We made things a little harder by tackling this month's Mix Rescue before a student audience.
Regular readers may recognise Jon Marsh from a Studio SOS visit we did a couple of years ago. I'd been asked if I'd do a mix of a student's track in front of a lecture theatre full of students at Worcester College, and the track the lecturer picked out for me was one that was recorded and mixed by Jon.
This wasn't to be a 'traditional' Mix Rescue because, other than a snare sound that Jon wasn't happy with, his mix turned out to be very capable. However, we decided that it would be interesting to leave hearing his mix until the end of the session, so that I could approach the mix in my own way and see how my interpretation differed from his.
The song started life as an analogue recording, and Jon explained how he had gone about the recording process. "The track 'Year of the Dog' is by Forgotten Sleep [www.forgottensleep.com], a Monmouthshire based three-piece alternative rock band. It was recorded at my Shed Studios in Herefordshire [www.shedstudios.com], with extensive help from the band's live sound engineer and friend, Richie Rees.
"The band comprises Matt on guitar and vocals, Hannah on bass, vocals and synth, and Gaz on drums. Recording was initially done onto a Tascam MS16 tape recorder (using brand new 456 tape), via a Behringer mixer, but the recordings were transferred to the computer for mixing and editing. Mastering of the track was done using Logic 's Multipressor and Adaptive Limiter.
"I mic'd the top and bottom of the snare using a Blue Ball on top and a Shure PG56 underneath. We used a separate hi-hat mic —a Samson pencil condenser — as well as stereo overheads. The drums were recorded separately from the rest of the band, using a Pearl Export kit with Remo pinstripe heads and Sabian AAX cymbals. I used a Pro Series kick microphone, placed half way into the kick-drum shell, pointing at the beater. There were Shure PG56 mics on the toms, plus Shure PG81s as overheads, with a Rode NT1000 as a front ambient microphone. Once everything was recorded to tape, it was all transferred into Logic Pr o on a first-generation 1.8GHz G5 with 2.5GB RAM, via a MOTU 828 MkII interface and Behringer ADA8000 expander.
"The bass guitar was recorded using an Ampeg V4B head going through an Ampeg classic 6 x 10 cabinet with horn. To mic it, I used an AKG D550 placed about two inches from a cone (slightly off-centre to try and get a more rounded, smooth sound). The distortion on the bass was a Rat distortion pedal, modified to give better bass response. The clean sound was just the amp, with the gain set to nine to drive the valves. The mic signal was tracked via a Focusrite Platinum Trackmaster preamp.
"Matt's 4 x 12 guitar cab was mic'd from around one-inch away, using an SM57 dynamic mic and a Rode S1 condenser mic, to give us a choice of two sounds to work with. His amp is a custom 30W design, made by a friend who spent months with Matt tweaking the design, trying to get the best sound out of it. Matt's Gibson SG was also processed through a Boss Blues Driver pedal. These mics were fed into the two channels on my SPL Goldmike preamp.
"All the vocals were recorded via a Rode NT2A with a pop shield, going through a Focusrite Voicemaster. The synth part was played on a Nord Lead 2, and the mix was monitored on Alesis M1A speakers. What I liked least about my mix was that the drums didn't cut through as much as I would have liked. I also found it quite hard to balance the lead vocal against the rest of the track, as it seemed either too high or too buried."
MP3 versions of both Jon's and Paul's mixes of 'Year Of The Dog' have been posted on the SOS website at: www.soundonsound/sos/jul07/articles/mixrescueaudio.htm.
To make life easy, the project was supplied to me as a Logic project file. In many ways, this was a fairly straightforward track to work with, as there was only one guitar part, one bass part and a short, spikey synth section playing over the drum track. However, the guitar and bass both made good use of distortion, which often makes it hard to separate the two. As I was doing the remix in front of a class of students, I decided to keep things simple by limiting myself to Logic 's own plug-ins. Although this decision meant I wouldn't be able to use some of my favourite plug-ins, we still arrived at a decent-sounding mix.
Rather than trying to shape the guitar and bass sounds using EQ, I used Logic 's Guitar Amp Pro, so that I could also try various amp and speaker emulations. For the bass, I wanted something that would round out the sound without robbing it of depth or attitude, and settled on the American Clean amp, with British 2 EQ, through a 4 x 12 cabinet, mic'd off-axis using a dynamic mic.
In the end I used both guitar mics (which had been recorded on separate tracks), with a different Guitar Amp Pro setting for each. For the dynamic mic I used the American Clean amp again, this time with American EQ and a US 1 x 10 cabinet. For the Rode-miked version, I chose a clean tube-amp setting, with Modern EQ and the same US 1 x 10 cabinet. This gave a solid sound with plenty of bite, but at the same time it focused the sound into a narrower section of the audio spectrum, making it easier to mix because there was less overlap with the bass. I also compressed the bass, because it was tending to sound a bit out-gunned in the sections where it was played without distortion. By extending the attack time to 14ms, I was able to retain the attack of the notes. The last bass note over-ran the final drum hit by a couple of seconds, so I made an edit to bring the last bass note forward, giving the impression of a much tighter ending. There were a few very minor timing discrepancies in other parts of the track, but nothing that detracted from the vibe of the song so, other than tweaking a tom fill near the beginning, I left those alone.
The drums had been pretty well recorded, except that the tom tracks had so much spill on them from the rest of the kit that it wasn't practical to gate them, and even if they were isolated manually in the waveform editor you could still hear the spill come and go as the tom hits played. My quick and dirty solution was to isolate the hits by silencing everything that wasn't a tom hit, then processing the result using Logic 's 'Audio to Score' feature, to convert the tom hits to MIDI notes. I had to change the note values afterwards to match up to my tom samples, but otherwise the process was straightforward and allowed me to trigger some suitable-sounding EXS24 tom samples to replace the originals. Once mixed in with the sound from the overhead mics this sounded perfectly natural.
The guitar had been recorded by close-miking an amp with a Shure SM57 and a Rode S1. Both signals were used in the mix. Even though the signal had already passed through an amp, Logic's Guitar Amp Pro was used to smooth the sound a bit more.
Listening to the kick drum in isolation revealed it to be lacking depth, so I used Logic 's EQ to boost at around 80Hz and 4kHz, which emphasised the thump and click of the drum. I also cut back at around 180Hz, where I could hear an unpleasantly boxy overtone. Then I gated the kick, adjusting the release time to get a natural sound. I set a fast attack on the gate, with 1ms of lookahead, so that the start of each hit would get through the gate intact. A little compression was also applied. To add more depth, I experimented with Logic 's Sub Bass plug-in, which, used sparingly, does give the impression of another octave below the original (though using this plug-in in this way really is a salvage manoeuvre, and in this case I could easily have got by without it). I also found that the click of the beater could be further enhanced by using Logic 's Exciter plug-in, the exact amount being adjusted when the whole mix was playing.
I was very pleased that Jon had recorded both the top and bottom of the snare drum, as it gave me a lot more scope to massage the sound, and it was the snare in particular that he'd felt sounded dull. Gates were used to clean up the top and bottom mics and, in addition to EQ on the top mic, I again employed Logic 's Exciter plug-in to add extra crispness. I also set up a dedicated plate reverb for use only on the snare drum. The lower mic picked up much more sound from the snare wires than did the top mic, so it was used to add in the required brashness. For the rest of the kit I used a shorter, bright room ambience, to give the drum kit some life without making it sound too wet.
The overheads sounded as good as you could expect using budget mics in a relatively small studio, so I used some broad 10kHz boost to add gloss to the cymbals, while rolling off the low end slightly to clean up the sound. A little lower-mid dip to tame some boxy overtones also helped. The whole drum kit was then routed to its own stereo bus, so that it could be controlled by a single fader and processed as a single entity if required. (If you use Logic and route via the buses in this way, don't forget to switch the plug-in delay compensation to 'All', rather than just to tracks and instruments.) I didn't use the ambience mic, as the room ambience from small studios is rarely very flattering.
As the synth sound almost blended with the guitar, I treated that to a tempo-sync'd delay and then left it pretty much alone. However, I had a bit more fun with the lead vocal. I added some compression, as usual, but also employed level automation to get the vocal to sit correctly in the various sections of the song. This invariably sounds better than relying entirely on compression, and because level automation is so flexible you can bring up individual syllables if they seem to be getting lost. For interest, I decided to try my favourite Platinumverb slapback effect, which has more than a hint of John Lennon-meets-Oasis about it, and creates a nicely live rock feel. It relies on setting a pre-delay time of around 80ms to create the slapback effect. You then move the early-reflections/reverb-tail slider fully to the left, so that only the early reflections are heard. This tight bunch of reflections sounds more organic than a single slapback delay, and adds a lot of character to the sound. I used Logic 's automation to increase the mix of the effect during the choruses. To this I added some gentle EQ and a hint of Exciter up above 5kHz, to emphasise the sense of air and to help the vocal project above the distorted guitar and bass. A similar but slightly less obvious effect was used on Hannah's backing vocal part, which I also tweaked in Melodyne prior to starting the mix, just to tighten up the pitching a little.
It's worth pointing out that the final EQ and balance adjustments on each of the individual parts were made with all the parts playing: if you work on sounds only in isolation, you may find that they don't sit together as well as you expected when you come to balance them in the mix.
With this preparatory work done, building the mix was pretty straightforward. I first balanced the drum-kit elements to produce a workable kit sound, then added the bass. Juggling the guitar level to keep it up-front without diminishing the impact of the rhythm section is always a bit tricky, but the reshaped tonalities achieved using Guitar Amp Pro really helped. I also panned the two guitar mics slightly apart to give a wider image. In conjunction with the level automation on the vocals, the mix actually worked pretty well and kept a live vibe, even though it was clearly recorded in various takes.
Although it is never a good plan to master tracks in isolation, I used Multipressor (just as Jon had done) and followed this with Logic 's Limiter plug-in to coax the maximum loudness out of the track without allowing it to clip. I chose a three-band setup on Multipressor, with crossover points at around 100Hz and 1.5kHz, dipping the level of the mid-range slightly to increase the sense of clarity and loudness. As usual when mastering, I chose low compression ratios and low thresholds, so that the entire dynamic range would be squashed slightly, rather than just the top few dBs being squashed hard. The limiter was set to trim a couple of dBs off the peaks but I didn't want it to become too intrusive, as can be the case when excessive limiting is used to try to inject more loudness.
Overall, I was pretty pleased with my mix, though I know I have a tendency to try to make some things sound more 'polite' than perhaps the composer originally intended! I checked out Jon's mix after I'd finished mine and was glad that I'd managed to come up with something a little different without departing too far from his original concept, though the way we both approached the vocal sound was quite different, as you can hear from the audio files on the SOS web site.
"I think that Paul's vocal mix is a lot more poppy than mine, down to the slapback reverb sound he used. In my mix, the vocals are lower down, as the band wanted the vocal mix to be similar to Kyuss (vocals quite buried). But, as Paul said, each person mixes differently.
"I notice that the drums sound more punchy and full in Paul's mix, and the guitars sound more defined. The bass sits very nicely in the mix but still manages to sound warm and fat. I really like the extra touch of putting a tempo-sync'd delay on the synth part.
"I also sometimes use the 'Audio to Score' feature, together with EXS24, to reinforce any lacklustre drum sounds, although I do try to get the sound right first!"