Thoughtfully prepared DAW templates can leave you with more time for the creative and musical side of mixing.
There are many possible approaches to communicating with a client, and what works will vary, according to your personality and theirs, your relationship with them and how you like to work. But I always like to ask at least a few questions about what they’re hoping for from the mix, and where there are any previous mixes, to find out just what it is they don’t like about them. As well as getting us both on the same page, what I’m really trying to do is hand over a bit of ownership of the mix before I even start; I find that most people seem to appreciate that approach.
A rough ‘demo’ mix, if there is one, is nearly always a good thing to have, and a few realistic reference tracks can also be useful. In the case of this month’s mix of ‘Can You See Me’ by Dominic Kershaw (who performs under the name Trick Bird), after our initial discussions, I felt that what I really needed to do was, to put it crudely, deliver a bigger, fuller, more exciting version of what Dominic had already done.
As I’ll explain, I was able to achieve this fairly swiftly, due to some simple tools and techniques I’ve cultivated specifically to get a mix up and running quickly. As I’ll explain, this approach enabled me to change certain elements quickly, towards the end of the process, and to perform some mix revisions without too much unpicking of what had gone before.
The multitrack files had been nicely prepared and everything was mix-ready. I’d established that Dominic’s main musical background was as a guitarist, and that this was one of his early attempts at singing the lead vocal parts. The effects had been printed on most of the guitar tracks, and it seemed obvious that a great deal of time and attention had been devoted to getting these to sound how he wanted. I was fine with this as they sounded decent and, in any case, it seemed to me, following our initial correspondence, that I should probably not look to mess too much with the guitar sound. There were a few different vocal tracks: a combination of the track labelling and listening to the rough mix suggested that three different parts made up the lead vocal sound; and two groups of stacked backing vocals provided two opposing refrains throughout the song. The programmed drums, comprising only five tracks, were refreshingly simple, and these were joined by a bass guitar and a selection of pad-style keyboard parts that supported the vocal melodies.
I’ve been hesitant about using DAW templates in the past — it always felt a bit like mixing by numbers, and I could see it leading me to choose the easiest option, rather than the best — but, having spent some productive time recently figuring out how a template could help me, I’ve changed my mind. Why? Quite simply, my template helps me reach the important stages of a mix much more quickly. It’s no problem at all to me that I’m repeatedly leaning on the same tools to achieve certain sounds and results — it’s a very ‘analogue’ way of thinking, and I get to know those tools inside out, but I’m still free to use every advantage of digital in-the-box mixing. My current mix template isn’t as sophisticated as some I’ve seen, but it works for me...
I tend to use Pro Tools’ ‘import session data’ facility to bring in a set of typical group busses, with a selection of plug-ins that I find often work for certain instruments. I’ll also import a number of send-effect options like delay and reverbs, a master-bus chain and my current way of setting up the routing going into the master bus, as I’ll explain later. The master-bus selection is important, as it immediately sets a tonal character for the whole track, and that influences all the mix decisions that follow. Mixer Andrew Scheps, who’s famously now working entirely in-the-box, recently described his master-bus chain as the equivalent of his analogue console sound, and that’s a good way to think of it. It’s quite common for me to have a number of processing tools set up, including tape emulations and saturation plug-ins, as well as the more conventional mix-bus compressor and EQ.
Once I’d imported everything I wanted, and had performed some further mix prep (things like colour-coding and stacking the tracks in my preferred order), I set about establishing a rough static balance. Having positioned Dominic’s mix in a convenient spot in my arrange window I used that as a reference, and quite quickly, over two or three passes of the song, arrived at what I felt was a reasonable balance.
Even at this stage, my stereo-bus processing had already begun to make the project feel a bit more ‘mixed’, and having got it to this point I could decide on what my initial priorities should be. I decided to start with the guitars, as it felt like I needed to make a few decisions on these before progressing things.
There were two main rhythm guitar parts: an acoustic guitar that had been recorded with a mic, and a distorted electric guitar. These were supplied completely dry. They were accompanied by assorted melodic guitar parts, though, on which the effects had been printed. Dominic had done a really nice job of these atmospheric parts, which tastefully supported the vocals in all the right places. So my job was just to find the right level and play with a few panning options. The drums sounded very mono, but I don’t have a problem with that in general — if everything sounds wide, then nothing does! But it did mean that I would want to look to the two rhythm guitar parts to give me a strong sense of left and right when the song really kicked in.
The acoustic guitar recording was largely fine, though I could hear just a touch too much room tone and the overall sound was a little bright for my taste. It doesn’t always work, but I often find that a simple bit of notch filtering can clear up unwanted room tone, and on this occasion I found that sweeping a very narrow cut-filter allowed me to identify where the problem was (550Hz) and a narrow dip at that frequency cleaned things up nicely. I then used a simple tilt EQ to lose some of the brightness, along with some high-pass filtering. The acoustic guitar, panned around 75 percent to the left, was now sitting nicely.
With the guitars roughly laid out, I returned to my natural habitat of fiddling with drum sounds. I needed to decide if I wanted to steer the programmed drum parts towards a more believably ‘real’ drum sound, or to embrace a more programmed aesthetic. Dominic had chosen good samples, but I felt the sounds had far too much sustain for this song; while they sounded good in isolation they weren’t adding any real ‘punch’ to proceedings. The kick drum also lacked a little low-frequency power and, given the nature of the bass guitar part, this meant the whole track felt a bit ‘light’.
I therefore spent a while playing with a transient shaper plug-in. By winding back the sustain, I was able to start steering things in the right direction. Still, I was fairly sure I’d need to augment things with a sample or two of my own. For the kick, I opted for a more organic, dry-sounding drum, combined with a sample from an old LinnDrum drum machine. With around a 50:50 blend of these with the original kick sound, things seemed much fuller, but I still needed to apply a generous boost at around 60Hz to get a satisfying bottom end. I now had what I considered to be a good-sounding kick which, crucially, retained enough of the character of the sound Dominic had chosen. It was a similar story with the snare sample and, with some very careful blending or reverbs, I was able to get the cymbals sitting nicely in the track.
The most significant aspect of the drum processing in this mix was my decision to get heavy-handed with some saturation tools. I normally have a bit (and sometimes a lot) of this kind of processing going on with drums, and typically have the whole kit running through either a tape emulation plug-in or something like Sound Toys’ ‘Decapitator’. Softube do a great little free saturation plug-in that reminds me of the sound of an old 1073 when it’s driven into distortion, and I used an instance of this on the kick and snare drums and really drove them hard; I loved how this blurred the old and new elements of the programmed drums into a more cohesive sound, and it really reminded me of some of the influences that popped into my head when I first heard the track (the Flaming Lips being the main one), so I went with it.
The bass guitar was one of my favourite things about the track, so I wanted to make sure it felt like a key element in my mix. I spent a while listening to the bass and drums in isolation, and was a little concerned that it didn’t seem to be quite pumping in unison with the kick drum. It didn’t feel like I could do a great deal about this, and it hadn’t seemed like a big issue when the guitars were playing. I did, however, use the old side-chaining trick of putting a compressor on the bass guitar that was keyed by the kick drum. After carefully playing with the attack and release controls, the bass guitar ducked by 1-2 dB on every kick hit. This golden-oldie trick can have a nice gluing effect for the drums and bass (if used in the right situation — I’d only set this up if I felt it were needed).
The bass part itself was great. I’d been supplied with three tracks: a straight DI, a miked Ampeg amp and some bass reverb, but I couldn’t discern a huge amount of difference between the two main options. In that situation, I find it’s a good idea to pick only one and run with it, so I chose to fashion the sound I was looking for around the DI. I’m often amazed at how much I use the ageing SansAmp plug-in that comes free with Pro Tools. Very often it does a really nice job of making Bass DI tracks feel more alive, and it did precisely that here. But my job didn’t end there...
A large chunk of the bass part was played quite high up the neck and, while this was effective melodically, the bottom end fell away a bit too much at these points. A generous helping of EQ boost at 50Hz sounded great when the lower notes were played, but there was nothing down there to boost on certain sections! Thankfully, there are a few tools that can help give the impression of lower frequencies that aren’t naturally there — essentially, they work by generating harmonics related to lower-frequency fundamentals, without actually generating the fundamental itself. I played with a few such options before settling on Waves’ RBass to add a little extra fullness. You do have to be very careful with such tools, especially if your monitoring system is at all suspect, and I made a mental note to pay close attention to this when I did my final checks.
The final touches to the bass were to boost some of the mid-range around 1.5kHz, which helped it cut through, and to introduce a little spring reverb as a nod to the direction Dominic had already established. I’m really liking a touch of spring reverb on bass guitars at the moment. Used sparingly, I find it adds a little space and vibe but without clouding things up.
Vocals are important in any track, but I felt an extra responsibility with this song, given that Dominic was not experienced as a lead vocalist: I always admire people who have the courage to step up to fill this crucial role, and I wanted to do what I could to convince Dominic that his vocals belonged in the track — I figured that was half the battle!
As I said earlier, the lead vocal comprised three main tracks, and it wasn’t obvious quite how these should be balanced. Having Dominic’s mix as a reference was therefore crucial, and I spent quite a while balancing and automating the three parts so that the result adhered quite firmly to his interpretation. That said, I really liked one of his falsetto-style parts — more, perhaps, than he did — so I accepted the risks associated with making this a bit more prominent. Using a tilt EQ, I really pushed the air on this vocal and, with a bit of reverb and delay, it almost began to meld into the keyboards and guitars, which I really liked.
It took a fair amount of level-automation work for me to get the vocals sitting where I wanted them, even after they’d already been compressed quite generously. In fact, tweaking the vocal automation was an almost continuous process during the last half of the mix, but sometimes that’s what it takes — it’s hugely important.
Effects-wise, apart from a sprinkling of short and medium reverb on the vocals, I found a ‘shimmer’ delay in Soundtoys’ Crystallizer plug-in, which added a really nice effect to the vocal without it becoming too wet or ‘washed out’. The lowest of the three lead vocal parts I left almost bone dry, though — this gave the overall vocal sound a greater sense of depth and solidity.
The mix was now roughly in shape, and I was able to begin using some of the routing features of my new mix template. I’ve been trying to settle on a way of being able to add very small amounts of level automation to a mix as a whole as I find it can really add some much-needed dynamics to static-sounding mixes.
Often, when you use mix-bus compression, if you automate subgroups like the drums to come up in level during a section of a song, that level change is fought by the mix-bus compression. What I now do instead, is route all the outputs from everything in the mix to a series of different stereo Aux channels (busses), before they reach the master fader. This allows me to split my mix-bus processing over different stages. The first bus, which every track in the mix gets routed to, has no processing on it at all. It is used purely so that I can adjust the level going into the first plug-in on the master-bus chain. This is handy, as it’s of paramount importance to keep an eye/ear on how hard all these plug-ins are being driven; it really changes how they react and sound. This bus then goes onto a stereo aux containing the bulk of the processing like tape emulations, EQ and mix compression. It’s the output of this bus to which I apply any master level automation. Finally, this signal feeds into the last master channel, which typically only contains a limiter, for adding a touch of level to pre-mastered mixes.
As well as giving me more control over the processing applied to the mix as whole, because I’m doing the automation post any mix-bus compression, the changes in level are not ‘absorbed’ by the mix-bus compressor. True, they will, of course, head off into a limiter if I’m using that, but it’s still happening after the bulk of the dynamic processing. Often, a 0.5dB boost when a song hits a chorus is all that’s needed to give things more impact. For the mix of ‘Can You See Me’ I mainly used this type of automation to refine how much impact the track had when the drums and bass kicked in at the beginning, and I found myself adjusting the amount used several times before we were done.
As well as giving more thought to how I set up a mix to run efficiently and smoothly from the outset, I’ve developed a new routine of checking a mix for real-world listening issues. I have an old Panasonic portable hi-fi that I use as a type of ‘grot box’, and I find that if I listen to this reasonably loudly it can highlight problems that small speakers may have with upper bass frequencies. I’ve kept a much closer eye on this area, as I noticed recently that the generous 180-220 Hz boost that I often find myself wanting to add to a snare drum can make the speakers on a laptop produce an unpleasant ‘thunk’ sound.
Also, although my mixing room sounds pretty good, there’s a minor issue around 100Hz, where some room modes create a null near the listening position. I work around this for the most part, but for peace of mind I use Sonarworks’ headphone-correction software to give me a neutral headphone option to check a mix on. I also make a point of giving most mixes a good blast on the car stereo, both with the engine on and off. And then on my hi-fi at home...
You can check on more systems, but it’s really a question of paying your mixes due diligence without going over the top. Between a decent mixing space, and this handful of options, I find I’ve done enough.
I got a little caught up trying to get the level of the bass guitar right, and after performing the listening checks that I’ve described above, I gave this a 2dB boost, which is quite a reasonable nudge, before sending my first draft to Dominic. This sounded too loud when auditioned against the drums alone, but the overall mix felt like it needed it.
After that, I was able to wrap the mix up fairly quickly with Dominic — after a couple of sets of revisions, he was really happy with what I’d done. On the first version I sent, he wanted fewer effects on the vocals during the intro and tweaks to the levels of some parts here and there. Because of the routing I’ve set up in my mix template, it was very quick and easy to make such changes without throwing the whole mix out of balance. For instance, I have a stereo aux channel that all of my effects sends pass through on their way to the main output, so it was a breeze to open up the mix and then automate the level of the effects down for the intro section. After a little conferring, we also decided to make the lower vocal part a bit more prominent in the final version, which seemed to work well.
I've focused a lot this month on how I’ve optimised my mixing workflow, and with good reason. While you might reach different conclusions about what’s useful, and might have different options available in your own particular DAW software, I’d really encourage you to invest time in thinking about this yourself.
For me, getting the template right helps to speed up various points of the mixing process, but the benefit isn’t simply that I save time. One of the key advantages for me is that by being more time-efficient in the early stages I can get to the heart of the mix as quickly as possible, while my ideas are still relatively fresh and intact — it’s hard to assess what a song needs after listening to it for eight hours straight! With this track, I was able to get things sounding well on their way after only one or two hours — and that meant I had plenty of mental energy left to enjoy thinking about the more creative and musical aspects of the production.
Communication with your client at the beginning and end of a project are arguably just as important. Getting a feel for how much information, such as musical influences, you need from them before getting started in order to deliver something they’ll like, and the way you react to any revision requests or negative feedback they send your way are worth consciously thinking about. Getting it right is easier said than done, but being friendly and accommodating, yet also confident, can go an awful long way to making a project run smoothly — and to ensuring the whole process remains fun for all concerned!
Trick Bird is the name given to Cambridge (UK) based musician Dominic Kershaw. Despite being a proficient home recordist, Dominic was having trouble getting his mixes where he wanted them, and he explained in his initial email to me that he was left a little underwhelmed by the results when he’d enlisted someone else to mix his first batch of songs.
Dominic has been writing and recording for a few years in various bands and currently also writes, performs and records with the band Motor Tapes. His well-received first solo EP was played on BBC Radio 6 Music, and this month’s track ‘Can You See Me’ is from his second EP, due for release soon.
You can find a couple of audio examples in the sidebar of this article, or below, to give you a feel for the track Neil is describing in these pages.
Download | 95 MB