Two recent hits have featured the same Etta James sample (taken from her rendition of 'Something's Got A Hold On Me'), but I'd rather pay money for this track than for Avicii's 'Levels' any day. It's not that Flo Rida uses the sample any more imaginatively — neither of them are particularly inventive there — it's the main groove of 'Good Feeling' that I think kicks Avicii's arse, even once you take into account Flo Rida's higher subjective loudness level and slightly faster tempo. Tastes differ, naturally, so line up 3:34-3:49 of 'Good Feeling' with 2:40-2:55 of 'Levels' and decide for yourself. If you happen to reach the same conclusion, the obvious question is: 'What's Flo Rida's magic ingredient?'
To my ears, it's partly the nature of the kick and snare sounds themselves. The layered clap/snare backbeat of 'Good Feeling' is certainly much denser than the rather spindly sprinkled clap of 'Levels', but the kick also seems significantly tighter. Sacrificing some pure LF level and sustain in return for a deeper arrangement 'pocket' between hits makes the groove seem lighter on its toes to me, leaving the more powerful and sustained low end of 'Levels' sounding rather sluggish and flat-footed by comparison.
The way the backing track responds to the main beat only compounds this. With 'Levels', the main tool used to give a sense of momentum between beats (besides the synth fall-off into beat seven of every two-bar section) is kick-triggered gain pumping, but it appears to be acting too quickly to provide a particularly good sense of 'suction' towards each new beat. Flo Rida's gain-pumping, by contrast, seems much better adapted to the tempo of the track, such that the synths rush up more audibly towards the off-beat.
The pumping feels a little deeper, too, which might partly account for my impression that both kick and snare come across more clearly (deeper gain-reduction on kick hits means that it suffers less from frequency masking). However, I suspect that two other factors are also responsible for this subjective effect. Firstly, the main synth riff has an eighth-note gap programmed into it on every snare hit, unlike that of 'Levels' (or indeed its own beat-less drop sections at 2:28 and 3:19, which seem somehow to throw the tighter main groove into sharper relief). Secondly, there are some definite reverse-envelope synth layers in there which build towards, and then cut dead on, each snare hit.
None of these differences might amount to much on their own, but their combined effect on the urgency and excitement of the groove is pretty compelling — for me, at least. And the good news for anyone chasing Flo Rida's sound is that the techniques I've mentioned above can easily be applied in even the simplest DAW system if you keep your wits about you.
Before we leave this production, though, let me just flag up one other little point of interest during Flo Rida's 'ooh look, there's a bandwagon!' half-time break at 2:49-3:04. (To be fair, I'd say this particular instance knocks spots off similar dubstep stunts I've heard recently on Alexandra Burke's 'Elephant' and Madonna's 'Give Me All Your Luvin'.) The upper-register arpeggiations here feature high-speed square-wave auto-panning as a stereo widening effect, which is something you don't hear that often. The advantage of this method of widening is that it doesn't result in tonal changes when summed to mono — most other mono-to-stereo expansion tricks either produce a chorus-like effect or suffer from comb-filtering when heard from a single speaker. The down side of the auto-panning, though, is that relocating the synth to the stereo extremes makes it recede by around 3dB in the mono balance relative to the stereo, atlhough that's not a huge deal in this particular context, as it's not a major hook. Mike Senior
I often talk about the need to mix (and indeed master) with your target listeners in mind, because different playback systems and listening situations affect what people are likely to hear and appreciate in the music. In practice, though, most mix engineers understandably hedge their bets somewhat, in order to cater for as broad a range of listeners outside their core market as possible — as a chart-dance producer might do, for example, by layering some mid-range harmonics over a club-tastic sub-bass for the benefit of the small-speaker listeners driving his iTunes sales. But what makes this Kelly Clarkson mix such an instructive listen is that it boldly follows a much more no-compromise line.
As I see it, this is a mix that's aimed squarely at the lowest common denominator as listening systems go: factory-floor boom-boxes, elderly AM car radios, office laptops, bar-room tellies, and ear-buds of every flavour. From a commercial perspective, the melodies and lyrics are primarily where the money's at, so what matters most under such adverse listening conditions is that lead vocals and important hooks sound clearer and louder than any of the competition. However, it's also beneficial if the bare bones of the groove and musical harmony also reach the listener intact, because these can both powerfully enhance the 'stickiness' of the hooks.
The mix's overall frequency fingerprint certainly fits the bill in this respect, by hitting the 1kHz region unusually hard. This spectral bracket is about the safest port in a storm, because it's lightweight enough to be transmitted over even the smallest mobile phone speaker, but nonetheless pretty robust when it comes to conquering 80mph road noise, passing through curtains, or echoing down long corridors. That 'Stronger' pushes the mid-range at the expense of the sub-100Hz zone should also come as no surprise, as it's the low end that usually presents the primary limitation when mastering for maximum subjective loudness: this master's metering 5dB peak-to-average, which is pretty unforgiving. Clawing back some illusion of fullness and power is then achieved by hyping the bass instruments in the less headroom-hungry area around 120Hz.
The stereo picture is pretty ruthless, too, effectively a kind of 'wide mono' that relies on its multi-layered mid-range instruments and a few effects returns to fill in the stereo image. As a result, the crucial relationship between vocals, bass, kick and snare remains rock solid in mono, but still with just enough harmonic information left in there to support the lead melody. All you really lose, in fact, is textural thickness, but this isn't a great loss, given that it can compromise intelligibility in real-world mono playback scenarios such as shops, stations, and cafes. There's nothing crucial panned heavily to one side either, so faulty stereo systems and shared ear-buds won't throw any spanners in the works.
Of course, the way the individual instruments are treated is also a vital component. Predictably, the vocal is compressed and automated to kingdom come, its spectral top octave turned up to 11 to shove Kelly's rock-diva nuances right up your nose. (The de-esser on that session is probably still in therapy.) Note, however, that the vocal also claims first dibs on the mix's 200-400Hz frequencies, always in short supply in pop mixes because of their high-risk status — too much energy there and you'll jeopardise the overall tonal clarity that's practically a stylistic prerequisite. The kick has enough high end to cut through on small speakers, while the snare has sufficient sustain to hold its ground despite the transient-flattening collision with the digital end-stops at mastering. The bass has to trust its luck with the 100-200Hz region, primarily, so it's fortunate that only the smallest systems fail to make a passable stab at reproducing those frequencies in practice, especially when that bit of the spectrum has been deliberately enhanced.
As far as general mix balancing is concerned, long-term dynamics are clearly playing second fiddle to audibility concerns here. So on the one hand you can hear the faders and/or buss compression working overtime to maximise every detail of the vocal transmission, as well as riding up backing parts to bridge the vocal gaps (check out the one at 0:24, for instance), while on the other you end up with choruses that struggle to imply any significant step up in energy, despite exceptional sleight of hand on the part of the mix engineer.
The outcome of all this is a mix that really pops out from the crowd in its intended environments. If you want an idea of how well it does this, try an experiment: mix the file with 0dBFS pink noise and then bandpass the result at 1kHz. It's amazing how much of the information survives such a punishing combination of frequency masking, distortion and restricted bandwidth — especially when you compare the havoc that similar abuse wreaks on a lot of less specialised chart mixes. But when you push a mix to such extremes, something's got to give: on studio monitors, 'Stronger' sounds like it's suffering from light sinusitis, and everyone seems squished up against a sheet of glass — even once you've turned the volume down enough to render the saturated high frequencies listenable. Mike Senior
After reading about Paul Butler's production style back in SOS March 2012, I eagerly ordered Michael Kiwanuka's album to hear the results. The eponymous single is as good a place as any to get acquainted with this wonderful-sounding record, and it's also full of production gems. I was particularly intrigued to hear the string sound Butler achieved by layering up a lone violinist (Andy Parkin), which is a dodge that SOS readers regularly ask about. My view is that the care he took in using two different violins, three contrasting mics and a selection of miking positions to differentiate all the overdubbed layers definitely paid off, because the final timbre and blend feel not only in keeping with the production as a whole, but also rather lovely in their own right. (For more tips on creating layering strings like this, check out the big feature we did about budget-conscious string-recording in SOS March 2008: /sos/mar08/articles/stringsinahurry.htm.) The way that background hiss enhances the production is also very effective, to my ears, and I'm convinced that a cleaner recording would actually have made Kiwanuka's voice (and the whole mix) feel a lot less airy than it actually does.
However, the aspect of this track that's grown on me most with repeated listens is the guitar part. Perhaps it's because I've endured far too many uninspired 'sing and strum' demos and open-mic nights from budding singer-songwriters, but it feels like a breath of fresh air to have Kiwanuka singing over a less threadbare figuration, especially one as endearingly childlike as this. But there's more to it than innocence, because there's a nice sprinkling of semitone dissonance in the chord at 0:08, for instance, and an unusual wide-spaced open fifths texture at 0:15 and 0:50. At 0:29, he introduces some little neighbour notes in the pattern's high voice, following that up with an even nicer middle-voice oscillation at 0:37. The break in the rhythmic pattern at 0:45 then serves to draw attention to the lyric "close my eyes, won't look behind”. These features may seem almost laughably basic on the face of it, but the measure of their success for me is that when Kiwanuka comes to rest for a couple of bars on the same triad the song opened with (at 1:02), it manages to feel like a real contrast. Mike Senior
Check the SOS Forum for more comments on the tracks discussed in The Mix Review this month: /mixreview