How to Make Grime/Garage Bass by artists like Flava D and Skepsis

UK Grime and Garage is emerging into mainstream dance hotspots around the world. Youtube channels such as Deeprot are gaining massive followings and spreading some of the UK's dirtiest basslines from the likes of London, Manchester, and Birmingham. This tutorial is going to show you how to generate one of the hottest sounds in the scene at present. Differing from the deep UK house "bass donk" as is popular in tracks by MNEK, and David Zowie, this grime bass is characterised by a slower attack, whiney FM wub most popular in tracks by the likes of Flava D (below), and Skepsis.

The first trick to getting any synth sound by ear is to analyse what the component sounds are. We will use Flava D's "In the Dance" as a reference.

Coming in at 0:59 in the video, we can hear the sound that we want to recreate. Assuming you are listening on good speakers, you'll hear the two distinct sounds which create it. There is a sub with a relatively short attack, giving the note a low frequency punch, then there is the mid-range whine.

We are going to make this sound in Serum but any synth capable of FM should manage fine. Serum was chosen because it's wavetables and controls are clear to see, but this sound will be achievable on freeware such as T-Force Alpha too.

The hardest part of deciding where to start is the waveform. We can hear that the sound is dark, moody and low frequency weighted. We can therefore assume that either bright waveforms such as square waves aren't being used, or they are being heavily filtered. We can also tell that FM modulation is being used, though detecting this may come with some experience with FM sounds.

Let's start with opening Serum to the initial preset. 

Right now, we aren't quite sure which waveform to use, but we have a fairly good idea that we want a deep sound, so perhaps sine, or triangle. I like to start with sines, and if need be, add more harmonics, so let's find a sine waveform, where we can add harmonics via the wave position. "Analogue_BD_Sin" is one of my favourite waveshapes as it enables various modulations of a sine wave, while still keeping it distinctly sinusoidal.

The whine sound is generated by Frequency Modulating (FM) one oscillator with another. From experience with music, we can roughly guess that there is one oscillator playing the root note, and one playing 7 semitones (fifth) above it. So let's duplicate the oscillator A waveform into oscillator B.

Shift oscillator B up 7 semitones. We don't know which oscillator is modulating the other just yet, so since there's only two options, let's try one, and if it doesn't work, we try the other.

After trying the Modulating oscillator A, with oscillator B, and vice versa, we realise that the sound isn't quite the same. It is still quite low frequency, and not as whiney as in the song. Lets try adding an octave to oscillator B so it becomes 19 semitones above oscillator A (12 semitones for the octaves, 7 for the fifth above it).

Again, having tried the FM on that, it's not quite as high pitched as we like, but by having the FM on oscillator A, we are getting closer to the sound. Let's try pitching oscillator A down two octaves, so it is three octaves, and a fifth below oscillator B.

All of a sudden, the sound begins to resemble the spread of frequencies that are present in the song's bass synth. Before we play with nailing the FM, let's get the other characteristics sorted. We have a rough tonal template, but not a dynamic template. Let's listen to the sound and see how we can change the ADSR to resemble it a bit better.

After playing about, it turns out that a low/medium attack, and low/medium release are pretty close. We can use this same envelope to modulate the FM, because it sounds like the FM is being modulated along with the volume. The sustain is good for now, but will likely need lowering later when you are polishing off.

Before we go any further though, we need to consider the sub. It is a fairly clean sine sound so let's add it in. We also know that it has a faster attack and so being the first element in the sound, we want that to be controlled by the main envelope. The oscillators will therefore be controlled by a secondary envelope, so take note of the settings and transfer them to envelope two. TOP TIP: "alt +click and drag" copies the envelope to another.

These settings approximate how the sub sounds, and we now know that the oscillators can be controlled by envelope two. To do this, we drag it onto the volume level knob on each oscillator and set the range. We will also do this to the FM envelope on oscillator 1, bringing it from zero to eleven at it's modulation peak.

The sound is quite a bit brighter than what is present in the song, but for the mid-range has the right harmonic content, to fix this, we will use a LP filter. This can be chosen from the filter menu, I only want a soft filter so I will choose MG Low 12, so the roll off isn't too sharp. This filter gets modulated by the same envelope as the FM and oscillator levels, play about with the range until you get close to the sound.

Now we are close. The trick here, is to play with the modulation ranges and levels until you get as near to the sound as possible. Once you are happy with that, you can start adding FX to give it that polish. Below are screenshots of my final patch (note Osc A and B switched positions, but everything else is similar).