One of our engineers visits reader Alex Joyce’s studio to help him bring the best out of his dubstep track.
Because I run a record label I receive a lot of artist demos every day. These vary enormously in quality, spanning everything from the totally unusable rubbish that somebody vomited out after playing with cracked software for a couple of weeks, to well-crafted tracks by established artists who are looking to reach out to new audiences. Most, though, including this month’s Mix Rescue track, fall somewhere between these extremes.
Alex Joyce is a music production student at Bath Spa University and is making his own dubstep tunes. Listening to his tracks, it was clear that Alex had already reached a fairly competent level of music production and I liked what he was doing musically too. I felt that he’d benefit from a little engineering advice, though. Alex doesn’t live far from me, so I offered to visit him in his home studio and see how I could help him apply a bit more polish to his mixes.
One of the most important aspects of production — quite apart from the quality of the gear you use and your skills and judgement — is familiarity with your production setup. Alex’s studio had acoustics and monitors that were alien to me. Also, due to the size of the available space, Alex has his monitors wedged into the corners of his room. This obviously isn’t ideal as it can cause the sound to become muddy and boomy, which can cause huge problems accurately gauging mix issues. As there wasn’t much choice in the matter (this wasn’t a Studio SOS visit, after all!), I took along some headphones that would provide a reliable reference. I don’t usually like to rely on headphones for mixdown work — I much prefer using my speakers in my own studio — but I’m very familiar with my Sennheiser HD25s, as I use them almost every weekend when DJ’ing, and as a makeshift mobile monitoring setup. They’re decent, if not ideal, and sometimes familiarity outweighs technical performance, particularly when you’re working with programmed sources, in which there are often fewer nasty surprises than in stuff recorded with mics. (Consider that Will from bass-music act KOAN Sound uses £15 Logitech gaming headphones to make a large amount of his music, and you’ll understand what I mean!).
My first port of call was to look over the DAW project itself. Alex had been using Logic Express. I’m not hugely familiar with Logic, so I decided that it would be best to guide Alex from an engineering perspective, and leave him firmly in control of the mouse. There were around 100 tracks in total, but while this may sound a lot, it’s common in this kind of music, due largely to the amount of sample layering used in the sound design. It’s also common to duplicate elements or entire channels so they can be processed differently in the various sections of a track, to keep the energy levels up or the mix balanced and consistent no matter how busy the mix becomes. Alex hadn’t labelled all of the channels, which would have made life rather easier, but as I’m quite accustomed to looking at audio waveforms on a sequencer it was fairly obvious what most elements were likely to be — and the rest was nothing a bit of soloing couldn’t reveal quickly.
The most obvious aspect of the track that needed work was the drums. They felt quite ‘pinched’, with very short, thin-sounding transients and an overall lack of low-frequency (LF) weight. They also sounded a bit lost in the busier parts of the intro, and particularly on the drop. (The ‘drop’, to those of you less familiar with dance music, is the part at which a dance track builds up from either an intro or a breakdown into its most high-energy section. As a lot of dance music does not follow a verse/chorus structure, the drop could be viewed in arrangement terms as the equivalent of a chorus.)
One of the key mistakes I often see less experienced electronic-music producers make is in their initial selection of samples. While, as in this case, it’s always possible to re-engineer or enhance a track’s drum samples, it’s generally much quicker and more effective to start with a sound that’s as close to the desired final product as possible. I suppose this falls somewhere in that grey area between engineering and changing the musical direction, and in this case I trod as carefully as possible, only adjusting the more functional elements like the ‘weight’ of the drums: I didn’t want or have the time to get into rethinking any synth sounds or melodies. Alex and I scanned through a couple of sample libraries, including my Sound Of Dodge & Fuski collection by Prime Loops (I know the content of that one very well!), as well as Vengeance Essential Dubstep Volume 1. We chose a much heavier-sounding kick, as well as a punchier snare sample. It’s really important to choose a kick and snare sound that complement each other in terms of balance, which is something I always judge by feel. Generally speaking, you want them to be of similar perceived volume (judge by listening; don’t try to read the peak or RMS values!). More traditional forms of 4x4 dance music, such as house and techno, often use a constant kick drum, with a clap layered on top of every other beat — and you could consider the heavily LF-weighted snares that are popular these days as a sonic substitute for the weight of the lost kick drum.
As the kick sample I chose was already nicely processed, it didn’t require a lot of work. I just added an EQ boost around 100Hz for a little extra weight, a cut around 300hz to make space for other elements in the track, and another boost around 5kHz, just to add some of the presence and ‘bite’ this track required. The snare required a bit more work, as we found that we needed to layer two sounds together to get the desired result. First, we picked a punchy snare with a nice weighted ‘slap’, and boosted the frequencies around 200Hz to increase its weight. We then found a 909-style clap, added a reverb to it as an insert (not only would nothing else be sharing this reverb, but given the number of tracks in the project it made sense to keep all the processing on the one track — and the insert approach can make things easier later on, as I’ll explain), removed any excess LF content below around 500Hz with an EQ in order to avoid it clashing or phase-cancelling with the main snare sound, and layered the two samples so they’d play at the same time.
Another issue was that Alex had layered about five channels of high-frequency (HF) percussion and hi-hats together to create his cymbal groove. On their own, the parts worked well enough together musically, but sonically it resulted in quite a muddled and harsh tone, and the sounds felt a bit lost during the busier parts of the track, where the dominant frequencies were clashing with those in the raspy bass noises. I slightly simplified the pattern, my focus being on keeping the hi-hat sounds that had more of a rounded, full sound to them — while removing those that only consisted of extreme HF content — and increased their volume in the mix to better balance them with the kick and snare layers.
With the drum sounds suitably overhauled, I wanted to draw attention to all of the musical layers, but without sacrificing the space in the mix that allowed the kick and snare to carry the track. As the core rhythm was an evenly alternating kick and snare in pretty much a standard 4x4 pattern, I decided to use Xfer Records’ LFO Tool plug-in. This works very much like the side-chaining of elements from a trigger track (à la Eric Prydz’s ‘Call On Me’), but it gives you incredible control over envelopes because it’s based on a highly customisable tempo-sync’ed LFO, rather than relying on an external audio source. A trick I use heavily on many channels in most of my tracks is to use this to create a very quick level-duck around every kick and snare hit, so that the initial punch of the kick/snare transient cuts through, but the ducked channels come back to their intended volume rapidly. This creates space for the drums, but avoids the ducking sounding too obvious. I ended up using LFO Tool on most of the music elements in the intro, as well as on the raspy bass noises on the drop. This allowed me to make a lot of the channels louder, which made the mix sound a lot tighter and fuller — and all without sacrificing any clarity or punch from those carefully chosen drum sounds. LFO Tool is best suited to tracks like this one, which are based around some kind of 4x4 pattern. You can draw in a more complex broken-beat pattern if you wish, but I’ve found this to be a bit more fiddly and less precise. This is probably why most drum & bass producers I know tend to stick to the more traditional side-chain trigger method, in which the kicks and/or snares, or a kick drum on a silent track (ie. pre-fader send, with the fader fully down, or no output channel selected) act as a cue for a compressor to duck the volume temporarily of the desired channel(s).
While I’m on the subject of side-chain ducking, note that you don’t necessarily need to duck the entire signal; multi-band compressors with side-chain input, such as FabFilter’s Pro-MB, Waves C6 and Vengeance’s Multi-band Compressor, and dynamic EQs with side-chain input, such as Brainworx’s bx_dynEQ V2, allow you to duck only certain frequencies. A common application is to duck just the LF content of a track, thus making space for a kick drum without changing the fundamental character of the sound beyond what’s necessary.
I made a lot of use of a freeware plug-in (again from Xfer Records) called OTT, which is an emulation of the OTT preset from Ableton Live’s excellent multi-band dynamics processor. It’s a great tool for any non-Live users and has become a bit of a stalwart of the bass music scene since its popularisation by artists such as Skrillex, and is great for sonically thickening elements in a track without needing to layer additional sounds to fill up the frequency spectrum. It’s particularly good on parts such as plucked sounds with reverbs, or complex twisting bass noises, as it really brings out the grit and more hidden aspects of the sound for a very ‘in your face’ sound. I used it on most of the musical elements to Alex’s track, as well as the vocal sample in the intro to try and make them sound ‘bigger’ and more balanced in the mix. There are, of course, many other alternatives to OTT in terms of multi-band compression, but there’s something about the character of this plug-in that gives you a very immediate bite, which is nice when you don’t want anything too subtle. And, of course, it’s free!
Some of the sustained musical elements were being drowned out a bit, due to frequency clashes with the other elements competing for the same space, and this is an issue which can’t be fixed using the duck-around-transient technique described above. There was a particular issue with the interaction between the vocal and the synth lead line in the intro, and to combat this I both boosted the vocal at around 2kHz, for added presence, and ducked the same frequencies on the lead line, to allow the vocal to breathe a bit easier. To be honest, this wasn’t ideal: if I’d been building this track from scratch, I’d probably not have chosen to have a lead and a vocal playing together, or I would at least have chosen a vocal pattern or synth sound that didn’t overlap quite so much in terms of articulation, timing or frequency. But I didn’t want to enter too far into the territory of re-writing/arranging Alex’s tune, and there wasn’t much else I could do. Importantly, it did improve things.
When it came to the bass line on the drop, I was again keen that I didn’t become too involved with the core sound design of the musical elements, so I focused on adding a bit of ‘beef’ to the existing sound. I grouped all the raspy bass elements together and then used OTT on the group, although with the ‘depth’ control set a bit lower than on the other elements. It’s very much a trial-and-error approach with this plug-in, as I find you get wildly varying results depending on the source material it’s deployed on, so make sure you use your ears. If you use the output volume control to balance the level with the original input then you can easily bypass the plug-in to judge what effects it’s actually having on the tone. As well as using this on drums, I’d say I use this plug-in on around 70 percent of synth elements in my tracks now, as it gives such a great warm, analogue-like richness to the processed sound and it’s really easy to tweak.
Alex had already placed a good limiter (Sonnox Limiter) on the master output. I often see this practice criticised, but you have to remember that this sort of music is often aimed first and foremost at DJs and, rightly or wrongly, aggressive limiting has become the norm now for most kinds of dance music. Still, there’s a knack to getting the best result and one trick I often use — and did on this track — is to insert two limiters in series. I’ve found that this gives you a louder end result without as much audible squashing (your track will really suffer if the limiting does too much damage to the all-important punchy transient of your drum sounds). I also know some people who make their kick and snare drums much louder than the rest of the track, so that when they squash the track with a limiter at the end it gives the whole track a ‘pumping’ dynamic, creating the impression of enhanced energy. One of the great, liberating things about electronic music production is that you’re not bound by convention as to what each instrument is supposed to sound like. I’m a great believer in experimentation, and I’d encourage everyone to just have a play, rather than obsessing over technical theories. The sonic quality of the end result is the only thing that matters.
I wanted to leave Alex with some useful suggestions for his next productions. I explained that, had I been producing this track, I’d probably have tried adding a tambourine layer with a hit at every sixteenth note. I usually pick two samples that complement each other, to avoid too much of a robotic character, and use LFO Tool with a slower ramping volume increase than I described earlier, so that the pattern adds a sense of movement.
Multi-band harmonic enhancers are also great on all forms of dance music. It’s not exactly cheap, but iZotope’s Ozone has a useful four-band exciter, which I find can really help to fill out a track and give it huge amounts of added depth and power — when not over-used! (Of course, many mastering companies now use software such as Ozone, and it’s likely that any record label that signed these kind of tracks would have this sort of treatment taken care of professionally).
I often find that people delay committing to decisions or taking risks, which can inhibit the development of a track, and a brilliant creative method is to bounce all of your bass line parts (including the sub) into a single audio file (obviously you can keep the source parts too in case you need to go back to them). This gives you a lot of freedom to chop, reverse and repeat parts of it in ways that you wouldn’t have thought of — or would not have been possible — when working with layered MIDI instruments. A lot of happy accidents tend to happen when working in this way, and it often helps you to switch the groove up a bit as the track progresses. I find it much easier to make quick alterations to audio than having to faff around with copying and pasting all the different MIDI sections of the various channels before bouncing to audio, as with this kind of music a bass line is often made up of five or more different synths running together to create the groove.
Towards the end of the mix, I tend to bounce more and more sections into audio like this. This is partly down to CPU hit as my mix increases in complexity, but it really does tidy everything up, which makes working easier (remember how many tracks were used on this project?). Anything that has reverb on it I’ll also often bounce into audio, as this allows you to very quickly cut tails short where you want — this is the other reason for which you could use reverbs as an insert effect, rather than the more traditional send effect. Simply being able to cut slices out of audio when they’re hitting at the same point as a key rhythmic element like a kick drum helps to bolster the effect of plug-ins such as LFO Tool in terms of making the mix as clear and punchy as possible, while allowing for maximum volume levels. Overall though, just experiment and have fun!
Alex Joyce: “I was very pleased with the progress Rob helped me make on my mix. We analysed the importance of sample choice in the context of mixing and applied new samples to great effect. We studied the balance of high and low end and quickly agreed that many elements were clashing in the high end. Rob helped me address this by suggesting different samples that operated more within the mid-range. Rob also introduced me to some very practical and useful plug-ins.”
You can find before and after versions of this month’s track, as well as the snare samples discussed in this article on the SOS web site.