Part 13 - Demo Dub Delete - the capture and creation cycle
In the previous chapter we looked at production and setting out the vision for the song.
So now we have our spark and spice mixed together and our direction set, how do we go about turning that into a reality?
If you're a proper composer or you're just really good at thinking about musical parts, it might be as simple as writing down everything you want to play, hitting record, and playing it all.
If you're anything like me then you might be quite a long, long way from that.
So I have a 'Demo Dub Delete' cycle, and you might find it helpful too.
Demo
Firstly, and most importantly, capture the essence of the song as completely and accurately as you can.
For me, this involves practicing to a metronome a few times and then recording to a click. As well as being a bit of good discipline, this also makes it easier to add midi parts later if the baseline is solid.
It can get a bit awkward if you have a tempo or time-signature change, ideally work out where those things apply and set up your DAW accordingly. Otherwise just accept that at some point you're either going to be working off grid or doing a fair bit of corrective editing.
Depending on your set up at home, and how quick you're acting, these first parts may end up featuring in the final recording, so personally I always trying and separate out the vocal and instrument parts right from scratch, and always try and properly 'perform' them.
Try not to settle for just an 'OK' take because, as well as being potential material for the final thing, a half-decent take will also give you a bit of passion to work with when you're creating your other parts.
You might be surprised at how often demo takes make it onto the final release.
Dub
OK, you've got your core song parts recorded, the whole thing has a structure, and you've taken a few minutes to mark up the different sections of the song* in the DAW to help navigate things. Now you need to add those extra parts that will take you towards your vision.
At this point the critical question to ask is 'what does it need?' When you're working on your own, probably with templates and a set of 'preferred' instruments (virtual or otherwise), it's very easy to get into a kind of 'production formula'; especially if you're working on an album or EP where you're aiming for a bit of sonic consistency across the record. But you still need to ask this question every time. It will save you a lot of time in the long run if you only create what you need, not all that you can.
But, assuming that I've started with a guitar and vocal track, and that I'm working towards some kind typical-band-style production, then the first thing I'll typically look at adding is drums or percussion.
There are a couple of reasons for this, primarily because I find it much easier to fit a bassline to a drum groove than vice-versa, and secondarily it means I can get rid of the click track and play any future parts over the top of the drums. This feels more natural and allows you to naturally build crescendos, drops and the other artefacts of tension and release with each subsequent part.
Having got the groove, the rhythm instrument and the vocal melody sorted I'll then generally add the bass to give everything a solid foundation and a structure to hang harmonies and other melodic parts from.
And then we ask that 'what does it need?' question again. There is a school of thought that says to compete commercially a song needs at least five different melodic parts or themes that come in and out during the piece. Personally I think that depends a lot on the genre, but it's not a bad rule of thumb for anything that you're expecting to sit in the 'popular' or middle-of-the-road categories.
So let's take our hypothetical example above and assume that the vibe we're going for needs a bit of electric guitar to add to our initial acoustic recording, and a bit of Hammond organ.
Now the easy, and boring, thing to do - and the thing many of us start with - is to take that initial guitar part, maybe a chord sequence, and pretty much duplicate it on the electric guitar and the organ. Maybe add a guitar lick or two or a chord inversion on the Hammond.
And then everything sounds like a mush in the middle, so we pan the electric guitar and organ out to each side and that's kind of OK isn't it? Until we listen in mono** and it all becomes a mush again.
So instead of just duplicating my chords on the guitar I'll put some time in and come up with another melodic part. Probably based around those chords but with less cross-over and probably working in different areas of the fretboard. And then, because I'm not any kind of keyboard player, I'll then go through a heavily manual process of working through those chords in the midi editor, playing with different inversions, and cutting stuff out until I've actually got an interesting melody or two going on there as well.
Pop some backing vocals on the choruses and bob's your uncle.
Right?
Delete
One of the most common, let's call them 'problematic artistic choices' - because they're not really errors or problems per se - that I hear in songs and mixes is one of continuity. And by this I mean a song will typically start with one or two elements, and then things will be gradually added and added as the song goes on, with the producer trying to create an ever 'bigger' sound at the final chorus.
And when you listen to it, inevitably things are added but then never removed, there's too much continuity. So the intro starts with a bit of acoustic guitar, then the rhythm section comes in during the first verse, then the electric guitar and the Hammond at the chorus... and then all of these elements stay for pretty much the whole track. So at the second chorus we need to add more stuff. And then more still at the third. And even more for the repeater. And it all ends up sounding muddy and flat.
So here's another rule of thumb for you: nothing should play all the way through a track.
Even the carrier that we talked about last time, the thing that sets the scene, that can be dropped once the vibe is established. Doesn't have to be for long, but you have to let something out to make space to let something new in.
As we said previously, the easiest way to make something loud is to make the previous section quiet. The easiest way to make something wide and enveloping is to make the previous section narrow and focused. So before you think about what to add in, look at what you can take out.
Be ruthless, go back to the 'what does it need?' question and ask that of every musical part in every part of the song. Sure, your Hammond part may chime beautifully with your electric guitar riff, but it'll have much more impact if it only features a couple of times in the song; not all the way through every verse.
That may be a lovely bass riff leading into the third chorus, but would greater tension be established by dropping it out?
Just because something is nice, or you enjoyed playing it, or it helps the mix balance doesn't mean it needs to be there. Or at least not there as much.
Here's a trio of examples from some of my favourite bands, really just so I have an excuse to go and listen to them again.
Public Service Broadcasting, Go: https://www.youtube.com/watch?v=BHIo6qwJarI
So much of the vibe for the song comes from that fast-strummed*** guitar, but even that really dominant carrier drops out around 2:30 to change the vibe.
The National, I'll Still Destroy You: https://www.youtube.com/watch?v=gPp-3QSXf1g
That drum machine part is the thing that ties all the different parts together but still drops out at key points to allow that contrast and resulting tension and release. And the fact that almost everything drops out at certain points means that there's then plenty of quiet to go loud from.
And finally, a masterclass on dropping things out and going big with very little: Josh Ritter's Thin Blue Flame: https://www.youtube.com/watch?v=EHdEt1tQNFU
There are only four parts to this song; vocal, electric guitar, piano and drums, but even so he's not scared to drop stuff out all over the shop.
So delete, delete, delete. And then maybe some more.
Unless of course you're one of those lucky people who can just listen to a piece and say, "Oh this needs two bars of mellotron in the second verse and some taiko drums in the pre-chorus," or whatever else the song needs.
But if you've already got those skills you're probably not the target audience of this blog... ;)
* Do get into the habit of doing this early on, it will make things much easier.
** Always check your recordings in mono, it will highlight all manner of issues that you can miss if you only listen in stereo.
*** Apparently this song is at this pace because that's as fast as he can physically play it.
Comments
Post a Comment