Week 31 was quite a journey, and a good deal of it was painful, so I wanted to write about it a little. It’s all part of the challenge! 

The track started with a simple idea, loosely informed by something I’d heard recently. I put the main parts together very quickly, and in the early phases things felt good – the basic sound was ok, and there was a decent feeling in the groove. 

Somehow I then complicated things, adding too many string parts and other synths. One of the synth lines in particular sounded great, but also quite trance-like – a bursting, bright open saw. So then the drums weren’t keeping up with the energy on that line, so I started switching the drums about a bit – compenstating for the new element I had introduced. 

And then the bassline was too laid back also, so I began trying for a totally new one. At this point, I experienced some dispondancy – nothing was fitting together, the track had lost direction, and its earlier coherence had vanished. 

Also there was a new track (by Marcus Santoro) that I was loving. Here’s the track:

In a most unskillful way, I compared the two tracks, referencing one against the other. The Santoro track was a lot brighter, so I started making mine more bright. But the crazy thing is that I had already balanced everything against a different reference track – one much more suited to what I was originally aiming for. 

Nuts!

I had to step away from the track for a bit. 

When I came back I decided to return to what I’d intended to do in the first place.  After that it became slightly more pleasant to complete. It was still painful though.

So here are the lessons I observe here, learnt through an unpleasant experience: 

Form a strong vision for the direction of the music at the outset. 

Don’t then deviate from this as it will only cause pain and make the music sound worse. The best tracks are coherent, with a simple message. Two messages is too many. 

Find an appropriate reference track, let this be your guide.

I found a great reference track that I used to check the overall tone of this track.

Becoming exciting by another track, then referencing against that was a really dumb idea. Choose one that sounds fantastic, and let it guide you in the choices you make. It’s not about copying the exact track – the one I used for this track is different from mine – very. But it has a similar frequency layout and tone.

The lesson I learnt here was that reference tracks are not just useful in mixing – they can help guide the direction. If Undercatt’s track didn’t have a bright open saw line, why on earth was I putting one in mine?

Don’t leave too many spare parts lying around. 

I had tried a few different synth lines, and I thought they might have worked at one point, so I kept them in on mute. This made the arrangement look bad, and provoked a loss of clarity in my design. 

Do not move on. 

I found designing the pads (strings) difficult in this track. I made some that were ok’ish, but I knew they weren’t quite right. I decided to come back and fix them later – perhaps one other parts were added they would make more sense. 

I see this as a bad idea. Pads are important to me, and I care about the atmosphere of the music I write. Knowing that these were only half baked made me lose faith in the track overall. I should have stuck with pad design until I had created some sounds which were punching their weight instead of dragging down the production process. 

Don’t give up.

I’ll quote, if I may, in full from an excellent book on music production by Dennis DeSantis. He says: 

Problem:

The closer you get to finishing the track, the more you realize that it’s a failure. It will be impossible to turn this into something you’ll be proud of. Why bother finishing it at all? Wouldn’t it make more sense to just abandon it and start over on a completely different project? 

It’s depressing to realize that you’ve made something bad. It’s even more depressing to realize it while you’re still working on it but after it’s beyond any hope of salvation. But in this situation, there are still valid reasons to keep going and finish the track anyway. 

Solution:

Most producers have started far more tracks than they’ve finished. It’s much easier to give up in the middle of a project and move on to something new than it is to see a project through to the very end. 

But what most producers don’t realize is that each stage of the music- making process is itself a thing that requires practice. We get to be better sound designers by designing sounds. We get to be better drum programmers by programming drums. And we get to be better song finishers by finishing songs. Because of this, the more songs we start but don’t finish, the more opportunities we miss out on to practice finishing. And as a result, we might continually improve at various aspects of the early stages, but we’ll never improve at actually getting things done. 

If you realize very early that what you’re working on isn’t going to be successful, you probably have time to change directions and make things better. But if you’re very late in the process of making the track before realizing that it’s not good, it might be too late to fix it without completely gutting it (which is essentially the same as starting over). In these cases, forcing yourself to finish—no matter how painful—is often better than giving up. You’ll not only get practice finishing, but you’ll also get practice failing, in itself a valuable skill to learn in a subjective and unpredictable business like art. The better you get at finishing, and the better you get at coping with failure, the better your chances will be the next time you begin (and, hopefully, finish) a project. 

If the track is really as bad as you think, maybe there is a natural
end point that’s earlier than where you’d stop with a track you were happy with. For example, it might not make sense to get your new track professionally mastered. And it might be a good idea to not share it with the public. Maybe it just goes right back into your Scraps and Sketches (page 74) folder, to be pulled apart for use in other tracks later. But the important thing is that you actually finish the arrangement, if for no other reason than to practice, improve, and experience how it feels to finish. 

Wise words. Taking the time out to sit down with this book for a while really unlocked things for me and stimulated me to finish the track. 

On to the next one. 

Week 7 was a different experience for me. Very often production is more fun in the early phases of making a track: there isn’t much going on – competing, overlapping, muddying – and the sound has a purity. Bit by bit this seems to erode as the track becomes more fully formed, and before you know it, you’re fighting to bring back the clarity you started out with.

This week was the other way around: I spent the first half of the week wondering whether things would come together in the end. Then I spent the remaining few days discovering that they did! If repeatable this would be concrete progress, and a direct consequence of this website and quest.

Perhaps finishing a piece of art is, itself, and art (and a science): distinctive, and a sheer discipline in its own right.

I have added a site-wide player at the bottom of the screen, so the music keeps going when navigating around the site.

It also makes it easier to access all the music on one place.

Pleasing.

Artificial Intelligence gains ground every year, now and can do some things better than humans, even when this wasn’t thought possible. They’re far better than humans at Chess, for example; now used by even the best players to check on their moves.

Despite all this, I don’t think computers will ever produce art in the same way that we do. Music – all art, really – is something special, conveying the very essence of the human condition.

The video above is a recent rendition of Bach’s BWV974 Concerto in D Minor, Adagio. Spotify has 100s of different versions of the exact same piece, and each player interprets it in a slightly different way. Some interpret it in a very different way. I happen to like Ólafsson’s version the very best of all. I can’t quite tell you why – I’d be committing the same mistake that the computer is about to make, below: of attempting to quantify the sublime.

Here’s a computer version I recorded playing that same, wonderful transcription, but with little of the soul:

A computer is a wonderful tool for making music – especially my kind of music. But it is not the composer, not the conductor, and certainly not the musician. A computer is merely an instrument.

Here’s a two minute explanation of why I’ve taken on the challenge of producing a track a week and publishing it here.