Mastering: computer versus human

Being a mastering engineer, I get asked about my opinions on cloud-based mastering such as Landr etc. quite often.

My opinion?

Mastering is a job that requires perception skills, i.e., has to be done by a human if you want it 100% right. Sure, we have technology to match an EQ curve and loudness figures, but there’s a lot more to it than that. It’s like having your text translated by a computer: it can be OK, but you won’t get that A+ result your work deserves. There can be so many problems that a computer simply cannot “hear”. In addition to that, a computer can’t be “highly familiar with a genre” the way a human can.

I recently had a phone convo with a person who started another cloud mastering service that does it for very a inexpensive price; he knows me and lives in the same country and he wanted to check he’s not stepping on my toes, and our thoughts were similar: we’re not sharing the same client base, because those who want that OK quick master and don’t want to pay a lot go for automated cloud mastering, but those who want the best result that stands up to commercial releases go for human-operated service. You can not discuss the mastering job with the cloud; you can’t say “can you still add a little bit of sheen, please?”, and the cloud cannot tell you “it’d really be best if you turned that one percussion down by 3 dB as it’s sticking out way too much in my opinion”.

Being able to discuss things with the engineer is more important than some might think, and the results of that and an ongoing relationship between a producer and an engineer can yield superb results, and that will never happen with a computer.
I’m not saying one shouldn’t go for a cloud mastering service in the least; I’m saying be aware what your money can buy and go for whatever suits your budget/expectations/needs the best.
Anyone interested in mastering done by yours truly can get in touch and peep my mastering service.