To begin, simply enter the text that you wish to be dictated into the text box, or alternatively,
upload a text file (.txt). Dictations start with a one second pause so that you can get ready. To
terminate the dictation in progress, just make a change to the text.
Three voices are provided. "United States (F)" generally works best for dictation; however,
the others are also available if you so choose.
Set the WPM to a rate between 10 and 300. At lower rates, the words are fed into the generator
one by one rather than all together so that the rate of dictation can be slowed down infinitely
without reducing intelligibility. At higher rates, this consideration no longer applies and
the text is generated as one continuous sequence. "Threshold" controls when this transition occurs. The
default value is 120 WPM, and it is recommended that you do not deviate too far from this value
for the best quality of audio output. Try it and see what you find optimal; you can set any
threshold value between 100 and 150 WPM.
Note that at speech rates below the threshold, audio generation typically takes longer as
algorithms run to calculate how to slow dictation to the desired rate.