We explore some ways in which creative mixing can add life to a sample‑based production.
In theory, sample‑based music should be easy to mix. There are no performance or tuning issues to contend with. There’s no need to disguise poor miking technique or bad room sound, and no unwanted spill to fight. Mixing is usually both a technical discipline and a creative one; with nothing needing ‘fixing in the mix’ a sample‑based production ought to give us free rein to concentrate on the creative side.
In practice, though, productions of this sort can be very challenging. This is often down to arrangement: even the best sample libraries will never sound convincing unless the composer has a good grasp of how to write for strings, brass, bass guitar, drum kit and so on. Sometimes the struggle is to make pristine sampled instruments fit alongside lo‑fi home‑recorded vocals and guitars. And, not infrequently, producers simply choose sampled sounds that don’t fit the production.
Rich Townsend is much too good a composer and musician to fall into any of these traps. Even so, he wasn’t entirely happy with how his own mix of ‘Blue Fingers’ had turned out. When he compared his sample‑based production with modern big‑band music and jazz captured live from real ensembles, something was missing. The question was, what? And could a Mix Rescue bring it back?
Working under the name Samplehound, he aims to release no fewer than 16 albums of new music by the end of 2023. Some will be purely acoustic, while others, like ‘Blue Fingers’, will feature electronic elements. Rich’s original version of ‘Blue Fingers’ can be found on the 2021 Samplehound album Cabin Fever.
The track itself is a sophisticated piece of jazz, built mainly on sampled versions of conventional instruments such as grand piano, Rhodes piano, bass guitar, drum kit and percussion, but also incorporating some more abstract synth and sound‑design elements. A very skilfully crafted arrangement meant that just pushing up the faders delivered a pretty effective balance, and Rich’s mix sounded excellent on its own terms. However, I could understand his sense that it was a little lifeless in comparison with his references.
To try to recapture this sense of liveliness, I decided to reimagine the project as a multitrack recording of a live band. How might it sound if these were real instruments, arranged within a big live room and captured using a main stereo pair augmented by spot mics?
The first step in realising this vision was to tackle one of the main issues that can make sample‑based productions hard to mix. Sample‑based instruments invariably make full use of the stereo field, with pianos, drum kits, string sections, even humble percussion instruments all spanning the entire width of both speakers. This makes them sound impressive, but when more than a couple of instruments are combined, the listener’s perspective gets confused. In a good recording of a real ensemble, each instrument has a concrete and well‑defined position within the stereo panorama. Recreate the same piece with samples, and everything is everywhere.
Rich was obviously aware of this issue and had narrowed instruments such as the grand piano and Rhodes, but his multitrack still contained a lot of expansive elements. This was especially true of the drums and percussion, where even the kick drum track had very significant stereo content. So my first step was to make almost everything mono, and use the dual panners in Pro Tools to give each source a plausible position within the stereo field. I decided, for example, that the drum kit would occupy the middle of the field, with percussion instruments arranged either side of it. I hard‑panned the electric piano and the acoustic piano to opposite sides, and used similar brute‑force tactics to position some of the synths. The main melody instrument was an interesting patch from Spectrasonics’ Omnisphere that morphed between two very different sounds; it too occupied the entire stereo field, so I made it almost mono and kept it in the centre.
It’s hard to accept that you’re laying the groundwork for a successful mix by making every individual instrument sound smaller and less impressive!
Mono‑ising the elements of a recording in this way is a leap of faith. It’s hard to accept that you’re laying the groundwork for a successful mix by making every individual instrument sound smaller and less impressive! Sometimes the process can reveal serious problems with mono compatibility, too, in which case you’re often best off simply throwing away one side of the stereo file and using whichever one sounds better. But it’s vital to keep the bigger picture in mind, and in this case, half an hour’s fiddling with pan pots gave me a stereo panorama in which everything important now had a distinct, well‑defined position. The individual instruments sounded less grand in solo, but the stereo presentation of the mix as a whole made more sense.
However, it didn’t fully solve the problem of confused perspective. Because the samples were all bone‑dry, the mix still sounded very claustrophobic, with no front‑to‑back depth and everything jostling for attention. It also lacked any of the psychoacoustic cues that, in a ‘real’ recording, would serve to bind different instruments together and place them within an environment. When multiple instruments play in the same room, even their close mics typically pick up some spill, but more importantly, they also pick up the ambience of the room itself. How obvious that ambience is in the finished recording depends on many factors, but even when we don’t hear it as obvious reverberation, it still tells our ears that all of these different sounds are being made in the same space.
Rich’s own mix seemed to have very little reverb. This made sense when I listened to his reference track, a surprisingly dry and up‑front recording of the jazz group Yellowjackets playing with the WDR Big Band. After some experiments, though, I decided that I wasn’t going to be able to make all the sampled instruments gel without making the mix a bit more ambient.
I started out with the purist goal of using a single stereo reverb to recreate a studio live room, and feeding everything into it in proportion to its left/right position and depth within the mix. A ‘Big Drum Room’ preset from EastWest’s QL Spaces II was perfect for making the many drum and percussion tracks sound as though they belonged in the same space. However, it wasn’t so flattering on the keyboards, synths or clarinet, so my purism evaporated and I ended up creating several other reverbs.
A different studio live room impulse response seemed to work better on the piano and Rhodes, and a hall patch from Liquidsonics’ Cinematic Rooms helped the remaining instruments to gel. I consoled myself with the thought that in a real session, it wouldn’t be unusual for drums and percussion to be recorded in a live room while piano and winds were in booths! This also had the bonus that I could send the drum reverb to the same auxiliary input as the dry drum tracks, so that any bus processing would be applied to both together. When you’re trying to create the illusion that reverb belongs to a source, this can make a big difference.
My selection of ‘realistic’ reverbs certainly helped make the mix blend together, but I felt it could stand a little more excitement. I reasoned that if I were mixing a live multitrack, I wouldn’t be limited to using the ambience that had been captured on the mics: I’d probably want to add some more characterful, artificial reverb and delay as an effect.
Again, trial and error was the order of the day, and the treatments that worked to my ears included a splashy plate reverb from SoundToys’ Little Plate, which I added to many of the drum tracks, and some more experimental reverb and delay settings from Baby Audio’s Spaced Out and SoundToys’ Primal Tap on the synths and lead instruments.
As I mentioned at the start of this article, sampled instruments don’t typically require corrective processing. There’s no proximity effect or harshness that needs to be tamed with EQ, no rogue notes that jump out, and so on. However, there’s a fine line between perfection and blandness, and sometimes the pristine quality of sampled instruments just feels too sterile. The perception of front‑to‑back depth also depends on tone as well as ambience, and rolling off high and low frequencies can be an effective way of pushing something further back in the mix when you don’t want it to sound too reverberant.
There has been much talk lately about saturation as a mixing tool, and to my mind, it really comes into its own in exactly this situation: when there’s nothing wrong with the sources, exactly, but everything sounds a bit too ‘hi‑fi’ and lacks character. Consequently, although I didn’t use any conventional EQ at all at the track level, and relatively little on buses, my mix contained numerous instances of SoundToys’ Radiator, Softube’s Harmonics, Acustica’s Taupe and Overloud’s Gem Tapedesk and Dopamine. I used Softube’s Bass Amp Room to give the sampled bass guitar some grit, and Line6’s Helix Native to make the Rhodes sound as though it too was being played through an amp.
By now the mix was sounding more lively and organic, but when I played back the song all the way through, I realised that a single, static balance wasn’t serving the music throughout. Some of the sounds Rich had used, especially that morphing synth lead, were very dynamic: a balance that placed them at the right level in one section made them too loud or quiet elsewhere, and compression alone wasn’t helping all that much. To address this, I began by creating a rough map of the piece, using Pro Tools’ Marker track to delineate different sections. I could then target my use of automation to ensure that the important elements could be properly heard in each section. With some of the sounds, it wasn’t so much the level that needed controlling as the tonality, and there were several places where I automated a low‑pass filter cutoff or multiband compressor threshold to get everything to fit.
The technical side of mixing is self‑effacing. We know we’ve achieved our technical goals when listeners don’t notice them. By contrast, it’s tempting to think of the creative aspect as being an opportunity to show off, adding ‘in‑your‑face’ effects and treatments that make the listener sit up and take notice.
If there’s a lesson to be learned from this particular project, it’s that this isn’t necessarily the case. Bombarding the listener with ear candy or warping sounds beyond recognition can be fun, but the kind of creativity you really need as a mix engineer is more subtle. It’s about having a vision for the overall sound of a project and being able to realise that vision in a compelling and natural way. Inasmuch as there was imagination at work in my mix, it lay in constructing a framework that would allow Rich’s musical ideas to come across clearly and naturally.
Creativity in music belongs most of all to its creators, and the contribution of the mix engineer shouldn’t work against those of the composers, arrangers and performers. A creative mix isn’t a series of stunts designed to grab the listener’s attention, but a set of decisions that illuminate the creativity already present. Changing the decay time of a reverb or the timbre of a tambourine won’t be obvious to the casual listener in a way that an extreme vocal effect might, but they are all part of a more general creative act. The unique contribution of the mix engineer — the thing that no‑one else can do — is not to ‘enhance’ these individual elements, but to make them work together towards a common goal.
Rich Townsend: “I’ve recently upgraded the acoustic treatment in my studio. When I listen to top‑quality commercial releases, such as Yellowjackets’ latest album, I get a wonderful sense of realism, the recording space, and fluidity of the high frequencies. When I listen to my mix of ‘Blue Fingers’, I don’t get those things. It sounds a bit like a digital recording from the early ’80s: a bit congested and flat, with HF that definitely cannot be described as fluid, and without a clear sense of space.
“I really like what Sam has done for the track. It sounds more coherent and I get a better sense of real people playing together (even though it was recorded one track at a time). I thinks that’s down to the sense of space that Sam has created very successfully around the music. Everything sounds more relaxed and organic. The piano solo sounds more human. The entry of the complex synth lead sound has been made into a real feature.
“My favourite change is the way Sam has adjusted the electronic piano part and panned one of the Battery drum sounds to the right, where it gives a real sense of a stereo soundstage. It never occurred to me that the sampled kick might have a wide stereo spread — this is something I will pay attention to in the future!”
To hear Rich’s original mix and Sam's remix, download the hi-res WAV files of the original and remixed tracks in this ZIP file: