Create Searchable Live Audio Transcriptions With Otter
iOS / Android: There is one part of reporting that I absolutely despise: transcribing. Whenever I interview people, I always record the interview on audio so I can come back and listen to it later. It’s a great idea in the sense that I can talk to people like real people, without a laptop in front of my face, but a bad idea in the sense that after that I have to relive this interview, four words at a time, because exactly so much I can type while listening to the recording before I have to pause to catch up. In fact, I record interviews because the same thing happens in real life, so I would never have quotes if I relied on my own skills. Now an app called Otter Voice Notes offers one way to solve the problem.
In recent years, several different services have emerged that suggest using artificial intelligence to decrypt things for poor people like me. I tried them all with high hopes, but was pretty much disappointed until I took the Otter for a test drive this week.
The application was created for the purpose of using it in meetings. Created by AISense, it uses AI to decode the audio it hears in real time, as well as to identify different speakers and extract keywords from a conversation. The beta version debuted in March, but the company is making it public today and offering users the option to purchase the new Premium version.
I’ve tried it in several job interviews, and while it wasn’t perfect, it turned out to be much better than some of the other options I’ve tried there. Also cheaper. Prior to Otter, I used Trint, which is useful but costs $ 15 per hour of recording. With Otter, you get 10 hours of transcription for free every month.
Like most AI-powered transcription apps, it works best when the sound is recorded close to the microphone rather than farther away, and when the speaker does not have a strong accent. I tried to transcribe Dr. Martin Luther King’s “I Have a Dream” speech several times using the same YouTube video, and when the phone was next to the speaker, the transcription was accurate. As I crossed the room, it got a little smaller.
When you record something like an interview, it extracts the keywords and puts them at the top of the page; touch them and you can see where they happened in the conversation. Similarly, you can also search for specific terms. The idea is that if you record an entire meeting with him and then want to remember what was said about the Jenkins project, you can find Jenkins and quickly find not only a big conversation about it in the middle of the meeting, but a shorter mention … closer to the end.
The app really works better for things like meetings, when you want to go back to a certain milestone, than interviews with reporters when you are looking for accurate quotes, although it can handle both.
In addition to recording, the app also offers the ability to include images in a string, so you can take a picture of a whiteboard or slide and display it at the moment you took the picture in transcription. It also offers text and audio export for sharing, and will offer call recording as an Android exclusive later this month.
Premium users (read: people who have tons of appointments to record) can get 6,000 minutes of recording per month, up from 600. In June, they’ll also get some advanced export options, including the ability to include or exclude speaker tags. timestamps during export, as well as the ability to export in txt, pdf and vtt formats, as well as in mp3, m4a and wma files.
You can try it now by downloading the app from the App Store or Google Play . The premium version is priced at $ 79.99 per year or $ 9.99 per month. Students with a .edu email address can subscribe for a discounted price of $ 2.99 per month.