Faster Transcription
Whether to have text written while you are still speaking.
EnviousWispr normally writes your text once you stop talking. Faster Transcription, which was called Live transcription until recently, writes it while you are still speaking, and it is off unless you turn it on.
Leave it off on Parakeet. On WhisperKit, turn it on if you have picked a language and you often dictate for more than a minute.
Turning it on
Go to Settings > Transcription and switch on Faster Transcription. The question mark beside it explains what changes for the engine you are on.
Looking for words on screen while you speak, rather than text written into your document? That is a different setting: see Live Preview.
Why you should leave it off on Parakeet
Parakeet writes your speech in overlapping pieces and joins them together. The joins are where mistakes appear, and the longer you talk the more joins there are.
Measured on 28 test recordings and a replay of 500 real dictations, against the same audio transcribed the normal way:
- Word errors went from 2.0% to 3.7%.
- Repeated or invented words went from 17 to 51.
- About 1 dictation in 24 lost its final words.
It also saves no time you would notice under a minute. It only gets ahead at around five minutes, which is exactly where it makes the most mistakes.
Why WhisperKit is different
WhisperKit keeps one continuous transcript instead of joining pieces together, so accuracy holds up better. It can still drop a final word now and then.
WhisperKit has to know your language before it can start. If your language is set to auto-detect, EnviousWispr ignores this setting and works the normal way, because guessing the language from the first moment of audio gets it wrong too often.