Today I'm focusing on work flow – the things I've learned and incorporated over the years to make the process of creating and finishing songs as efficient as possible.
The main difference between what I do now and what I used to do lies in my mindset. I used to open up my sequencer with a “let's be creative”, or “let's see what happens” mindset. It was quite enjoyable, experimenting and playing around with different sounds, and no doubt built experience in what works and what doesn't (though this learning process probably isn't optimal, in terms of the time spent). I don't do that any more. Now I have a clear goal when I'm in the studio. I know what I want to get out of the session – whether it be composing a melody, building up a mix, adding finishing touches or experimenting and playing around with different bass sounds (yes, I still do it, but now I'm aware that's what I'll be doing when I start the session!)
Something I have made extensive use of during the years are production templates. These are constantly changing – if I find I'm using a certain limiter more and more, at some point I'll modify my template so the limiter is ready to use without me needing to load it in.
Some people are against using templates, claiming they will lead to “cookie cutter songs” - every song which uses the template will sound pretty much the same. To a degree, they're correct, though it depends on how broadly you interpret the expression “pretty much”. I think “pretty much” could well apply to songs by people who don't use templates, unless they change all their tools and learn a new style of mixing between each song. However, if we are using a template and there are certain settings applied when it is loaded up, and we don't look at these settings during our current song to make sure that these settings are meaningful, we do run the risk of coloring our sound in a particular way, which the song may not require. And if we're using a template where actual sounds are loaded in – say, a kick drum – then this will lead extremely quickly to “cookie cutter” territory. This could be okay when working on an album, where a consistent foundation is desired, but could otherwise lead to stagnation. It will not provide experience in mixing a broad range of sounds.
Templates are useful when they contain tools and routings which we will use 95 times out of 100 anyway. Manually setting these up and loading these in those 95 times is quite likely to slow us down and kill our flow when we'd rather just be getting on with making music!
There are many effective ways to configure templates. As with music, there's no “one size fits all”. I'll go through my current template, maybe you'll pick up on some things to try.
I have a number of audio channels set up, ready for instruments which will invariably find their way into a trance song – a kick drum, a hihat, a clap/snare, a crash, a “special fx” channel, low and high basses, a couple of leads and a couple of pads. I have EQs on a number of individual channels, which are initially turned off. A couple of other processors are on some of the channels – processors which color or thicken up sounds. Once again, these start in the “off” position. It's handy to be able to simply turn them on and start tweaking if I feel a given sound will benefit from the treatment. The instrument levels are set at nominal starting positions, rather than leaving all the faders at 0 dB. The faders won't stay in these positions (I'm not a believer in “my kick drum must be at -8 dB” - it totally depends on the sound of the kick drum and the other instruments around it). The nominal level settings simply ensure that sounds come into the mix at ridiculous levels when I put them in.
I also have a number of groups set up, which the various instruments are routed into. The groups I currently use are: Kick/Bass, Synths, Drums, Loops and Pads. Some of these have highpass filters set up at nominal frequencies – a decent starting point which I can refine when the sounds come into the mix. There are various EQs and compressors which are ready to be turned on, including sidechained compressors – I have a MIDI triggered sidechain input channel set up, ready to push other sounds out of the way of the kick drum. Using groups makes it easy to make broad adjustments to levels and EQ, rather than individually adjusting a number of channels.
I have four send channels ready to go – two delays (one shorter, one longer) and two reverbs (a small room and a large hall). Send channels make it easy to add some sonic interest to sounds, and sending a few sounds through the same send channel at various points in a song gives the song greater cohesion. Sending a number of sounds into the same main reverb ensures they all sound like they're playing in the same space. I have to admit, I rarely change the delay times on my short and long delays – this is one area where my template is leading into “cookie cutter-ness”.
These groups and sends are then routed into a master channel, which contains a number of processors: a compressor with very restrained settings, an EQ (initially turned off) and finally a limiter (which starts off producing zero gain).
Well, that's my template. It allows me to get into the process of making music incredibly quickly. I'll now move onto my broad method of creating a trance song, from start to finish. The actual process will always differ in some ways (it's hard to fully predict the creative process), but this method works for me currently.
A song starts with the song – I generally have separate composition sessions where I come up with melodies, chords and basslines. Most often the sounds I use are sounds which get my creative juices flowing, and don't end up in the final version of the song. So, I have the main melodic elements ready to go when I start building up a production. From here, I'll start with either the melody, finding a sound which brings it to life, then build a suitable drum and bass foundation to go with it, or I'll start with the drum and bass foundation. Around half the time I'll build up the foundation from nothing, the other half I'll dip into my treasure trove of kick and bass foundations, usually opting for a setup I haven't used before. These kick and bass foundations usually arise out of sessions where I try to recreate other people's songs. Even when I don't get that close to the original, I'll still often end up with a great set of sounds to use in a future production.
I get the main elements in place – it is essential that the kick and bass lock together seamlessly. A strong foundation makes the rest of the mixing process much easier – it practically all falls into place, having a solid reference point to build on. If I start with the lead, then the kick is chosen while the lead is playing, then the bass (and bass pattern) is chosen while both the kick and lead are playing. After the first sound, none of the sounds are auditioned in isolation. Sounds in isolation have no meaning.
At this point, if I didn't start with the lead, I'll add it now. I'll generally have a feel for the type of sound I'm looking for, and reach into one of my synthesizers or samplers. Once I find and tweak my main sound I'll generally layer it with another sound – if my main sound is direct and centered, maybe I'll layer it with a wider, more washed out sound to lushen things up. Or I'll layer a fuller sound with a thinner sound.
Then I'll add the other drum sounds – clap/snare, hihat, crash, loops. I'll take my time going through my sample libraries. I'm looking for sounds which:
sound different to what I've used before
come through clearly when all the other sounds are playing
sound like they belong with the kick (and bass), during the intro and outro of the song.
I'll often apply a stereo enhancer to loops, and have them out wide while the main drum elements are in the centre.
After that I'll add a pad – lately I've been using just the one. Again, I'm looking for something which sits nicely in the mix, but also sounds quite nice on it's own, if it happens to be isolated at some point in the song.
I'll perform some preliminary processing to get things sounding decent – adjusting highpass filters on sounds and groups, to leave the low end open for the kick and bass. I'll often add a doubler/ unison effect to a lead sound to thicken it up.
At this point I like to get the mix sounding as “finished” as possible. I turn on the speakers and check out where the bass is sitting (I do the vast majority of my mixing in headphones). Then I export the main sequence as it currently stands, and run it through a frequency analyzer, which I then export and bring into a program I've set up where I can compare my song to a ballpark of other great sounding songs in my style. This tells me where my overall level is at, and where my frequency balance is at, relative to the ballpark. This lets me know if any levels are unbalanced, or there are resonant frequencies in some instruments. I make level adjustments and repeat the process three or four times, after which the mix usually sounds “finished”. Quite often I end the session at this point, to get some distance from the mix. The next day I'll be able to listen and evaluate it more objectively before deciding to turn the main sequence into a complete song. Some songs get shelved for a while at this point, if I don't think they're strong enough in their present form.
When I've decided to turn the main sequence into a full trance song, I get away from the computer and imagine the overall journey, the flow of the song. What will happen during the intro? When will instruments come in, how will they come in? How will the main melody be introduced? Suddenly, all at once, or gradually evolving from a pared back version? I write down how I imagine the journey to go. Then I go back to my sequencer and write in a guide track (I currently use Cubase. Not every sequencer may have this, a text track where you can specify the various sections of a song). I write in where instruments come in or drop out, where levels or filters should be automated, and so on. This is very handy for keeping the overall song in mind when working on sections.
Then I record all the instruments. I use a number of hardware synths for a lot of my sounds, so here is where I write in all the notes and MIDI automation to control the various synth parameters (resonant lowpass filters feature heavily in trance). Once I'm happy with all the notes and automation I record the incoming audio, then that track is "locked in". I proceed in this way through all the sounds I'm bringing in from external sound sources. I really enjoy committing to sounds and finishing songs. In my early years of mixing I was using only software synths and I'd leave my sound choices wide open right up until a song was finished - I'd find myself still agonizing over whether my lead sound was good enough, well after choosing the original sound. I have a better selection of sound choices these days (this is for me personally - I don't buy into "better" and "worse" sound sources per se), and I find it so much easier to commit to sounds early and get on with the process of finishing the song.
Once all the sounds have been recorded I get my drum programming in place - where they have fills or drops leading into transitions, where levels, filters or other parameters can be automated to heighten the effect of certain sections. Often I extend this to other instruments as well - maybe a few bass notes can be cut out heading into a transition as well. I don't feel I need to stick with exactly what has been recorded if a bit of creative editing at transition points will work well. I also automate send effects - for example, the feedback control on a delay, or maybe the overall volume of a reverb (rising up, or even having the reverb cut out briefly for effect). I'm aiming to get as much of the flow in place from the main sounds before adding any other effects.
Then, when I do get around to placing the special effects sounds, they're the icing on the cake rather than trying to carry a lot of the flow by themselves. I look for a good selection of uplifters, downlifters, crashes and other effects as desired. I look for a decent range, and see if I can use them two or three times - more than this and they can sound very overused (especially if they're quite distinctive), less than this and it can take away from the overall integrity of the song - it sounds like I've pulled out all my special effects and dropped them in all over the place.
Then the song is done. I'll export it as a WAV file and listen to it now and then over the next few days, making notes for adjustments. After a few days I'll try out all the adjustments I've noted and re-export the song. Then I like to let some time pass before hearing it again, to come back to it with fresh ears. Usually at this point I don't pick up anything which can be further improved (maybe after another 6 to 12 months of experience I'd be able to improve it).
I hope you've enjoyed reading about my work flow - feel free to ask me some questions!
Keep making great music!
Fabian
No comments:
Post a Comment