As a journalist, transcribing audio is my least favorite thing to do. It usually takes more time than the actual interview itself. But, it’s necessary to get it right to avoid misquoting or misinterpreting comments.
I’ve made use of some other artificial intelligence-based transcription apps and then moved that text over to Microsoft Word, but Microsoft is now making that process easier for me, other journalists, students and anyone else that needs to process recorded audio.
On Tuesday, the company announced Transcribe in Word, a new feature that allows users to record conversations directly in Word for the web for automatic transcription. The service, an advancement of the Azure Cognitive Services AI platform. detects different speakers and allows users to revisit parts of the recording by playing back the time-stamped audio.
The transcription can even be edited if the AI speech recognition technology isn’t entire accurate, said Dan Parish, principal group PM manager of natural user interface and incubation at Microsoft, in a blog post.
Your transcript will appear alongside the Word document, along with the recording, which enables you to leverage your transcript to create great content in the way that is best for you. Say you want to pull the perfect quote from an interview to support the main point of your story—just click the plus icon on any line of the transcript and voila, the exact quote is inserted. Want to send the entire transcript to your colleague? Simply click “add all to document” and your full transcript will be laid out in Word.
Users can also upload audio or video recorded outside of Word. In person, I usually record audio on my phone. These days, it’s mostly done via Zoom or Microsoft Teams. I typically have to take that audio and send it through a transcription service and then copy and paste it into Word to clean it up.
According to Parish, this would eliminate one step.
Transcribe in Word enables you to stay focused on your conversation in the moment, saves you valuable time and energy by transcribing it for you, and is integrated into Word so you can focus on the message of your document and not fuss around with different windows or applications.
Currently, English is the only language supported for Transcribe in Word for Microsoft 365 subscribers. It’s supported in the new Microsoft Edge and Chrome browsers. There is currently a five hour limit of transcription time per month for uploads and a file size of 200MB.
Transcribe for Office mobile will be coming by the end of the year.
Microsoft’s speech-to-text feature, Dictate, was also upgraded with additional voice commands on desktop and mobile.
Say things like “start list” or “bold last sentence” to let your ideas flow without stopping to adjust your text. Voice commands understand a variety of symbols so you can add things like “ampersand” and “percent sign”, and you don’t have to sound like a robot! We’ve based commands on the way people naturally talk so that you can capture your ideas easily. So saying things like “dot dot dot” when you can’t remember “ellipses” works just as well.
Dictate with voice commands is available in Word for the web and Office mobile for free when signed into a Microsoft account.
Voice commands for Word desktop and Word for Mac users will be coming by the end of the year for Microsoft 365 subscribers.