LANDR’s automated mastering service just got better... and it no longer requires a connection to the cloud.
When it comes to AI‑powered audio processing, LANDR were well ahead of the curve. Back in 2014 they released the world’s first automated mastering service, a cloud‑based offering that worked out what processing was required to ‘master’ an uploaded file, and then allowed you to download the result. Naturally, it was greeted with both fascination and fear: LANDR couldn’t possibly do what a skilled mastering engineer could do... could it? And if it could, would it render the mastering profession obsolete? As with pretty much all things AI, the reality was more nuanced than the speculation: it wasn’t the death knell for the mastering profession, but it did do a pretty impressive job of ensuring a mix had a pleasing overall balance, that vocals came across with sufficient clarity, and the stereo spread, dynamics, and loudness were appropriately controlled. In short, if you fed a half‑decent mix into it, LANDR was a quick and pretty consistent way to ‘finalise’ a mix to a better standard than many hobbyists working in home studios could.
LANDR has grown over the years: today, not only do they own VocAlign and Revoice Pro developers Synchro Arts, but agreements with various third parties enable them to offer some impressive subscriptions that include a DAW, a range of sample libraries, video tutorials, plug‑in instruments and effects, and online collaboration and distribution services. They confidently claim that LANDR Studio is “the only subscription you need to create, master and distribute your music” and although some might disagree, it’s not an outlandish claim.
For all that, the essential proposition of their mastering service has remained the same — you bounce your stereo mix, upload it and invite LANDR to do its thing, then download the result. Until now, that is! LANDR’s new Mastering Plugin is available for macOS and Windows and supports the AAX, VST3 and AU plug‑in standards. The USP, of course, remains similar: an AI‑powered process analyses various characteristics of a stereo mix, figures out what processing chain might best be used to master it and then applies that processing. But the analysis and referencing is now performed locally and, once the processing is done, the user has far more control both within the plug‑in and more widely in your DAW. For instance, it’s a simple matter to go back to your mix and raise/lower a fader, or revisit a specific instrument’s processing before inviting LANDR to try again. No printing, no uploading or downloading: it’s all much more conveniently integrated into the DAW‑based workflow.
To get started, you insert the plug‑in on a track (LANDR suggest the master stereo bus) and the GUI informs you that you must initiate playback in your DAW. When it detects that playback has started and there’s a signal, it begins its analysis: on‑screen messages suggest that this entails measuring the frequency response, dynamics and stereo width, but judging by the results, I assume several other factors are assessed behind the scenes. Unlike some other AI tools I’ve tried,...