We spend a lot of time rescuing sub‑standard recordings and arrangements in Mix Rescue, but this month it really was all about mixing and mastering.
All too frequently, Mix Rescue seems a misnomer for a series that would be more appropriately titled Arrangement SOS, Recording Rescue or Performance ER. In other words, the problems that make a track difficult to mix usually turn out to originate at the pre‑production and tracking stages, and require more than tweaking a few faders to sort out.
So it was a welcome surprise to receive the multitracks of Superfox's 'Hear It In My Head', because for once it was hard to find fault with the material, the performance or the recording. The band's catchy blend of pop, soul and funk, which slightly recalls Maroon 5, was clearly well routined, thoughtfully arranged and tightly played. And having hired a professional studio for the basic tracking, they had emerged with a sparkling set of clean WAV files, boasting good tones and excellent separation.
Despite this, the band were having trouble producing a finished product that did justice to the raw ingredients. Eager to get the track done and out to the world, they contacted Mix Rescue to see if we could help.
A quick listen to Superfox's own rough mix suggested a couple of areas for potential improvement. Perhaps most obvious was the low level of the lead vocal, which tended to disappear when the mix got busy. The drum balance also sounded a little odd, and reverb was getting a bit out of hand, so I could see why the band felt the mix as a whole lacked impact and life.
The multitracks happened to arrive just after I'd done a fresh install of Avid's Pro Tools LE 8 on my office Mac. So, armed with an M Box 2 and a pair of headphones, I thought I'd follow Paul White's example from October 2010 and have a stab at mixing the track without using any third‑party plug‑ins.
There were 31 audio tracks in total, but it quickly became apparent that not all of these would be needed. For example, the five main electric guitar parts were each represented by three tracks: an amped sound miked separately with an AKG C414 and an Audix i5, and a DI'd version. I decided to reduce each to a single part straight away, so muted the DI and 414 tracks in favour of the i5 versions.
The eight drum tracks included the expected stereo overheads and close mics on kick, snare and (gated) toms, plus three additional tracks labelled 'Hats', 'Ride' and 'Lstn mic'. This last, I assume, was tracked using the listen mic off something like an SSL mixer, and certainly had a character of its own, but I couldn't immediately see a need for the other two, so decided to leave them muted unless such a need emerged later on.
The tracks were all very cleanly recorded, and little EQ seemed necessary beyond high‑pass filters on the overheads, toms and snare, plus some mid-range boost to emphasise the attack of the toms. There was, however, no room-mic track, and no perceptible ambience on the overheads, so although it was not hard to get a basic drum balance, it all sounded rather flat and artificial, almost as though it had been built up from samples. Clearly, some reverb was in order, and I ended up using two separate instances of the AIR Reverb plug‑in. One was set up as a global effect, generating only early reflections, and I sent a little of the overheads to this to liven them up a bit. The other used a version of the factory 'Drum Room' preset, modified to make it nice and bright. I added sends to this from the snare and from both tom mics, and routed its output to the same auxiliary channel to which the drum tracks themselves were all bussed.
Harry Soar's drumming was very controlled, so compression wasn't required as a corrective measure. However, it's rare to encounter a commercial mix in this genre that doesn't feature compression used for effect, and some quick experiments with the stock Digirack Compressor/Limiter III suggested that a healthy amount of squashing on the drum bus, with a fast attack and release, added excitement and urgency. The snare part in the verses involved a lot of quiet stickwork punctuated by louder hits, and I added a Bomb Factory BF1176 plug‑in to its close‑miked track in order to balance the levels of the two. I also used the Expander/Gate plug‑in to tame the ring of the toms, and to gate the reverb feed so that only the louder snare hits generated reverb. Finally, with lots of guitar tracks to accommodate, I decided to pull in the width of the overheads to make room in the stereo field.
As well as the drum kit tracks, there were two mono percussion tracks: a shakey egg, which was present virtually throughout the song, and a tambourine, which helped to lift the choruses. I hard panned these left and right respectively, and then left them well alone for the time being.
So to the vocals. Listening to Eddie Glossop's lead vocal in solo revealed a rare question mark over the recording itself, in the shape of a noticeable harshness on louder sections. Attempts to create a typical modern rock vocal sound by using masses of EQ and compression simply made it sound scratchy and nasty, so I reined things back to a few dB of gain reduction and a small boost in the 3kHz area, and resorted to extensive level automation to help the vocal sit at the right position in the mix.
Again mindful of the guitar‑fest to come, I decided against using reverb on the lead vocal, and instead created two global delay effects — one a short, bright slap‑back without much feedback, the other a much longer and warmer panning echo with several repeats. I then used automation to balance the two sends from the vocal track, leaning mainly on the slap delay in the verses and bringing up the longer delay when things needed to be a bit more epic.
I also remembered the old trick of thickening a vocal track by sending it to a stereo pitch‑shifter, and shifting the left and right signals up and down a few cents respectively. However, the suite of plug‑ins bundled with Pro Tools LE doesn't include a real‑time pitch‑shifter, so as a substitute, I tried creating two mono aux tracks, panning them hard left and right, and using two instances of the (mono only) AIR Frequency Shifter plug‑in to detune them. The resulting effect wasn't the same at all, but in moderation, I quite liked it, and used it on several of the other sources besides.
There was a stereo track of backing vocals, too, which required minimal treatment beyond some reverb and the usual level automation.
I didn't anticipate too much trouble from the rhythm guitars, which were all tightly played and well recorded, and so it proved. Most of them appeared only during the chorus, and after a bit of mucking about with faders, they sounded fine to me, apart from one electric part that benefited from a cut at 1.7kHz to stop it fighting with the vocals and snare. Like the drums, the guitars were recorded very dry, so I set up another global reverb, and also used small amounts of the slap delay and early reflections patches I'd already employed elsewhere. I hard panned all the electric guitars left or right, and that sounded respectable enough, so I didn't attempt anything more subtle.
There were also several lead parts, plus a stereo acoustic guitar, which opened the track and reappeared during the choruses. The acoustic sounded great just as it was, but needed narrowing to make sense within the context of the mix. In the verses it was joined by two lead parts, which again I panned hard left and right and anointed with slap delay and early reflections. One of these was a busy, clean funk part; a little compression with a shortish (4ms) attack time helped the transients pop out, while I used level automation to ensure that the details could be heard and to duck some of the 'chuck‑a‑chuck' damped sections between phrases. The other was an almost keyboard‑esque melodic part, which, to my ears, cried out for some sort of 'ear candy' treatment to make it more engaging. I found this in the AIR Vintage Filter plug‑in, using its envelope follower to set up something a little like an auto‑wah in the verses. I bypassed this in the chorus, there being plenty else going on there to engage the listener's attention.
Finally, there was a proper overdriven lead guitar, which executed a nice solo after the middle eight and some melodic twiddles over the final chorus. Guitar solos are often an opportunity to wheel out some fun effects, and this was no exception. I panned the source track hard right, then sent it to a mono aux panned hard left; this had a short delay set 100 percent wet, taking advantage of the Haas effect to create a wider sound. Both of these tracks then fed a stereo aux channel adorned with an AIR Dynamic Delay plug‑in, set up to create a long ducking delay, and a flanger to add some movement. Plenty of level automation was also required to get the guitar solo to sit nicely with the vocals and the rest of the mix.
The last major instrument to worry about was Liam Wragg's bass guitar. As is often the case, this took some time to get right, even though it was well played and well recorded. There was quite a contrast between the bass part Liam played in the verse, which was punchy and staccato, and his more legato chorus bass line, and any mix treatment that worked well for one didn't seem to work for the other. In the end, I created two different aux channels with different effects on them. One used the AIR Vintage Filter to add low‑mid thickness and substance, while the other employed the AIR Fuzz Wah plug‑in to add grit and bite. I then used automation to balance these and the lightly compressed dry signal to taste: the filtered version helped add substance in the verses, while the distorted version helped the bass cut through the wall of guitars in the choruses.
With the basic mix done, I began to feel that there was one area where the arrangement could use a little help. After the middle eight and guitar solo, there was a drop, where the band returned to the sparse feel of the opening verse before kicking back into the final chorus. It seemed to me that the effectiveness of this drop could be enhanced by a little bit of creative editing and processing. I cut out and muted the relevant sections of the shakey egg and funk guitar tracks, and used another instance of the AIR Vintage Filter plug‑in to screw down the drum bus, the aim being to achieve maximum impact when the big drum fill kicked back in at the end of this section. To emphasise the spacey quality of the drop, I also duplicated the lead vocal line 'To know that we'll always be friends', moved it to a new track, shifted it a bar later and applied the AIR Lo‑Fi plug‑in to create a ghostly one‑off delay effect. An hour or so of fiddling with plug‑in automation later, and I felt I had something that really added to the track.
The final stage was to apply some sort of master processing to bring things up to a respectable level, and hopefully to add some beef. Up to this point, I had felt more liberated than limited by the lack of third‑party plug‑ins on my computer. The AIR effects sounded good, and the Pro Tools EQ did its job well. However, while the Compressor/Limiter III plug‑in is easy to set up and gave clean, transparent results on individual channels, it can't compete with some of the more luxurious third‑party mix compressors on the market, and the only dedicated mastering plug‑in bundled with Pro Tools LE is the ancient Maxim limiter. I used these to bounce out a rough mix to play to the band, but decided that final mastering would have to wait until I had access to something a bit better.
Happily, the band liked the rough mix, and had only a few changes to suggest. They had the excellent idea of adding a tambourine sample to reinforce some of the snare hits in the verses, which worked well; they also spotted that the last note of my ghostly vocal delay in the drop clashed with the underlying harmony, so I used the off‑line AudioSuite pitch‑shifter to push it down a semitone.
Until this point, I'd only heard the mix on headphones, so I took a raw 24‑bit master home with me, partly so I could check it on speakers and partly to take advantage of some proper mastering tools. I'm not usually a fan of multi‑band compression, but in this case, Waves C4 did a good job of thickening things up without affecting the essential liveliness of the sound. I also applied a touch of the same company's S1 stereo width enhancer, and then ran the mix through the tape emulation section of Magix's AM‑Track, which is quite a good way of making a mix louder without too many adverse side‑effects. For the final limiting process, I turned to the new Slate FGX plug‑in, which is gobsmackingly effective at making things loud while retaining the integrity of transients and other dynamic elements.
In the end, then, this month's project turned out to be less about rescuing and more about just mixing. As such, I think it underlines how crucial it is to get things right in the first place. When a band who can play use a good studio to record quality material, as was the case here, it becomes so much easier to steer the mix back onto the right path, even without the aid of fancy third‑party plug‑ins. Mixing a well‑recorded track is primarily a matter of balancing levels, with processing and effects used to enhance good source material rather than disguise its faults. If only it was always like this!
Superfox are a four‑piece pop‑rock band from Sheffield, who have made great strides in the short time they've been together. The full line‑up is: Eddie Glossop, lead vocals and guitar; James Glossop, lead guitar and vocals; Harry Soar, drums, and Liam Wragg, bass guitar. A powerful live act, they've played shows alongside Bon Jovi, Little Man Tate, Wishbone Ash and Saul Williams, and their material has been widely aired on local radio.
Superfox's Harry Soar comments: "When we tried to mix our track, we couldn't find a way to get the balanced, powerful sound that we as a band love. Radio‑friendly bands seem to find bags of depth and character in their mixes. Yet we couldn't find such depth in our track and we couldn't put our finger on what it lacked.
"Then we heard the new mix and we loved it immediately! It managed to balance all the many different sounds in the mix without losing sight of the song. The extra echoed vocal lines and reverb effects made the sound much more rounded and polished, and the snare effect in the middle eight was a great touch. These little things helped to create a mix that has more "flavour”. The direction Sam has taken was spot on, and exactly what we were hearing in our heads (no pun intended...) when we thought of the track. We can't wait to get people listening to it, so it'll be going up on Spotify, iTunes and our MySpace page very soon!
"Thanks so much to Sam at Sound On Sound for taking the time to work some magic on our track. You're a legend!”
We've placed some 'before and after' audio files on the SOS web site so you can hear for yourself the changes that Sam made to Superfox's track. Go to /sos/dec10/articles/mixrescueaudio.htm to hear them.