Preview results is a configuration option that enables “interim” speech recognition output to be presented while the AI processing is still finalizing. At times, preview results may show inaccuracies that are corrected in the final results; however, this visual feedback to the user during dictation can provide reassurance that audio processing is working. This also supports users ability to monitor quality of dictation in real time, making it easier to spot potential errors or edits to make.
Do not enable this feature if you only want to display the final speech recognition results and don’t have stringent latency requirements. If, however, you want to lowest latency for presenting speech recognition output to users, then enable preview results.Be sure to only display preview results temporarily and write final result output into the target editor.