Three things I learned about the music industry from releasing my own music
Before embarking on the worlds of business and technology, I spent almost a decade as a recording artist, releasing three albums and touring all over Europe in the lower rungs of the alternative scene. Actually, I say “before”, but of course the creative instinct never leaves you, and so this month I found myself readying new music for release for the first time in five years.
They say that there’s no better way to understand a user than to be one. Blokur’s mission is to fulfil the potential of the internet for the creator. So what did I learn from releasing my own music?
Making music at home is almost no restriction at all
With lockdown preventing me from getting to my studio (or the local church where I recorded the strings last time around), it turns out that it is more than possible to achieve your creative goals at home without spending too much money. I invested in a portable vocal booth to disguise the sound of recording vocals in my youngest daughter’s bedroom, a digital version of a classic synth, some audio processing plugins that were on a lockdown discount special and…that’s it. (OK, I did have a good mic and interface already). And if you know the meme, it’s good enough for Four Tet, right?
Besides the technical limitations of recording at home, the other restriction is that you only have yourself to check things against. But I found a huge wealth of information online to help with everything from building techno kick drums to mixing vocals. And when it came to mastering, reaching one of the world’s best engineers was as simple as uploading to an online mastering service.
Getting your music online is easy peasy, but what about finding an audience?
When I last put out an album, securing retail distribution for a CD release was still a key part of the plan even for small indie releases like mine. Sadly many of the indie record stores that supported those releases just don’t exist any more — and those that do are closed as I type this due to the pandemic. This time then, I just uploaded the songs to an online distributor. It took no more than a couple of minutes. I did, though, find myself wondering how many streams I would have to generate just to cover the distribution fee.
The other question is how on earth you make an impact when everybody with a laptop is releasing music into the world (and when playlist pitching can be opaque — at least to a newcomer). For that, in the UK at least, the BBC remains a key supporter of new music. I uploaded the first song to the BBC Uploader on a Thursday, and found it included in a BBC podcast the next Saturday morning. And DJs like Tom Robinson and Huw Stephens continue to be the first place that many songs and artists first get exposed to a mainstream audience.
It’s easy to make mistakes with metadata (even if you’re supposed to be an expert)
I have spent the last three years building a platform that processes, matches and cleans music data. I can tell you the ins and outs of music industry data standards, copyright structures, and data relationships. I have had a hand in designing algorithms for automatically matching songs and resolving data conflicts. And yet I am slightly ashamed to say that I found myself making errors when generating metadata about my own songs.
First, when registering the recordings with my neighbouring rights organisation, I took the option to clone my first song to make it easier to register the second, forgot to edit the ISRC for the new track, and ended up with two different recordings with the same ISRC. Then, I began a registration of the composition with my performing rights organisation only to realise halfway through that I was about to create a duplicate in the system, because I would need to log in separately with my music publisher account to register the song with the publishing information. And to complete the hat-trick, I performed the reverse of my first ISRC error, by failing to notice that the distributor would generate its own ISRC for the track if I did not check a box to let them know that I had one already, thereby detaching the track from its counterpart composition in the PRO database.
These are all things that can be unpicked. But with 40,000 tracks being uploaded to Spotify every day, it’s crazy to think how much data detritus we are sending out into the world. And these are errors — for example, the streaming service ending up with a different ISRC to the one linked to the composition in the PRO database— that can prevent money flowing where it should.
And so — surprise! — the experience of releasing music again made me even more convinced of the need for accurate data to help musicians and music companies get paid. At Blokur we just aspire to do it on a slightly larger scale.
