The Future Of Humanity Is Being Human
Posted on 12/16/2017 12:56PM
The Future Of Humanity Is Being Human
Over the last three weeks, I have been involved in two online competitions (Groove3 Master Off 1&2) that pit human Mastering Engineers against online mastering services such as LANDR. It has been interesting, fun and ultimately made me a few more dollars in business (I won the first one-- number two has ended due to some issues with the poll). There has been a lot of conjecture and posturing from both sides. Calls that only a human can improve a final mix or that one service or another is better than a human ME most of the time-- both of which are utter nonsense.
The fact is, technology has revolutionized music, production and every facet of human life. Our industry is no different. The democratization of music production is well documented. Is it good or bad? Yes. So what do we do about it? We adapt. We find a niche and to do all we can to play to our strengths and let the “bots” have their strengths (rather than deny they have any).
A mastering service is fast, cheap and depending on the mix and the specific service, can provide good results. (If it’s a single song (not arranged in an EP, LP etc) and you don’t have a specific sound in mind (just a ballpark). Can an automated service improve your song? Yes, absolutely. Can it beat a human according to personal tastes on a given day? Yes. Two track processing can be done, to some degree of success, with algos. If it is better than a particular human is very subjective and depends on the song, and the taste of the individual listener. In previous tests it suggests humans win more often, but not all the time and that will probably improve for the services over time.
So what is absolutely the domain of humans? Where do we have a niche, or an area of dominance? Places where you have to work WITH people. A human mastering engineer, can take in a stream of requests, references and match it with their own taste-- try and elevate the audio to a tailor made version of the music that brings out all the emotion and intention of the artist. A human can take big risks, we can do a heavily processed “vibe” that really changes the sound and feeling ot the track or we can leave it alone and let it be what it is.
We can manage expectations. We can nurture developing artists, we can provide feedback (when asked) that can improve mixes before they are completed which avoid the need for big mastering moves that can cause side effects. We can recognize genre norms and make sure the track works in that context. We can talk an artist down when they want to give up, or push them when they need to just go back to the drawing board one more time.
Humans are best at being human and managing the very elements that separate us from technology. Emotions, irrational behavior, adapting to odd working habits, prioritizing very disjointed and esoteric ideas(this is a metal record, we want to respect the same principals as Steely Dan… huh?) and turn them into a practical application of art.
Most of all, humans can build trust over time, we can develop long lasting relationships with artists that give them a safe creative space where they can do their best work and trust they are getting an objective opinion they know means well.
Humans also recognize when something is WRONG. An algo will never know when a hi hat count in, a throat clear or a door closing os an error or a choice. Humans can ask, a machine will never know that, even if it can one day detect clips, ticks and other sonic issues.
A human will never make the best EQ choice 100% of the time, or get the most level out of a track without distortion or even get the tightest low-end every time. We can however RESPOND to mistakes and know how to improve and the best way to get there. We can, if we are diligent, make 100% of our customers know we value them and their music, and in the end that is a service worth paying for.
Let young artists with no budget use an automated service, or somebody with no time to wait, or somebody that has grown bitter from too many Mastering Engineers not understanding their needs or getting annoyed at all their revision requests. If a service gives you what you want, use the service. There is room for both if we are honest and I am not threatened by it. People ultimately use me because they like me, my work and the way I work with them.-- not because I can beat a “bot” 99 times out of 100.
Lastly, as of now-- services are not able to do targeted project mastering (if I am wrong, let me know). Balancing eq and loudness over a whole project is a key feature of mastering, and services aren’t able to do that. I am sure that will come, as well as doing DDP (will the CD EVER die), vinyl premastering and tips and tails. As of this moment, those things are still human.
What we can learn from a an automated services, is they don’t judge their clients. It doesn’t get annoyed you are changing your mix twelve times, it just bills you. It doesn’t argue with you or try and impress you with its $10,000 compressor. All of our strengths as humans are our weaknesses, and we have to be diligent as these missteps can effect the image of the entire community.
This is a service industry and good service will always be rare and valuable. If you are a human, focus on being the best human and let the machines do what they do best.
Good write up, John. A little hard to read or follow. Perhaps due to writing so many thoughts rather quickly? But many strong points made. Nicely done. I sorta wonder if a human will compete with an automated service as a human automated service.