the acquisition of the actual dictation skill. While it really is as easy as just speaking, many people are unprepared when they turn this feature on and in the absence of speech, filler words such as "like", "um", "so" may be inserted either consciously or unconsciously which the computer will of course pick up. I have always recommended utilising this tool sparingly and in short increments at a time (one to two sentences).

In the natural process of speech, context of language is implied through the use of pauses or intonation of voice. In written language, this is represented through the use of punctuation. Most dictation software will recognize the basic punctuation conventions when given the command (e.g. period, comma, question mark) but will not be able to insert it automatically. A recommendation for all users is that they listen to the final results to correct any errors that may have occurred during the transcribing process.

Back to r&w for Chrome overview

Back to r&w for Edge overview