This month, we optimise speaker placement, tame some early reflections, and get our hands dirty with a problematic mix.
Matthew Pinfield plays in a band called the Young Runaways, and he produces all their recordings, as well as teaching music at a secondary school. They are a six-piece, with the usual drums, bass and electric guitar plus the less usual violin, trumpet and trombone. Their aim is to combine pop songwriting with orchestral arrangement sensibilities. They've been together for about a year with this line-up, and are releasing their debut EP via Commercially Inviable Records on 3rd October 2011.
Matthew uses a current Intel iMac running Logic Pro, teamed with an M-Audio ProjectMix interface, plus Yamaha HS80M and Genelec 8020A monitors and a fair selection of mics (mostly SE models). He'd treated his attic-room studio using old heavy curtains, blankets and duvets, in true SOS budget fashion, and had achieved fairly workable room acoustics, but he called us in because he still couldn't get the clarity and punch he wanted from his mixes, and he was also worried that they weren't translating well to other systems outside the studio. The room is fairly small, at around 2 x 3 metres, and with some odd ceiling angles because of the shape of the roof, but Matthew had his monitors facing down the length of the room, which tends to create the fewest problems.
On our arrival, Hugh and I were greeted by Matthew and Young Runaways member Lucy, who plays violin. Matthew fixed us up with drinks and some chocolate Hob Nobs, and as we chomped our way through them, we played Hugh's test-tone and reference-track CD through the system to check out the monitoring. The Yamaha monitors were perched on some DIY platforms made from cut-up sleeping mats, and positioned close to the back of the unusually deep desk, slightly behind the iMac. The Yamahas sounded somewhat bass-light and coloured through the mid-range, while such low end as there was sounded rather uneven, and the stereo imaging was also pretty vague and confused. The desk was actually positioned behind an electronic piano, which explained the surprisingly deep monitoring arrangement.
The Genelec monitors, which were mounted on stands off to the sides, sounded rather boxy by comparison, so as I fumbled with the controls on Hugh's camera, he made his usual trip to the back of the monitors to see how they were set up. Sure enough, Matthew had rolled off a lot of bass from the Yamahas and had done some tweaking with the DIP switches on the back of the Genelecs, so step one was to set everything back to flat. But before we played the test CD again, I suggested we move the Yamaha speakers further forward, to minimise the area of reflective surface between the monitors and listening position, and to bring their baffles forward of the iMac screen. We also replaced Matthew's DIY speaker platforms with Auralex MoPads, which made a significant improvement in cutting the amount of vibration getting to the desktop, and then angled the speakers up slightly towards the listening position. We were concerned that the Genelec speakers were set up rather far apart, so we repositioned these on top of the Yamahas but angled them downwards, using the spare foam wedges that came with the MoPads to achieve the required downward angle.
A re-run of the test tones showed a more even, though still not perfect, low-end response. Nevertheless, it was still far better than we would have expected from the room size and shape. A few more minor adjustments to the speaker positions and we had it as good as we could get it, with the final tweak being to reduce the high-frequency output of the Yamaha monitors by 2dB, to make the sound less aggressive. The bass end sounded far more consistent and better balanced now, and there was a noticeable improvement in stereo imaging, along with much less mid-range coloration. Switching to the Genelecs also gave similar improvements, and the forward-sounding nature of these little speakers provided a useful real-world comparison against the fuller and slightly more recessed character of the Yamahas.
There was little to do in the way of room acoustics, as Matthew had done a pretty good job with his blankets and duvets, though the side mirror points were still partially exposed, so we put up just two Universal Acoustics 600 x 600mm panels, one propped up on the window ledge and one on the right-hand wall. This improved the stereo imaging further, and so was worth doing.
Now it was time to look at one of the mixes Matthew was having problems with, and as soon as he hit play we both identified the main problems as being too much reverb, too much going on in the lower mid-range, and a lack of transient detail. Actually, the mix wasn't that far off the way you'd expect a song to be mixed in the '80s, and the arrangements were really good, so I was looking forward to rolling up my sleeves on this one.
On 'Closer' (the track I focused on), Matthew had sung and played everything himself, including live drums, yet had still managed to make it sound like a live performance, which is no easy job. He'd used Logic's Amp Designer plug-in for all his guitar sounds, and had miked up his drum kit and got a pretty good sound from it, despite the relatively small recording space. Most of the guitar parts had been doubled, and there was also an orchestral section running through most of the song, comprising live violin and sampled strings, with trumpet and trombone parts arranged over several tracks. There were also loads of backing vocals and doubled vocal parts. Matthew had polished up his main vocal line using Melodyne, and was using Logic's Pitch Correction software to keep the layers and backing parts sounding tight.
Over the course of an hour or two, we made a lot of tweaks to this mix, so I'll just go over the main points, as much of what we did has been covered in great detail in previous SOS articles. Matthew's iMac was struggling with all the separate instances of the Space Designer reverb he'd added to many of the individual tracks, so they had to go, and were replaced with a more conventional aux-send and reverb-return arrangement. This involves using only a few reverb plug-ins, but placing them on buses rather than individual tracks. By sending varying amounts of the tracks to those buses, you can adjust how much reverb is added to each track.
Next, it was on with nearly all the mute buttons so that we could start re-balancing with just the bass guitar and the kick drum, using low-cut EQ to take out any subsonic mush. The kick needed gating to take out spill from the other drums, and I used a limiter, rather than a compressor, to give it a bit more definition. It also sounded the least satisfying of all the miked drum tracks, so we ended up with some subsonic cut plus a little 90Hz boost to give it some thump, then added a little cut around 180Hz to stop the mid-range getting too 'tubby'.
I boosted the high mids on the kick to bring out the 'click' a little, and to make it more audible alongside the bass. The bass benefited from slightly less compression than Matthew had originally applied, and as its main character was in the 250Hz region, I lifted that slightly to give it emphasis. Some high cut was also applied to the bass above 5kHz, to reduce string noise.
Once the kick drum and bass were sitting comfortably, I brought up the snare, which Matthew had miked top and bottom, and used some delay on the lower head mic to give the snare a slight doubling effect. He wanted a brighter snare sound, so rather than using EQ, I used the Enhancer plug-in on the lower mic to add some snap, and then added some delay and grit with Logic's PlatinumVerb, by combining pre-delay and a very short reverb. All the diffusion and density controls were dialled to minimum, and the ER/Tail slider was set to favour the early reflections. Once balanced with the dry sound, this produced a really solid and lively snare tone.
Like the kick, the toms needed gating to get rid of the inevitable resonant mush, while the overheads, which actually sounded great, were given a little high-shelving EQ boost above 6kHz, just to lift the cymbals slightly. We then re-routed all the drum mics to a bus, so that the overall kit level could be controlled by a single fader.
Matthew had heard about parallel compression but wasn't quite sure how to set it up, so to give the rhythm section more depth, we set up an aux send on both the drum bus and the bass guitar, and fed that to Logic's Silver Compressor — one of its low-overhead plug-ins that nobody ever seems to think of using. Set to a high ratio and with around 25 to 30 dB of gain reduction on the peaks, it delivered a wonderfully brutal, squashed sound that, when mixed in with the dry signal, really beefed up the sound without blurring the transients.
Three tambourine tracks had been added over some sections of the song, though all of these had too much reverb and sounded far too wet, so they were simply dried up and balanced with the drums.
Next came the vocals, and in this case the dripping reverb was replaced by a short plate (around 0.6 seconds). This was fed from an aux send, with a tape-delay plug-in placed before the reverb to give us a fairly long, repeating echo, so now we had a dry vocal followed by some ambient delays, rather than by a solid wall of reverb. The reverb amount was adjusted as we added parts back into the mix, so that the effect was only just audible in the final balance. This gave a nice up-front and modern-sounding vocal with just enough reverb and delay to stop it sounding stark and exposed. We added a little shelving EQ above 8kHz, plus a decibel or so of very broad boost at around 1kHz just to help it push through the mix. Around 6dB of compression was enough to even up the lead vocal and get it sounding confident and consistent.
All the layered vocals were stripped of their reverb, and the Speed controls on the Pitch Correct plug-ins slowed down a little to reduce the chorus-like artifacts you often hear when multiple tracks are too tightly tuned. Matthew had made the assumption that the more layers you added, the more reverb you needed to glue them together, but we demonstrated that the exact opposite is true. By using just a dash of the main vocal's delayed reverb, the layered parts gained a lot of clarity but still sounded rich and textural.
On a multi-layered vocal harmony part, I demonstrated how to further thicken the sound using subtle pitch-shifting to add or subtract between five and eight cents from the original pitch, and then layering this in with the existing sound.
A similar detuning strategy was applied to the guitars, all of which had been doubled. The acoustic guitars simply needed some low cut to to lose the mud without reducing their definition, and they needed virtually no reverb, as doubling the tracks added the necessary complexity to the sound. The electric guitar sounds Matthew had chosen were actually really nicely done, with just a hint of edge to give them attitude but not enough to make them sound messy. He'd used a modelled spring reverb, which sounded appropriate, and that just needed backing off so that there was a little less of it.
Matthew's orchestral samples already included some ambience, so didn't really need much more in the way of reverb, although one of the more prominent violin parts did benefit from having a bit sent to the vocal reverb/delay aux to soften it a little. There was clearly little I could teach Matthew about arranging in general, though I did point out that I thought the song got a little too busy too early, and that it would benefit from the 'mix backwards' approach, where you set the balance for the loudest section (in this case the end of the song), and then work backwards to remove parts and adjust balances, so the song can build. I suggested dropping out the orchestral sounds right after the intro and then bringing them back around halfway through the song. There was also a section rich in vocal harmonies and backed by the acoustic guitar that, when stripped of the orchestral sounds, created a nice breakdown.
Other than that, our interventions were largely confined to removing low end from the pad and ear-candy parts to keep the lower mid-range free from clutter. I also set up my now-familiar chopped-16th gated effect on one very heavily reverb'ed guitar track that was intended to serve as a sound effect. I did this using Logic's Tremolo plug-in, set to a tempo-sync'ed square wave so that instead of the effect clouding the mix, it actually reinforced the track's rhythm.
Having done a bounce of our revised mix to compare with Matthew's original, we passed on a few tips relating to general DAW usage, not least the benefit of setting up a good starting template from which to work, with all the commonly used tracks named and labelled, and, where appropriate, given icons and colours. This template could also include basic send effects, such as the vocal treatment and parallel compression we'd used in the previous mix. We also looked at Logic's Screen Sets feature, which allows different windows or combinations of windows to be called up using numerical keys. This makes life much easier when working on a single screen. Again, these can be built into the template song, so that you have an easy way to flip between the arrange window and mix window, for example.
I had already shown Matthew the benefits of bussing effects, such as reverb, to save having 20 different instances taxing the CPU at the same time, and also that combining common groups of sounds, such as drums, backing vocals and orchestral sounds, on their own buses made the mix much easier to manage.
The first thing that became clear during this session was that having a set of well-recorded, well-arranged tracks makes mixing very much easier than having to deal with sub-standard recordings. We also found that, with the exception of the kick track (which required a bit of work to get it sounding OK), Matthew's mix started to sound better as soon as we hit Bypass on the mountain of plug-ins he'd deployed!
Current production fashions dictate using reverb in a fairly subtle way, so having the ambience mixed lower led to a more modern sound. Filtering out unwanted subsonics and keeping the lower mid-range under control, meanwhile, meant that we ended up with a cleaner-sounding mix.
Matthew also said he'd learnt that sometimes choosing an unsophisticated plug-in can produce great results — he wouldn't normally have tried the Silver Compressor, yet it turned out to be ideal for parallel compression, and whereas he'd normally choose Space Designer for reverb, the gritty, trashy sound of the PlatinumVerb with the density dialled right down was perfect for adding attitude to an over-polite snare drum. We'd also managed to treat his mix perfectly adequately using Logic's own plug-ins — so while it is nice to have something esoteric to hand, we demonstrated that you can usually get perfectly good results through the careful use of what is available, as long as care has been taken with the original recording.
Matthew said: "I'd like to say a huge thank you to Paul and Hugh for the Studio SOS. My room is now a pleasure to work in and my mixes translate considerably more accurately to other listening situations. Among other things, the stereo imaging is markedly improved and the clarity of the bottom end through my monitors now is unbelievable. My room is considerably easier to mix in after your adjustments and it was fascinating to watch you work on my projects. I learnt a great deal on the day that will stay with me as I continue to develop as a producer. Once again, thank you for your input and assistance. It means a great deal.”