Our tame mix engineer was asked to inject some dark, moody aggression into this month's track. Read on to find out how he did it...
The audio files available on this page accompany the Mix Rescue column for SOS August 2012, featuring the song 'Fine Lines' by Toronto-based band The Guest Bedroom. The filenames are hopefully fairly self-explanatory, but the descriptions below should help you understand a little more about what you're hearing.
Mix Rescue | SOS August 2012 - The Guest Bedroom by Sound On Sound
This is what the snare-drum close mic sounded like in the final mix, processed with just a 2dB low shelving cut at 420Hz from Reaper's ReaEQ and heavy compression from Antress Modern's freeware Deathcore compressor. The compressor was set to a 10:1 ratio, with fast attack and release times. Both high-pass and low-pass side-chain filters were engaged, and the Color button's HP and DS lights were both active.
Without the heavy compression from Antress Modern's freeware Deathcore plug-in, the snare close-mic doesn't provide nearly as much sustain, as you can hear if you compare this example with the SnareDeathcoreIn file.
This is my full remixed drum sound, but without the group-buss compression. In addition to the polarity inversion facilities built into the Cockos Reaper mixer, I used Reaper's built-in ReaEQ, ReaComp, and ReaXcomp processors, as well as the following third-party freeware plug-ins: Voxengo MSED, Elysia Niveau Filter, Antress Modern Deathcore, Sonimus SonEQ, and Smartelectronix Ambience. Slate Digital's Trigger was used to layer an additional sample to support the kick-drum's low end.
Here's the final drum sound from the remix, complete with group-buss compression from Universal Audio's EL7 Fatso Sr. Compare this directly with the DrumsCompOut file to highlight the balance and tonal changes this fairly heavy-handed processing incurred.
By default, both the attack and release settings for the Universal Audio EL7 Fatso Sr (which I'd inserted over the drum group) were set faster than I wanted, and this file demonstrates why I chose to lengthen both these settings: on the one hand the fast attack was robbing snap from the snare and low-end weight from the kick; while on the other the fast release was reducing the psychological impact of the compressor's gain-pumping — a slower setting gave a better illusion of your ears caving in under the pressure! The differences are quite subtle, though, so compare this file side by side with the DrumsCompIn example to hear them at their starkest.
Another of the default settings in Universal Audio's EL7 (which I'd inserted over the drum group) was the built-in side-chain high-pass filter. By default this was set to 240Hz, but this didn't deliver the degree of gain-pumping I was looking for, so I reduced it to 120Hz. Taking the cut-off frequency any lower down the spectrum had an adverse effect on the weight of the kick drum, however, as this audio example demonstrates — I've bypassed the high-pass filter altogether, and it really loses ground at the low end compared with the DrumsCompIn example file.
During the main instrumental riff sections I felt that the drum room was being ducked a bit too forcefully by the drum-group compression, so I fed a little of the room signal past that compressor, as you can hear in the DrumsCompIn audio example above. If I hadn't done this, I'd have ended up with a sound as in this audio example — notice how the cymbal sustains in particular are rather jerkily modulated by the double kick-drum hits.
Here's a basic balance of the raw DI and miked-amp bass tracks — I used a couple of decibels less of the DI. The phase-cancellation between the two signals is sucking out all the low midrange, and makse the musical line very inconsistent because some frequency components are so far out of kilter, especially the low end.
Here's the same balance of miked-amp and DI bass signals you heard in the Bass01_BasicBalance file, but now with the DI signal's polarity-inversion switch engaged. Although the tone is still quite unfocused and woolly, the low midrange is much fuller, and the musical line a lot better controlled.
By adding a delay of around half a millisecond to the DI signal, I was able to bring the bass part's midrange into much clearer focus, as you can hear if you compare this example with the Bass02_Polarity file. It's also worth listening side-by-side with the Bass01_BasicBalance example too, because the contrast demonstrates quite how much of an impact adjusting polarity and phase relationships can make to this kind of combined tone.
Although the bass sound in the Bass03_MicroDelay is nice and solid, I did feel that it needed a touch more low end within the context of this mix, so my final mixing tweak on this instrument was a touch of 100Hz low shelving boost from the freeware Sonimus SonEQ plug-in. If you compare this example with the previous three files you can hear that the EQ isn't making nearly as big a tonal difference as the polarity/phase adjustments.
This is the original lead-vocal recording for a section of this month's Mix Rescue remix, without any processing. Notice how the tone overall feels rather dull, even though a lot of the louder vowel sounds ("ah”, "eh”, and "ee” in particular) have some uncomfortably piercing frequency components.
Running the vocal through the HarBal mastering software, I used its IntuitQ harmonic balancing routine to try to tackle the strongest of the narrow frequency peaks, resulting in a frequency curve that had big dips at 900H, 4.4kHz and 11.5kHz in particular, as well as a strong compensatory peak at 6.6kHz. If you compare this audio example to Vox01_Original, you can already hear what a big difference this single step already made to the subjective smoothness of the sound — although I still felt there was more to be done on that score.
Having bounced the vocal track back out of the HarBal software (which only works off-line), I first cleared up the tone further with Reaper's ReaEQ plug-in (a high-pass filter at 250Hz and a further broad 7dB peaking cut at 420Hz) and then drove the signal hard into ToneBoosters's affordable Ferox tape-emulation plug-in — again, as a means of trying to round off some more of the troublesome narrow-band frequency spikes. Line this file up alongside the Vox02_Harbal example to hear the differences most clearly.
The final stage in my attempts to banish the vocal's sporadic harsh frequency peaks involved compressing its 1-9kHz region hard in three octave-wide frequency bands using Reaper's multi-band ReaXcomp processor. As you can hear, the results aren't exactly natural-sounding, but at least the mid-range spikes don't tear your ears off any more!
Once I'd vanquished the harshness problem to my satisfaction, the rest of the lead vocal processing for my remix was fairly straightforward: a decibel of 900Hz boost from the freeware Sonimus SonEQ; a good dose of full-band 8:1 compression from Stillwell Audio's The Rocket (operating in its oversampling Decadence mode with a 91 percent Impetus setting); a 8dB side-chain EQ boost at 7kHz to prevent The Rocket from over-emphasising vocal sibilants; and a good deal of fader automation to keep a consistent balance in the full mix.
In order to allow the lead vocal through the mix better, I applied a number of EQ cuts to backing parts to reduce their masking effects, as you can hear in this example file. The EQ comprised a 7dB third-octave cut to the drum overheads at 7.7kHz; a similarly narrow 7dB cut at 6.5kHz on the full drumkit group channel; a wider 7dB cut at 5.7kHz on the toms group channel; and another narrow 2dB cut on the organ distortion channel at 2.7kHz. Now compare this file to the VoxAntimaskingEQOut example to get a feel for how much of an effect these EQ tweaks had on the vocal balance.
For comparison with the VoxAntimaskingEQIn file — I've bypassed the four EQ cuts specifically targeted at reducing lead-vocal frequency masking.
This example showcases some of the delay effects I used to blend and widen the final mix — I've muted the drums here so that you can hear them better. The lead vocal and bass both have a short stereo pitch-shifted delay patch applied to them, and the lead vocal also has a heavily EQ'd single-tap echo from Univeral Audio's EP34 Echoplex emulation for blending purposes (as well as to add a sense of 'liveness'). Another short echo from Togu Audio Line's freeware TAG-DUB-II plug-in has been used to blend the guitar, and also works to widen the guitar image by virtue of its being panned to the opposite side of the stereo panorama.
For this example I've bypassed all the delay effects you can hear in the DelayEffectsInNoDrums file, so you can hear how much difference they make. Notice in particular how the bass, guitar, and vocal sounds all become more 'point-source' and don't gel as easily with the organ part.
Here's the original mix of The Guest Bedroom's song 'Fine Lines', as sent in by SOS reader (and the band's bass-player) James Toth.
This is my final remix of 'Fine Lines' based on the same set of multitrack files, although I did add a kick sample from Slate Digital's Trigger plug-in and a few background chords from Nine Volt Audio's Metal Guitar REX loop library. The mix was carried out in Cockos Reaper using a variety of third-party freeware and payware plug-ins.