In a wallet‑friendly departure this month, we rescue a reader's track with nothing more than careful use of Logic Pro's built‑in plug‑ins...
Kate Ockenden works in the music performance and technology departments at Kidderminster College and has a Logic Pro‑based recording system at home for producing her own songs. She wrote and composed the jazzy, funky composition, 'Get It If You Want It', which features her singing both lead and backing vocals, as well as playing and programming all the instrument parts. For the drum sounds, she'd used a kit from East West's Colossus and a loop from Logic's library, and for the strings East West Quantum Leap Symphonic Orchestra Gold Edition. Most of the other parts had been done either in Colossus or with Logic's own bundled instruments. The vocals were recorded using a Rode NT2000 mic and a TL Audio Ivory Series 5060 channel strip.
Although Kate was happy with the arrangement, which I felt worked well for the style of song, she wasn't happy with the overall sound — but wasn't quite sure why. She was also conscious that her own voice tends to get quite edgy when she sings loudly in her upper register, and although she'd tried to tame it using Logic's de‑esser plug‑in, this wasn't really doing the job.
No doubt these problems were due in part to the monitoring limitations in Kate's studio, so I invited her to bring her mix over to my studio, where we could listen to it over my Event Opal monitors in a room that I knew well. Any MIDI tracks were converted to audio before she brought the project to me, in case I didn't have all the same plug‑ins, and the audio files were bounced — though it sounded as though some processing had already been applied to some of the tracks.
I was very tempted to throw all my lovely UAD plug‑ins and Lexicon reverbs at the job, but as Kate was keen to learn anything new that she could about mixing, I decided that I'd stick to using only Logic's built‑in plug‑ins: that way, she'd be able to take the mix home and dissect it to see at her lesiure what I'd done to it. I know from experimenting with some of my own third‑party plug‑ins that the final mix could have sounded a little bit smoother, and perhaps a bit punchier than the one we ended up with — but even restricting myself to Logic's plug‑ins, I managed to make a positive difference to the tonal balance and overall focus of the mix that Kate really liked. After all, not every one can afford to get all the plug‑ins in the world, and the key to making good recordings and mixes is to make the best of the tools that are available to you.
On playing the mix over my monitors, I easily identified the vocal brightness that had worried Kate, even with no EQ or compression on the track. I was also aware that the low end of the mix was slightly muddy and congested, with some boxy overtones. This had the effect of putting the whole mix out of focus and making it sound less polished than it should. Some of the parts had been cut up and moved to improve the timing, so before getting into the mix, I did the usual housekeeping thing of trimming out silences at the start of tracks that might contain small noise and also adding short fades to the starts and ends of some edited sections to avoid pops and plops.
As it was the low end of the mix that worried me the most, I started out by isolating the bass guitar to see why it wasn't sitting correctly in the track, and found that Kate had fallen into the trap of trying to make a bass instrument sound more prominent by EQ'ing in some bass boost — when what was really needed was more mid‑range detail to give the sound some character and definition. I ended up treating the bass using Logic's compressor in its Platinum mode, with a 2:1 ratio, and with the threshold set to reduce the loudest peaks by only 5dB or so. I suspect that the original bass sound came from samples anyway, so it was already reasonably even in level. A 10ms attack and 60ms release time worked well enough, keeping the sound nice and percussive. My EQ setting ended up with a 24dB/octave roll‑off below 65Hz and a generous hump at 365Hz to bring out the mid‑range punch, and to add some definition when heard over smaller speakers. With the main issues of the bass sorted, I could get on with the rest of the mix.
One of the more useful aspects of the standard Logic channel EQ is its built‑in spectrum analyser, though this needs to be set to its highest resolution mode to show you anything useful. (If you're not using Logic, there are plenty of similar analysers available that can do a similar job.) I often use it on bass instruments and drums to see what's going on below 40Hz or so — as that's largely inaudible, and is thus wasted energy — but I also use it on other instruments to see what's happening below the normal range of the instrument. You can't always hear these low frequencies in a mix but they can combine to pour a layer of mud over the bottom octave or two of your overall sound.
My strategy for this particular song was fairly straightforward, as Kate had done 90 percent of the work, in choosing appropriate sounds and writing the arrangement. I'd sort out the bass guitar sound, tweak the drums where necessary, and then 'bracket' any other sounds that were spilling into too wide a section of the audio spectrum, using 18 or 24dB/octave low-cut filters, combined with some HF cut where necessary.
As we've pointed out before, there's only a limited amount you can do by working on a sound in isolation, because what matters is how it works in the context of the overall mix. For example, you could shave the low end off an acoustic guitar and string pad to the point where they sound painfully thin when solo'd, but they can still work fine in the context of the track. As a very general rule, the bass end of a track should be adequately catered for by the bass instrument, in this case the bass guitar, and the kick drum — and nothing else should be allowed to stray into their territories without very good reason.
I also wanted to find a way to smooth out Kate's lead vocal, while still keeping it sounding present and up‑front in the mix. This was something that proved more of a challenge, because the raw vocal track got quite edgy in parts, but was perfectly well‑balanced in others. Rather than go back to Logic's de‑esser, which is a bit on the basic side, I decided to try the Multipressor multi‑band compressor, with just the sibilant frequency band set to compress during the loud, over‑bright sections. To achieve this, I juggled the band boundaries so that band three ran from around 2‑6kHz and set that with a fairly high ratio, then adjusted the threshold so that I got 10dB or so of gain reduction on those bright peaks. Anything falling below the 2kHz filter or over the 6kHz filter passed with no compression, as the other bands were set with their thresholds high enough to avoid any compression, and although the vocal still seemed very up‑front, this process did take some of the harsh edge out of it without introducing the lisping artifacts that heavy de‑essing sometimes causes.
I also applied a little EQ cut at 3.7kHz, as there seemed to be an aggressive peakiness in that area. Logic's Compressor was inserted at the end of the chain, set to a 2:1 ratio with an Opto character, 10ms attack time, and auto release. This squashed the vocals by only 4 or 5dB on the loudest peaks and helped it sit better in the track without making it sound unduly processed.
Kate had set up some subtle delays, which remained low in the mix — I left these as they were — and she'd also recorded some backing vocals. One of the backing vocals sounded very high in pitch and seemed to have been processed in some way, so I applied Logic's Pitch Correction plug‑in to all of them, to tighten them up a bit. I didn't use pitch‑correction on the lead vocal, as it is sometimes difficult to avoid give‑away side‑effects, but as the backing vocals had a slightly processed feel anyway, it worked fine here. I also copied my Multipressor de‑sser to the backing vocals, as these had a similar edginess.
On a song of this type, I find that adding much in the way of reverb tends to detract from the intimacy of the vocal. I used a hint of short plate with 80ms pre‑delay and all the low end rolled out below 200Hz, courtesy of the Space Designer convolution reverb, on the main vocal line. For the layered chorus vocal, I opted to use Logic's cheap and cheerful Platinumverb, with the balance slider set to favour the early reflections, and around 65ms of pre‑delay. When mixed well down behind the vocal, this gave a sense of space and presence without making it sound washy. Again, if I'd been let loose with my usual plug‑in armoury, I'd have probably used a short EMT plate with pre‑delay — probably using the UAD Plate 140 plug‑in, which is an excellent tool.
Kate had recorded a short vocal link section around three-quarters of the way through the song, directly after a piano instrumental break, but aside from applying pitch‑correction to two of the four parts and adding a little of the same short plate reverb used on the main vocals, I left them as they were. It would have been easy to simply copy the Mpressor settings again, but it's important to use your ears — and in this case, I actually preferred the sound of this section without any vocal compression.
Kate's programmed drum parts sounded pretty good right out of the box. She'd used two kick-drum sounds, which I rebalanced slightly, but the only significant tweaking that was required was on the snare drum, where I used some compression (4.6:1 with around 5dB of gain reduction) followed by EQ to enhance the definition of the sound, and to remove some low‑end boxiness. A steep low‑cut at 66Hz was combined with gentle boost at 200Hz and 4.5kHz, which gave a crisper sound without the boxy overtones. The hi‑hats were also treated to low‑cut filtering, plus a little 'air' boost at 5kHz and up.
All the drum parts were routed to a single stereo bus, so I inserted a low‑cut filter at 65Hz, using an 18dB/octave slope and while this didn't detract from the punch of the mix, it did help to clarify the low end. There was also a track made up of drum loops that sat nicely under the programmed drum parts, so again that was really just a matter of balance: modern samples are usually pretty well‑recorded, so you can often get away without having to EQ them at all.
At four points in the song, there were staccato string samples playing four to the bar, and I simply bracketed these with EQ to keep the main focus of the sound between 500Hz and 6kHz and remove any unwanted frequencies. Kate had also included a Fender Rhodes part, to which I added a gentle phaser plug‑in to give the sound a bit of movement. There were additional piano lines, again from sample‑based instruments, which I just balanced with no further processing. I was pretty happy with Kate's original pan positions and the mix was generally quite symmetrical.
While I would have liked to treat the overall mix with my UAD Manley Massive Passive EQ plug‑in, and a fancy bus compressor, in this instance I simply inserted a Logic compressor over the whole mix, set to a VCA character, with 10ms attack time, auto release, and a low (1.2:1) ratio. With the threshold down at ‑27dB I was getting no more than two or three dB of gain reduction, but it was enough to glue the track together. Finally, I applied my recently‑resurrected trick of using two limiters in series, using Logic's limiter in both slots. Normally, doing this would make no sense, but Logic's limiter has both hard and soft‑knee options, with the latter softening the way in which peaks are dealt with. Setting both limiters to soft‑knee, and adjusting each to trim only a couple of dB off the overall level makes the mix sound loud and punchy, but without squeezing the life out of it — which is what you'd get if you tried to take that much off with a single instance.
It's all too easy to try to fix everything in the mix, but that approach is never the best one if you can help it: this remix job was made much easier because Kate had spent time and effort coming up with a really good arrangement, and choosing good basic sounds.
I enjoyed the challenge of trying to polish the mix using only Logic's plug‑ins, and it is a credit to Kate's initial work that I didn't feel it necessary to use any mix automation — although if I'd had more time, it would probably have been possible to improve the levels of some vocal phrases marginally, by the careful use of level automation. They say a mix is never finished — but you do reach a point where you have to ask yourself if any further work will affect sales!
The improvement in clarity through filtering out subsonic frequencies in the bass and drum parts, combined with some gentle bracketing of other instruments, to stop them spilling down into the upper bass region, really made a big difference to the clarity of the mix, and the frequency‑selective compression helped to smooth out the vocal timbre to a useful degree.
If you were to ask me what I missed most about having to stick with Logic's own plug‑ins for this remix, it would be not having the choice of more refined‑sounding equalisers, or access to my favourite UAD Plate 140 reverb — but at the same time, I hope I've managed to show just how capable a DAW's own plug‑ins can be if the original recording is made with care, and if you learn how to use the tools you have at your disposal.
Kate: "I'm extremely happy with the outcome. Being a perfectionist means that I'm seldom happy with my mixing skills — partly because my background is more musician than technician, and partly as I can drive myself mad and end up over‑tweaking a mix — but Paul has shown me how a clear approach, with logical and simple adjustments, can, in turn, bring great clarity to my mixes. Now, when I'm feverishly tweaking, at least I have better techniques and a better understanding of the gear in my audio toolbox. It seems that I don't need to go and buy expensive new tools to get a balanced mix — although I will always be tempted!”
This month's track is from Kate Ockenden, who works at Kidderminster College. The track, 'Get It If You Want It' is destined for an EP which will be released later this year, and will be sold on Kate's web site, which you can find at www.kateockenden.co.uk.
If you'd like to hear Kate's original mix and Paul White's remix, along with some before and after audio files to illustrate the Mix Rescue in more detail, pay a visit to the SOS web site at /sos/articles/oct10/mixrescueaudio.htm.