After the Covid outbreak, many conferences and similar events were moved online or into a hybrid format. For me as researcher, this means that I now only rarely present my research outcomes live in person. Instead, I record more and more videos of my conferences presentations to share online as part of online conferences. Some of them are also on YouTube.
I noticed that I tend to use filler words like “uhm”, “äh”, “like”, etc, which are even more annoying in recorded and publishded videos than in live, in-person presentations. Of course, I practice my talks before recording, but it is difficult to get rid of bloopers and filler words completely without editing.
Since I am not used to video editing, manually searching for and removing filler words in long presentations would have meant hours of work. Hence, I searched for some tool to support this editing process and found Descript, which does exactly what I needed: It can record presentations, converts speech to text, automatically detects and removes filler words, improves audio, and more. To edit and remove parts of the videos, simply delete the corresponding auto-generated text like you would in Word.
Demonstration
Here is a short demonstration of how it works:
Of course, it is not perfect yet and sometimes words are recognized incorrectly. But overall, I was very impressed by how well it usually works, plus there are frequent updates and improvements. It has certainly saved me hours of editing and significantly improved my videos. Since it took me a while to find the tool, I wanted to spread the word in this blog post, hopefully saving others some time.
Product Video
For more available features, check out the product video/advertisement, which I find very entertaining:
Try It Yourself
You can try Descript for free for some hours (but with watermark, I think). After that, there is a discounted plan for students/non-profit, which might be useful, or more powerful plans for creators, professionals, or enterprises.