Resound and Descript are two products that offer AI-powered audio editing for creators looking to save time and speed up their workflows.
First things first: Hats off to Descript. They’ve made a really powerful product that lets you edit audio and video just like editing a word doc. It’s a powerful tool that we’ve used ourselves in the past. Well done.
Both Resound and Descript are powerful AI-powered tools for creators, so which one is right for you?
Let’s dive in and find out.
Resound is the AI podcast editor for creators. Built by the same team of Professional audio engineers that produced Culpable and 15 other #1 shows on Apple podcasts. Resound automates tedious editing tasks like finding and removing filler sounds, long silences, mixing and mastering your podcast, and much more coming soon.
Resound officially launched in June 2022 but started as an internal project at Resonate Recordings in 2019.
Descript is an audio and video editing app that lets you edit audio and video like you edit a word doc. They offer audio transcription, editing, AI audio and video editing features, and most recently with their acquisition of SquadCast, remote recording.
Descript was founded in 2017 by the former CEO of Groupon, Andrew Mason. Since then they have raised $100M in venture capital.
Resound and Descript have a lot in common. They both offer audio editing features, AI-powered editing features like filler sound detection, and AI mixing and mastering. But there are a few key things that make Resound different.
In summary, Descript is building a text-based audio and video editor for creators. Resound is building the most advanced AI audio editor. If you’re looking for a tool to automate your post-production then Resound is for you. But if you’re not yet convinced, please read on!
One of the core differences between Resound and Descript is the editing experience itself. While Descript offers a transcript-based approach to editing, Resound puts the focus on editing your waveform by displaying edits overtop of the waveform. The purpose of this workflow is to ensure that you have complete control of every single edit you make.
Review each edit one by one, finely adjusting the boundaries as needed. Our unique review process gives you total control over your edits while maintaining the promise of a faster editing experience.
We’ve been building and fine-tuning our own machine learning models since 2019 to specifically identify ums and ahs in your audio. Some might call it obsessive, but we believe in this problem and have the highest standards for solving it with accuracy and fidelity (and trust us, we’re still not done yet, but we’ve come a LONG way).
We didn’t outsource this problem to a third-party speech to text tool, like some tools do. Instead we trained our own machine learning models while collaborating with audio engineers. That means we have greater accuracy of the start and end time of each edit, sound smoother and more professional, and can pass along the additional control of our models to you.
We’re also just getting started. We’ve been hyper-focused on filler sounds (ums, ahs, ers) but will soon build on top of our work to expand to additional problems like filler words (like, so, right, you know), and much more.
Resound is the most advanced AI audio editor. Our mission is to automate post-production for podcasters and creators so they can save time. We’re not interested in building an all-in-one platform. Trust the best hosting platform for distributing your show, and trust Resound for the fastest and most accurate podcast editing. We’ve spent the past eight years editing podcasts, and the past five building the machine-learning models that power Resound.
We believe AI should be used to automate tedious work, but never at the expense of replacing your creative work entirely. Practically speaking, this means that every feature we build in Resound has two audiences in mind: The hobbyist creator and the paid professional audio engineer. Resound started as a project to help empower our internal team at Resonate Recordings work faster, and do less tedious work, but our goal was never to replace those individuals. Rest assured that everything we build will consider both the hobbyist and the professional.
Resound is being built by the same team of professional audio engineers behind Culpable and 15 other #1 shows on Apple Podcasts. The original Resound team was spun out from Resonate Recordings in 2022, led by Jacob Bozarth, and we continue to work closely with other Audio Engineers at Resonate to get oversight and feedback on new features. We also have a unique data agreement between our companies that gives us ethical access to see how professional audio engineers make editing decisions.
One of the obvious ways that Descript is more powerful than Resound is their video editing offering. For video podcasters and creators Resound does not currently offer features in this category.
However, we hope to remedy this in the future, while staying true to our roots as an audio editing tool.
The Resound Editor has gone through many improvements in 2023, but it’s still lacking a few key features that are expected from other Audio Editors. You can’t (yet) see multiple tracks at the same time, arrange tracks on the timeline by dragging them around, or add fade-ins and outs to the end of clips in the editor.
Alternatively, Descript takes the lead in offering more features in their Editor.
Because Resound has been laser-focused on solving the problem of detecting ums and ahs with our own machine learning models we have not yet expanded to detecting other types of mistakes (like, right, you know). But in the near future we plan to add this.
Resound is hyper-focused on automating post-production, which means we don’t offer remote recording, hosting, text-to-speech, or voice cloning services. If you’re looking for a tool that offers these in one place, then Descript wins in this category.
While Resound offers a waveform-first approach to editing, Descript’s main selling point is the option of editing your audio and video like a word doc. In fact, they were the first tool to do it back in 2017 (as far as we know), sparking several others to follow suite like Podcastle, Spext, and others. If you are moving around different sections of an interview, refining a script, or storyboarding, then Descript is especially helpful for these edits.
Descript began as an audio editing tool but expanded to become a tool focused on video editing over the past few years. They offer b-roll stock footage, allow you to import and export video formats, and can easily be used to storyboard, arrange, and edit video footage to your liking.
Another core feature that helped land Descript on the map is Overdub. This feature, powered by their own Lyrebird AI, gives you the ability to synthesize your own voice, which means you can type in words to Descript and get a lifelike creation of your own voice instead of having to re-record parts of your audio. You can also use their built-in AI voices to add voiceover to your show.
Descript lets you quickly import audio and video from third-party sources, and also lets you export your audio to hosting platforms and other places.
This is an area where Resound plans to expand but is currently lacking.
Another side benefit of their transcription and video editing features is that Descript lets you quickly and easily generate video clips for use as Shorts, Reels, and other social platforms. You can even get captions that get embedded within the video to boost engagement and for people who are scrolling social media with the sound off.
Descript’s filler word detection is fairly simple and easy to use, but the main thing it lacks is precision around the boundaries of each edit. Our research indicates they use a speech-to-text tool for finding these words which is fine in theory, but leads to lower-quality results especially for words like um and ah, which SST models are not as good at identifying since they are non-words.
Alternatively, Resound has spent the last 4 years developing proprietary ML models to detect filler sounds with an extremely high level of accuracy in terms of how many we detect and how tight the boundaries are to each edit.
We’ve heard several of our users that have tried both Resound and Descript comment on this. Editing filler words in Descript is a fairly easy process, just a few clicks really. However, once you review your edits it’s not easy to go back through and see which edits were removed and review them one-by-one.
Resound, on the other hand, gives you the ability to review each edit one-by-one in a fast and gamified workflow that’s meant to keep you in control at all times.
As we mentioned above, Resound has some weaknesses like everyone else. But we’re working hard to fix them as quickly as possible. Here are a few things you can expect in Resound soon…
Are we missing something? Request a feature in Resound here.
Resound and Descript are two excellent products that offer AI-powered audio editing.
If your main goal is to clean up your audio and speed up your workflow, we recommend going with Resound.
If you’re looking for a video editing solution that offers AI-powered features, or if you like editing on a transcript instead of a waveform, we recommend Descript.
Lastly, it’s worth noting that this isn’t a binary decision. You can run your audio through Resound to clean up the simple mistakes and enhance the sound, then pull it into Descript to move content around with the transcript-based edits.
Still on the fence? Try Resound for free.