Subscribe

turn it up —

AI can now master your music—and it does shockingly well

Suddenly, everyone can master their own music.

-

AI can now master your music—and it does shockingly well
Enlarge
Aurich Lawson | Getty Images

It was just a stray remark at the end of a guitar lesson last fall.

My teacher is a veteran musician whose band has had both major label and indie record deals, and he loves the analog, the human, the vintage, the imperfect. So it didn't surprise me to learn that he still likes to mix tracks with an old analog board or that he has a long-time "mastering guy" who finalizes the band's albums.

What did surprise me was the comment that he had for some time been testing LANDR, the online music service that offers AI-powered mastering. Pay a monthly fee, upload a well-mixed track, and in a minute, the system spits back a song that hits modern loudness standards and is punched up with additional clarity, EQ, stereo width, and dynamics processing.

After mastering, the end result should sound "more like a record," but this stage of the music-making process has always required subtle value judgments, as every change operates on the entire stereo track. And well-trained humans excel at subtle value judgments.

So I was expecting some line about the slight edge that ears and hands still held over our machine overlords. Instead, I heard: "In the last year, LANDR has improved so much that it now sounds as good as, or in some cases better than, things we've had mastered professionally."

AI-powered mastering systems allow endless tweaking (mastering engineers generally offer a specific number of revisions). They return results within a minute (mastering engineers might need up to a week). They are comparatively cheap (mastering engineers might charge anywhere from $30 up to a few hundred bucks a track, depending on their experience). And now they can sound better than humans? This statement, coming from a guy who won't buy a guitar made after 1975, was high praise indeed.

I flagged the remark as something to investigate later.

A few weeks after our conversation, Apple released version 10.8 of Logic Pro, its flagship digital audio workstation (DAW) and the big sibling to GarageBand. Stuffed inside the update was Mastering Assistant, Apple's own take on AI-powered mastering. If you were a Logic user, you suddenly got this capability for free—and you could run it right inside your laptop, desktop, or iPad.

So between my guitar teacher and Apple, the consensus seemed to be that AI-powered mastering was in fact A Thing That Could Now Be Done Well. Which meant I had to try it.

See, in 2023, I built a small home music studio for under a thousand bucks—gear is ridiculously cheap and high quality compared to the old analog era—and taught myself tracking and mix engineering. This was a surprisingly technical process. It took months before I knew my 1176 from my LA-2A or my large condenser from my dynamic mic or my dBFS from my LUFS. Mic placement alone, especially for complex analog instruments like acoustic guitars, took considerable trial and error to get right.

Alongside this engineering boot camp, I began writing songs as part of my long-sublimated desire to don the tight leather pants and bring real rock'n'roll back from the dead. I liked some of my compositions, too—not always a given in this sort of creative endeavor. The lyrics felt clever, the melodies hummable. I began to record.

But something stood in my way. Even as I learned proper technique and my recordings went from "meh" to "now we're talking," they never quite possessed that pro-level radio sheen. How to get it? Many people mumbled "mastering" like it was a dark art, as though some hidden final alchemy would "glue" your track together and lacquer it in a high-gloss finish, and they said this alchemy could only be performed by grizzled mastering engineers working with special mastering compressors and EQs of almost unfathomable complexity.

Nate Anderson Nate is the deputy editor at Ars Technica. His most recent book is In Emergency, Break Glass: What Nietzsche Can Teach Us About Joyful Living in a Tech-Saturated World, which is much funnier than it sounds.

Promoted Comments

Great article for the home studio enthusiast. I own Ozone and just used it on a country pop track. I have mixed feelings about Mastering. I work my ass off mixing a track for balance and dynamics and then it goes through Mastering where software or a human will EQ, limit, play with the phase, etc.

It's almost insulting, but there's no way around it.

The history of mastering is tied to vinyl record production. There are technical limitations to the energy and frequencies that can be pressing into a hot piece of plastic. THAT has been the traditional role of the Mastering Engineer - to tweak an already good mix for the limitations of vinyl.

Today, Mastering has just become a race to get it loud and bright at -14 lufs.

Don't mind me, I'm just a grumpy old man.
Classically trained, working professional performing musician, and mastering studio owner here.

This unfortunate comment is just called working with a bad mastering engineer, but we’re not all like this. Respecting and enhancing the musical choices of the team who worked on the track before it got to you is something any high quality human mastering engineer will have baked into their DNA. It can be hard to find this kind of collaborator, though.

My suggestion: vet your human mastering engineer by personality and the musical experience to work across multiple genres. Stay away from the grumps and ppl with “signature sounds”, and look for someone who is kind and asks musically collaborative questions before they agree to take on the job. You want someone who will turn down the job if it’s not the best fit for their studio, experience, and artistic bent.

I also suggest sending your track out to a few different engineers to see who can make your music sound its best. Most of us do free test masters nowadays so you can compare directly to LANDR, etc.

This is a both/and, not an either/or. I love that my field is becoming less of a dark art populated by grumps, and these tools are indeed great for learning and they have indeed democratized music production.

I still get many clients who prefer human mastering. There is a creative collaborative element to human mastering that I’m not worried about tech taking away: getting objective ears on your project that can come up with new ideas you might not have thought of is a big reason why humans can be worth it.

But this requires working w an engineer who is collaborative, and also with a musician who wants to be collaborative too. Some musicians only want to work alone and put “good enough” tracks out directly onto the internet purely as a creative outlet, and for those musicians I think algorithmic tools probably are the best option. Certainly nothing wrong with taking this approach.

What any pro will tell you, and my personal opinion, is that music production at the pro level is about putting together a great team that can work together to get something greater than the sum of its parts. The quality bump you get if you can do that successfully is a big part of making your music stand out in a very crowded field.

Music is an inherently collaborative art form, more like a big team making a great TV series or movie than a solo painter working alone in their basement.

But not everyone needs or wants to take this kind of approach.

I do want to note that not all human mastering is prohibitively expensive: charging $100 per track would be $20 each for a 5 piece band. Just do a single and see if you like working with that person, it doesn’t need to break the bank. Or take advantage of a free test master from a human; it will take longer but can be worth it in certain situations.

It’s really the questions: “why are you making music”, and “how much do you value collaboration”? For many people attempting to become professional musicians who get paid to do this, the extra 10-20% quality bump you can get from having a high quality human do your master is completely worth the money. 80% quality really isn’t good enough in the highly competitive professional setting of making music.

The “professional setting” for making music also isn’t some rarified space only Taylor Swift can afford, and I actually do object to that kind of black and white thinking in the article. Many humans that do this charge affordable prices that are cost effective, valuable services their customers appreciate. High quality human production is actually cheaper and more accessible now than it’s ever been in human history.

All of this doesn’t also mean automated tools aren’t great for certain other use cases. As a working pro, I really like that I get the question: “what exactly do you do” less often now that LANDR and Ozone have helped more ppl understand how hard this is to do.

One other thing: in my own work, I’ve been pleasantly surprised at how little young musicians ask for super loud masters. The loudness wars really are over and musicians of all ages are into dynamic contrast now, which is awesome.

Great article, and thanks for bringing some more education to an often misunderstood field.
I use iZotope tools on every single track I produce/master. However, I agree it is by no means a replacement for the human engineer. At the very onset of a project, once I get the main instruments chosen or tracked, I will throw Ozone on the stereo mix and let it do it's auto-learn thing. I do this for two reasons: (1) it does provide a rough sense of how the different instruments play off each other and compliment or overlap, allowing to tweak individual VST instruments during the writing process and (2) I feel it provides additional motivation for the writing process to hear what the end result is capable of being even at the very early stages of an idea. However, when it comes time to break out stems and perform final master, I just about never use any automated settings or presets. At most, I will use the Auto-EQ learning based on another professionally mastered track in the genre I'm targeting, and then tweak from there applying all other modules manually.

There is one other nice thing about these pricier VST tools, especially for beginners. After applying any AI or automated adjustments, these tools will show how the values have been adjusted, as opposed to just uploading a track to an online service and getting a flat file back in return with no explanation of the "magic" behind the scenes. Looking at how the values were adjusted and then tweaking them further afterwards is a great way to grasp certain concepts. I wish these tools were around when I started doing this professionally back in 1999.

But as many experienced professionals know, the real magic is the mastering that happens on individual tracks/stems. I have yet to find a tool that can correctly create the appropriate parallel (NY) compression or mid-side effects and things of that nature.

tl;dr : I typically rely on the AI and auto-learn features to enhance the writing process, but rarely ever use the same settings on the final master.
Thanks for this article, and for sharing your music!

I'm FAR from a professional musician, but I dabbled in Logic with some home recordings before covid, and so during covid I was part of the team that moved our church music ministry online. We ended up doing 14 "large" virtual choir projects in 12 months, plus other smaller musical offerings for weekly streaming services. Many of our large projects had ~30 instruments in the orchestra and the same number of singers, all individually recorded at home on smartphones. Fortunately, I discovered RX and Ozone early: RX to clean up home recordings and Ozone to do some basic mastering so I could focus on the mix and blending 60-80 individual recordings together into something musical.

I agree with others upstream that Ozone 9 tended to over-compress in its first pass, but that wasn't necessarily bad for streaming. I always took it as a starting point and after a few projects I had some presets that I stuck with. Without the initial starting points though, I would have had no idea where to begin with the mastering process. Ozone 9 is not (officially) supported on Sonoma, and I haven't shelled out for upgrades yet. I liked the sound of the Apple-mastered track and will probably give it a try on my next personal projects before automatically upgrading my Izotope suite.

(At church we also took advantage of covid to redo the AV in the sanctuary and now have a great Allen-Heath board at the center of our sound system. I'm barely qualified to even look at it, but learning to mix live performance of everything from traditional choir, modern praise band, and small-scale musical theater productions has been a lot of fun.)

Channel Ars Technica

Related Stories

    Today on Ars