You are here

Sound Design In The Mix

Neue York are a three-piece band from Hertfordshire in the UK. Their debut EP NYXX is due to be released on 1st September 2020 on Radical Eye Records.Neue York are a three-piece band from Hertfordshire in the UK. Their debut EP NYXX is due to be released on 1st September 2020 on Radical Eye Records.

Mixing can be a creative as well as a technical role, and the right production effects can transform a track.

Technology and budgets mean that roles within the music industry have slowly blurred over the years. I'm very used to acting as both the engineer and mixer, but I also often slip in and out of a producer-like role that I refer to as 'creative engineering'. In this role, I might offer advice on a melody or use my experience to help develop a section or sound, but without crossing the line into actually 'producing' the track.

I'll try to define my role with the client up front, and when I'm paid to 'mix' a track, the job is often just what you'd expect it to be. But sometimes my ears still yearn for a little something extra that will carry the production to its full potential. In such cases, my first instinct is to reach for effects that can enhance what's already there, but if that doesn't cut the mustard I'll probably try adding in some new elements. I'm not talking about overdubbing a face-melting guitar solo or splicing a whole new section into a song. Rather, I'm talking about 'ear candy': subtle, sometimes momentary textures, layers and sound-design elements that will enhance the existing material. The sort of sounds that the listener might not always notice immediately, but which leave the song sounding that bit 'flatter' when they're removed.

Laying Foundations

To illustrate what I mean, I'm going to use examples from my mix of 'Silence' by three-piece indie/art-pop band Neue York. This song involves a four-minute atmospheric build to a crescendo, which is followed by a long, slow fade-out. Layers of acoustic drums, electric guitars, a brass ensemble and a few VST synths are topped with a dual male/female vocal. I'll describe how I used these techniques to add interest to the vocals and drums, before discussing the new elements that I added to this production and why I felt they were needed.

Valhalla's Vintage Verb is great for long reverbs that can sound really present without muddying the mix, though sculpting the reverb with EQ can help it fit in even better.Valhalla's Vintage Verb is great for long reverbs that can sound really present without muddying the mix, though sculpting the reverb with EQ can help it fit in even better.

There are two vocals in this song, and it was immediately obvious that a long room reverb would enhance their atmospheric quality. To that end, I brought up Valhalla's Vintage Verb on an aux track. A long (85ms) pre-delay helped keep this reverb out of the way of the dry vocal sound, and I gave it a very long (6.82s) tail. With a long reverb like this, you have to make sure that you keep the vocal sound, and the mix more generally, clutter-free. So I set Vintage Verb's low- and high-cut filters (to 460 and 5410 Hz) and used Pro Tools' stock EQ III plug‑in for more detailed sculpting of the lows and to cut a little harshness. Compressing the reverb a little with Pro Tools' BF-76 plug‑in gave the effect even more sustain.

Rather than stopping there, however, I was keen to see if I could make the reverb even lusher and smoother, to the point where it would become a feature of the track in its own right, without obscuring the vocals or rendering the lyrics intelligible. To achieve this, I inserted another compressor after the reverb, this time setting Waves' C1 plug‑in to respond to an external side-chain signal sent from my vocal master bus. With a high-ish ratio (about 8:1), this ducked the reverb sound by around 4dB while the dry vocals were present, and the roughly 400ms release meant the reverb swelled up smoothly in between phrases.

Finally, a gentle widening of the reverb track, courtesy of Waves' S1 stereo imager, created a little extra space in the centre of the mix for the dry vocals. Subtlety is the key to using stereo wideners effectively: they can sound seductive while you're setting them up, but pushing them too hard can create a very unnatural/phasey listening experience, especially on headphones. It's also well worth checking that any widened elements remain present enough when you're listening in mono; your mono and stereo mixes won't ever sound identical, but even with a slightly different, drier, balance the mono mix still has to deliver the goods.

Drying The Drums

Drums are also fertile ground for the use of 'character' treatments. Almost any weird or wonderful effect can be put to good use on drum tracks, but there is a lot that can be done simply by balancing the raw elements of a drum multitrack in interesting ways. In this case, I decided that an almost unnaturally dry, up-front drum sound would be an effective counterpoint to the big vocal verb.

UAD 1176 Rev A and SPL Transient Designer plug‑ins were used to 'suck' ambience from the drum room tracks for a claustrophobic, dry sound.UAD 1176 Rev A and SPL Transient Designer plug‑ins were used to 'suck' ambience from the drum room tracks for a claustrophobic, dry sound.

I was provided with a pair of stereo drum room tracks which sounded great in isolation, but in their untreated state, they didn't help achieve this goal. Rather than ditch them altogether, I EQ'd them to create a more claustrophobic and somewhat unnatural sound, removing some upper mids around 3kHz and rolling off the high end from around 4kHz, with a gentle, 6dB/octave low-pass filter. I then set UA's characterful UAD 1176 Rev A ('blue stripe') to 'stun' (the famous 'all buttons in' mode, with the fastest attack and release) so as to add some grit and to level off the peaks. Finally, I turned to SPL's Transient Designer plug‑in, setting the sustain to -24dB to really suck out the ambience.

All this processing combined to inject a nice crunchiness into the drum sound and enhance its width without the need for any additional room reverb. The contrast between dry drums and wet vocals worked well as a starting point, but I wanted the mix to evolve in sympathy with the track. I set up an instance of AudioEase's Altiverb convolution reverb on an aux track using the Studio Davout C preset, sent pre-fader signal to it from various elements of my drum mix, and automated the aux track's level so that it gradually increased over the length of the track. This took the drums from being super-dry at the start into an exaggerated big reverb at the crescendo.

Gradually automating the volume of this long reverb's aux track helped the drum sound evolve throughout the song, up to the crescendo.Gradually automating the volume of this long reverb's aux track helped the drum sound evolve throughout the song, up to the crescendo.

Ultimately, I want my clients to listen to the mix and get super excited.

Yearning For More

The bass on 'Silence' was provided by a Hammond organ, which I had doubled with a Roland SH‑101 soft synth emulation in order to even out the low end. To treat this duplicate track, I created two aux returns. One hosted another Altiverb instance (the Chapel Studio IR preset) and one Soundtoys MicroShift. I rolled the top and bottom off both effects, and this all gave the new bass sound a sense of space and width, without muddying up the low end or taking up space in the high frequencies.

Reverb coupled with stereo modulation added depth and width to the SH‑101 bass part.Reverb coupled with stereo modulation added depth and width to the SH‑101 bass part.

The SH-101 emulation was used to augment a Hammond bass part that suffered from an inconsistent low end.The SH-101 emulation was used to augment a Hammond bass part that suffered from an inconsistent low end.

The SH‑101 was providing the foundation of the bottom end, which left me free to try to exploit the potential of the Hammond as a source of 'ear candy'. After some experimenting, the resulting signal chain began with 6dB/octave high- and low-pass filters to isolate the midrange. Next, one band of Ohmforce's Ohmicide multiband distortion (I can't express just how much fun this plug‑in is!) created a really gnarly distortion. This sounded cool but rather static, so I deployed a mono-to-stereo instance of Soundtoys' PanMan, sync'ed to the track at a super-fast 1/64th random rhythm, making the part leap like a lunatic around the stereo field. Now we were talking! Finally, I rolled off some high frequencies generated by the distortion, to tuck the part back in the track. This created a cool effect that followed the bass line, and could be automated to swell and create extra depth and excitement where desired.

You'll notice that this, like almost all of the character effects I added, was heavily band-limited. You don't want your 'ear candy' fighting with the more fundamental elements of the mix; and doing this also allows you to add a lot more of it!

A new (emulated) Juno 106 part was automated to create an evolving stereo pad sound, which enhanced the synths already in the production.A new (emulated) Juno 106 part was automated to create an evolving stereo pad sound, which enhanced the synths already in the production.

The track already featured a cool, warpy pad sound, which I really wanted to swell. Reverb and delay alone didn't take it far enough, so I decided to add another part in. I doubled the original part using a Roland Juno 106 emulation, and then automated some controls (using my keyboard's assignable pots to 'perform' these moves, rather than programming them with a mouse), and ran a few passes of the track whilst adjusting the frequency of the VCO section and the level of the triangle wave in the DCO.

Rather than discard the original Hammond bass track, it was used to create an interesting glitchy, distorted stereo effect, courtesy of filtering, distortion and an autopanner.Rather than discard the original Hammond bass track, it was used to create an interesting glitchy, distorted stereo effect, courtesy of filtering, distortion and an autopanner.

Wanting yet more lushness from this cool, evolving sound, I added the wonderful Goldplate, a dynamic plate reverb emulation Kush Audio developed with ReLab, and stumbled upon a cool trick: playing with Goldplate's pre-delay setting during playback warps the reverb tail, in the same way that a hardware Electro-Harmonix Memory Man or Roland Space Echo warbles when you play with the delay time. In this case, I automated the pre-delay so that it slowly swept between 15 and 28 ms, imparting a lovely, slightly wonky sound to the reverb. I then automated the volume of this synth track, so it swelled up to create the washes of sound alongside the original synth pad.

Automating the level and pre-delay time of a Kush Goldplate reverb served to enhance the Juno pad sound and create a swelling effect.Automating the level and pre-delay time of a Kush Goldplate reverb served to enhance the Juno pad sound and create a swelling effect.

Piano Man

At this point I was generally happy with the mix, which now evolved slowly and nicely, and had real depth courtesy of all the reverbs and swell effects. But I still felt that when the drums kicked in (at 1m 12s) something was lacking in the upper register. I liked the sparse drums, sub-bass and vocal, but felt that the bass notes needed somehow to be accented on the downbeat. Rather than change the bass sound itself, which was working, it needed something extra, something bright, with the sort of rich overtones that you get from sustained low piano notes played hard...

Low sustained piano notes were used to help accent the downbeats of the bass melody and add upper frequency content.Low sustained piano notes were used to help accent the downbeats of the bass melody and add upper frequency content.

That thought was a 'lightbulb' moment. I quickly rigged an AKG C414 B‑ULS mic in front of our studio's upright piano and recorded a simple, single-note phrase that followed the bass. Some gentle EQ got rid of the super-lows (below 64Hz) and brightened it a bit, and a touch of saturation and compression from Waves Renaissance Axx gave this sound the right amount of aggression, pulling it a little more up-front. Finally, some long, bright reverb from Altiverb's AMS RMX16 preset (5 Hall B3, decay 7.0) really helped to add more depth to the piano notes.

An AMS RMX 16 emulation helped to further sustain the added piano notes, giving them depth and width.An AMS RMX 16 emulation helped to further sustain the added piano notes, giving them depth and width.

Signing Off

Ultimately, I want my clients to listen to the mix and get super excited. So I'm always chasing those spine-tingling sounds and moments, and I find that adding the sort of subtle elements and 'ear candy' I've described here can help me deliver them. There's a definite line I don't want to cross, though: if I can add an element that takes the production up a notch without the client really noticing, everyone's a winner, but unless they've specifically asked me to do so, I never want them to feel like I'm 'playing' on their record. Happily, in this case, the client loved the additions. So much so, in fact, that they wanted me to push them up in the mix. Naturally, I was happy to oblige!

Audio Examples

On the SOS website, https://sosm.ag/ear-candy-1020 you can find a few SoundCloud audio clips from this mix in lo-res MP3 format, demonstrating how Matty's mix treatments and new parts lifted the production. Alternatively, download the ZIP file of hi-res WAVs and load them into your own DAW.

Package icon sounddesigninthemix1020.zip