Joker Audio Song

[Ray Kowalewski]

Giventhe title of this column, it's little surprise that we end up dealing with a lot of productions that have made various fundamental recording, arrangement, and/or mixing misjudgements. As a result, I regularly find myself making broad-brush changes to all of these during the remix process, as well as resorting to some fairly drastic remedial measures on occasion.

Not so with this month's project, though, because the MP3 that SOS reader Alex Hunt sent in had already cleared the most important hurdles. The writing and arrangement were strong; the band had performed with conviction; the recordings had been approached sensibly; the timing and tuning were solid; and the mix was doing a very creditable job of supporting and enhancing the end result. Despite all these good things, however, the mix sonics didn't quite hold their own when lined up against relevant commercial releases from bands like Interpol and The Vaccines, and Alex had struggled to achieve meaty guitar sounds without losing the presence of the vocals or the power of the drums. So he asked me to cast an eye over his mix settings to see if I could help him cover the final furlong.

Clearly, it made no sense for me to start work afresh from the raw multitracks, given that so many aspects of the mix were on target, so instead I loaded Alex's Steinberg Cubase project onto my own computer system so that I could start work from where he'd left off. Opening the file, there was ample evidence of a commendably disciplined mixing approach. All tracks were succinctly named, colour-coded, and collected into sensible folder and audio-group heirarchies, while the structure of the timeline was neatly indicated in Cubase's dedicated Marker track. Many of the recordings had been carefully edited for arrangement, comping, and correction purposes, as well as being multed across different tracks where appropriate, to allow easy section-to-section settings changes. At the bottom of the track list, a few commercial releases had already been imported and assigned to an alternate output bus for referencing purposes. In short, enough for me to be able to hit the ground running!

Overall, then, there was nothing I could reasonably find fault with at first glance, so I began to dissect things channel by channel. I did this by first muting everything, and then reintroducing tracks one at a time, so that I could more easily isolate small factors that might be hindering progress.

Alex had also taken a sensible approach to EQ as far as small-studio, in-the-box mixing is concerned, conserving processing resources by using Cubase's built-in EQ for everyday clutter-removal and frequency-balancing purposes, and then breaking out higher quality, DSP-hungry plug-ins (such as Universal Audio's UAD2 Pultec and Neve analogue models) only where more extensive subjective tonal sculpting was required.

I brought the drums back in first, comparing each individual channel to the full drum mix by toggling individual Solo and Listen buttons. The first thing that drew my attention was the overheads bus, to which Alex had applied Empirical Labs' Fatso compressor. While this is well-regarded as a drum compressor, what it was doing here was pushing down the snare-drum peaks coming through these mics, so that the snare sound in the mix was heavily reliant on the inevitably unnatural-sounding close mic. Ditching the compressor and rebalancing the cymbals back to roughly their former balance rendered the snare slightly fuller and more believable, as well as gluing the close mic better into the full kit.

Some of the difficulty in achieving vocal presence during the choruses seemed, to me, to stem from a harshness around 4kHz in the drum overheads, so I dipped out 5dB with a medium-narrow (Q=1.4) peaking filter, just during those sections, to combat this. The outside kick mic was also contributing a bit too much thrashy HF cymbal spill, so a few decibels of that bit the dust too. In addition, though, Alex's wide low-mid peaking cut on the overhead mics at around 140Hz was removing a bit too much warmth from the kit's timbre, while insufficiently tackling both lower-end flabbiness and a woolliness to the snare in particular. Moving the existing high-pass filter up from 100Hz to 170Hz and then narrowing the peaking cut to centre it on 270Hz seemed to improve matters on both counts, and enabled me to mix the overheads a little higher in the balance, binding the whole drum kit together more effectively.

Although the two vintage reverberators that these UAD plug-ins were modelled on are rightly revered as studio classics, in this mix they had been applied to the drums for the wrong reasons, such that their over-warm, unnatural-sounding sustain characteristics were cluttering the low midrange unhelpfully.A couple of warm-sounding send effects felt as if they were surplus to requirements: a short patch from Universal Audio's EMT 140 Plate on the snare, and a rather incongruously long, rumbling reverb from their EMT 250 on the toms. Neither of these made much audible impact on the full mix, in terms of blending or sustaining the close-mic signals, and neither was realistic-sounding enough to give any lifelike illusion of a surrounding acoustic, so removing the sends completely felt like no great tragedy, and clarified the low-mid range to a small but perceptible degree. In addition, Alex had applied a thick-toned hall patch (from Cubase's Reverence convolution processor) to the snare alone, and although I liked the way this filled out the body of this sound, the reverb tail was rolling around between hits unnecessarily, so I gated it to tighten things up again.

In place of these excisions, I added a brighter, 'stone room'-style reverb patch with a high-pass filtered instance of Cubase's own Reverence convolution processor, as this did a much better job of bonding the different signals into a single, coherent kit sound, as well as adding a nice hint of space. However, listening in mono revealed that the Middle component of the stereo effect return was contributing a rather woolly tone to the drums, so I shelved 4dB out of it using the M/S mode of DDMF's great little LP10 equaliser plug-in. Finally, I applied some automation to the effect-return channel to support the song's long-term dynamics, opening up the drum room more for the pre-chorus and chorus sections.

Having muted Alex's two EMT reverb emulations, I replaced them with a brighter and more natural-sounding drum-room ambience from Cubase's built-in Reverence convolution plug-in, following it up with some MS equalisation from DDMF's LP10 to aid with mono compatibility. The level of this reverb was then automated to support the ebb and flow of the musical structure.

My only significant remaining concerns with the drums related to their transients. In the first instance, the snare's 'bite' didn't quite seem to balance that of the kick drum, where I felt both were equally important to the groove. Fortunately, the compression settings on the snare group channel (both over-snare and under-snare mics had been recorded) showed that Alex had used quite short 1ms attack times, so it was easy to reinstate a bit of leading edge by just inching those up to 6ms.

Even after this, the overall drum sound still felt a touch spindly, so I decided to try some analogue-style 'warming' as an additional step, in this case using Tone Booster's great TB_Ferox tape emulation. I ended up driving this quite hard, and although that did change the sound quite a bit, what I was hearing seemed to suit the band's overall vibe, as well as rounding the spikes and thickening the snare's sustain in a nice way.

Simply setting too fast a compressor attack time on the snare channel robbed the drum of too much of its 'bite' in the original mix, so this was increased for the remix, allowing more of the drum's leading edge through.

The bass guitar was already pretty adeptly mixed. Separate mic and DI tracks had been mixed to a common buss, and then treated with EQ, compression (both full-band and multi-band) and parallel distortion processes, to create a sound that really worked in context. That said, I did make a couple of small EQ adjustments. Firstly, because the low-mid range of the drum parts had now been tightened up, I figured that the bass could afford to make more use of this region, so I eased up a decibel or so on the buss channel's broad 214Hz peaking-filter cut. My second change was a 1-3kHz peaking boost from Blue Cat's freeware Triple EQ plug-in. Although I added 10dB, the audible effect of this was nowhere near as dramatic as you might expect from that figure, partly because the information in that spectral region was so low-level to start with, and partly because I inserted the EQ before the channel dynamics, which inevitably fought against the tonal change to some extent.

The various guitar parts were also very respectable, on the whole, and had been adapted to their mix context very well, by virtue of careful multing, balancing and group processing. However, there was a sense that the chorus sections were getting rather crowded by the available parts (hence the heavy masking of the drum and vocal sounds), despite the missing weight and expansiveness that Alex and the band had been hoping for.

My tactic was simply to turn up the guitars a few decibels and see where that left me. Immediately, I found myself reaching for the EQ to tame the 4kHz region of the main double-tracked rhythm part with a 4dB cut, as well as removing a 1kHz boost Alex had applied to the chiming higher harmony part and limiting that track to round off some over-zealous picking transients. The additional tremolo part added for the second chorus also felt over-bright to me, so I reversed its 2dB 5kHz EQ boost into a 5dB cut, while the low-mid range seemed over-egged in the light of the other guitar and bass parts too, so I raised its high-pass filter by an octave or so, to 500Hz, to stave off woolliness. Again, those EQ alterations might seem quite assertive, but in reality their effect on the mix was quite subtle because of the subsidiary nature of the part.

3a8082e126