Français
“Typically an interview will take 20 or 30 minutes, and we’re going to use a minute of sound, so Trint helps us quickly find those snippets.”
Richard Coolidge is Senior Managing Producer at PBS NewsHour, and has been using Trint since 2017. As second in command to the Executive Producer, Richard is responsible for managing people, systems, and overseeing workflows and content.
PBS Newshour has two main uses for Trint:
1) Finding key sections of interviews to use in tape packages
2) Recording pre-interviews with potential guests so their team can get a transcript and share it with relevant stakeholders.
“We use Trint to choose which soundbites we're going to use from an interview.”
A typical interview recorded in the field might last up to 30 minutes, but the team might only use a total of 2-3 minutes of material from that interview in the tape package. Trint is a vital tool used in stripping out, editing, and choosing the best soundbites for their broadcast.
Creating transcripts of the audio in seconds means that the NewsHour team can quickly search through the audio from their recording, pinpoint the key quotes that they want to include in the tape package, and share it across their team.
PBS NewsHour’s secondary use of Trint comes into play when the team is interviewing potential guests for their evening show. NewsHour uses Trint to provide pre-interviews to get a sense of what the potential guests might say on whatever issues they’re discussing. NewsHour then uses the pre-interview to help decide which guest we might ask to be on the program, and how correspondents would frame their questions for the on-camera interview.
Running the transcript of that interview through Trint means NewsHour is easily able to transcribe the audio into text. This can be reviewed, and sent to higher-ups to approve the content and themes that the guest intends to touch on.
In the tech-led world, all newsrooms are on a race against the clock to publish their news stories as quickly as they can, without skimping on quality.
Sending journalists out to manually type up transcripts is time consuming and expensive, and with a 90% accuracy rate, PBS now relies on Trint to transcribe their interviews – so all their team needs to do is review it and make any minor corrections. Once upon a time this could have taken 24-48 hours. Now, it can be done in a matter of minutes.
“Our Trint transcripts come back with well over 90% accuracy.”
PBS mostly uses Trint for pre-production. But covering TV, social media, podcasts, and even providing news to their partner radio stations, Trint makes it easier to pull out quotes and stories for each medium.
“Say we’re doing an eight minute package… and we interview four or five people over a space of a week out in the field. After we film each interview, we get the audio off the camera card and run it through Trint to get the audio back with transcripts. Then, the producer can go through them easily, online, without having to print out hundreds of pages.”
Richard started out in 1988, when computers were a fairly new addition to newsrooms – and lacked most of the capabilities we have at our fingertips now. As digital technology evolved, the media slowly embraced it and opened up space for news producers to do their own filming and editing.
Now, producers, reporters, and other members of the media are able to tackle stories that they would never have had quick access to before. And transcribing audio to text in seconds is just another example of that.
As technology advances, it’s the role of journalists to keep on top of the evolutions that can keep them focused on real-time reporting rather than using up their time on the mundane.
“A.I. and audio to text is just another step in a direction where we're able to take advantage of more technology.”
---
