There are a few automated mastering services available now for musicians on a budget who need to get their mixes mastered. Here’s our (admittedly biased) take…
The best approach to mastering is always to “do no harm” and to respect the sound, feel and emotion that exists in the song already. However, there is always the possibility that something in the mix is preventing the music reaching its full potential and this is where working with a human mastering engineer is essential.
The problem with using automated mastering AI services is that if there is an issue in your mix – an unwanted click or crackle that should’ve been edited out, too much low end energy, an essential element of the song getting lost in the mix – the algorithm can’t email you to say “hey, you might want to turn that Bassoon down a bit before the second chorus” – it’ll just assume you intended it to be that way and blindly apply bucket-loads of compression and EQ regardless of the nature of the music you’ve fed it.
That last sentence is maybe a little unfair. There are some automated services that can provide decent sounding results. But whether they will work for your music really is down to luck. It’s rather like using a preset patch in a compressor plugin. You might get lucky and find it makes your drums sound amazing. Or you might not and it stomps all over them – or does nothing at all. Some human tweaking is almost always required to get the best out of a preset.
It seems that with everything AI is being programmed to do, its abilities will only improve with the technology. But I don’t think it’ll ever be possible to program an algorithm to be sympathetic to your needs as an artist, to have compassion and respect for a fellow musician, or to understand fully the complex and highly nuanced ways we talk about music.
The best approach to mastering is always to “do no harm”. And with this in mind, the best thing a human mastering engineer can provide is not fancy analogue compression, spectral processing, or high-end studio gear. It’s a fresh pair of ears and an unbiased mind. And a friendly voice saying “Hey, did you mean for that Kazoo solo to be that loud??”