We coax different sounds from identical guitar recording setups and try to inject some attitude into a ‘70s rock track.
The song ‘Hussy’ was written and recorded by Alex Kamburov, who also played all the instruments himself other than the drums, for which he turned to his DAW’s loop library. Alex provided the growly vocals too, though there’s also a one-word female vocal part, which I’m guessing wasn’t him! Alex told me he was aiming for a sort of AC/DC meets George Thorogood kind of vibe, but that he felt his mixes lacked the necessary weight and attitude, and that the drums in particular sounded far too ‘polite’. There were a few more specific requests, too, which you can read about in the ‘Rescued This Month’ box below.
What we started out with, then, was a mix comprising a modest number of parts — something comparable with what you’d have expected back in the heyday of tape. In addition to the drum tracks (one file containing kick and snare, another for overheads, and a further one containing some occasional crash cymbals), there was a bass guitar part, doubled ‘chuggy’ rhythm guitars, a lead guitar track, a short wah-wah guitar layer at the end of the song, Alex’s main and doubled vocals and the spot female vocal.
I started by listening to the individual parts, to see if there were any obvious problems. The lack of multitracks for the drums didn’t bother me, as there was nothing wrong with the parts — they sounded exactly like a clean recording of a good drum kit prior to any form of creative processing — but I felt I might need to tame the snare a bit to achieve the right balance. The part played along at a fixed tempo (I assume it was created originally by playing to a click track) and so my Logic Pro DAW software was able to extract a stable tempo from it, and this allowed me to work to a grid. While it isn’t always necessary to work to a grid, it can make life much easier if there are timing issues that need to be addressed, as turned out to be the case here. With the tempo identified, it was simply a matter of adjusting the track start position so that the first downbeat fell at the start of a bar.
The rhythm guitar parts were tight and in tune, and I couldn’t see a problem in shaping them further using EQ or amp and speaker modelling. In the end I used both. I was less happy with the bass, though, as it lacked any real depth and it also had some timing issues. Its bright tonality also made it sound rather rattly, and it was played in a legato style which made all the notes run into each other.
Vocally, the recorded parts suffered a little popping and they were quite breathy, but in the context of the rest of the track I felt they were OK in this respect. Alex delivered them with an approximation of a Billy Gibbons growl, but the end result lacked authority. Ideally the performance would have delivered all that, but I felt I should be able to address this with a little processing.
Finally, then, I assessed the guitar overdubs. The wah-wah part was fine as it was, but there was a little suspect note bending in the solo. I tackled that with Logic’s Pitch Correction plug-in, which is similar to the original Auto-Tune, automating it to step in only when needed. The easiest way to do that for lead guitar parts is to leave it switched in all the time but to set the correction speed fader to a slow setting except for those notes that need fixing, where it can be sped up as necessary. An offline pitch process such as Melodyne or Logic’s Flex Pitch would allow more precise control over the shape of the note bends, but I didn’t feel it was needed here — basic pitch correction worked as well as anything short of re-recording the part.
Before resorting to more extreme measures for the bass, I tried a number of different EQ and amp-modelling approaches. Yet, due to a combination of the legato playing style and fret noise, I just couldn’t achieve a sound I was really happy with by processing alone — and in any case, there were occasional notes where the timing was noticeably out in relation to the drums. Timing is easy enough to fix with the slicing and warping facilities available in most DAWs, but when the tone and the timing are problematic, you’re often better re-recording or recreating the part. So, I decided instead to use Logic’s pitch-to-MIDI facility to extract a MIDI part from the existing bass audio part. Because of the sound and legato style of the original part, this conversion tripped up in a few places, but with a little manual intervention I was able to reconstruct the original bass part as MIDI and then move any errant notes so they lined up tightly with the drum beat.
Using the MIDI track to control a sampled bass guitar from Spectrasonics’ Trillian gave me a tight bass part with plenty of tonal depth, but it was a bit too clean for this song. I could have used various plug-ins to mess it up a little but in the end I managed to get a convincing sound by adding back in some of the original bass part. The timing differences were evident on some notes, of course, so I used Flex Time to line up the offending notes in the original bass part alongside my new MIDI part. I know many DAWs have an audio-quantise feature, but the note envelope of the bass part was so indistinct that I didn’t hold out much hope for it doing anything useful, and there were few enough problems that the manual approach was fine.
Sorting out the bass part, then, took quite a bit of work, but thankfully it proved to be the most (in fact the only)challenging part of the job. With most of the corrective donkey work out of the way, I could turn my attention to the more creative elements of the mix.
I steered clear of using too many esoteric plug-ins in this mix where possible, as I thought it would make the processes described applicable to a wider number of users. I set up a plate reverb on send one, as that type of reverb seemed to me to be the best fit for this style of music. I wanted a more solid and a slightly ‘dirtier’ drum sound, so I hit the kick with some 5:1 ratio compression, followed by Logic’s Sub-bass plug-in to add a little more depth. Logic’s Overdrive plug-in then contributed a little ‘hair’ to the kick sound and, finally, I used an EQ cut just below 1200Hz to tame the snare sound, which, as I said earlier, shared the same track as the kick drum. I also added some subsonic EQ cut, just to remove any of those energy-wasting frequencies that fall below the range of most loudspeakers.
The separate crash cymbal track had no spill, so I did nothing with this other than balance it with the rest of the kit. For the overheads it was a case of using EQ to add a bit of a ‘smile’ curve, with the dip centred at 790Hz, and some more low-cut filtering below 50Hz.
Logic’s cheap and cheerful PlatinumVerb was used to add some early reflection rawness, and then I used Logic’s basic Limiter in its soft-knee mode to take around 5dB off the peaks, which gave the kit a more energetic sound. I didn’t intend this to be a subtle mix, so my next step was to send all of the drums to a bus, and on this bus to insert both an Exciter and another Logic limiter. Once again, I used the limiter in its soft-knee mode, just to skim another 3-4 dB off the peaks. The Exciter was there to add a little edge, while the overall limiting pumped up the energy to a useful degree.
For the two rhythm guitar tracks, I ended up putting each through a different amp emulation and panning them slightly left and right. EQ lift was also used after the amp emulations to put the peak in the guitar responses in different parts of the spectrum, so as to help separate them further. One emulation was of a Vox-style amp, with the other a Fender Bassman type of sound. I didn’t need much drive on the amp models, as the original parts were already suitably distorted.
A slightly different approach was taken with the lead guitar sections, as I wanted to give them more of a stadium-rock flavour by using faked double-tracking and delay. To create the double-tracking effect I copied the parts onto two tracks and inserted a pitch corrector and a different amp model (Orange and Fender this time) on each. The pitch correction speed was ramped up during the suspect part of the solo and the rest of the time they were set to be much slower but, importantly, still different (approximately 350 and 500 ms reaction time). The idea was to create some subtle pitch variation between the two parts. The copied part was then delayed very slightly relative to the original (around 40ms), producing a fairly convincing double-tracked sound. Finally, I added a little tape-style echo to the copied part, with a delay time of around 520ms and enough feedback to create a repeating tail of echoes.
I’ve already touched on the bass part, but it’s worth explaining what processing I used. Once I’d sorted the timing on the original part I used Logic’s Bass Amp Designer, with its graphic EQ set to give a peak at around 1.5kHz, and also a touch of compression, courtesy of the same plug-in. The Trillian electric bass part, with which this was layered, was again processed via an instance of Logic’s Bass Amp Designer, but this time with more subtle EQ, high amp gain to add a bit of attitude, and no compression. When balanced together, the two bass parts blended well, to my ears, creating the illusion of a live, amped bass part.
For the vocals, I was tempted to use the excellent Waves Real ADT plug-in, which does a great job of mimicking the ADT effect that was so popular back in the day, but instead I decided to use the same technique as I’d used on the guitar parts — it’s a similar effect which, although slightly more complex to set up, can be achieved with the bundled plug-ins of most DAWs. So, it was again a case of copy the part, delay one slightly to give an ADT effect, and then use pitch correction on both, but set to different tracking speeds to add some non-cyclic pitch variation. I find this sounds more natural than using pitch-shifters to add a fixed amount of detuning. A low EQ cut was used to remove unnecessary low end, and I allowed myself the luxury of the Waves Vocal Rider to even out the levels — although I followed that with some fairly high-ratio compression just to keep up the sense of energy. If you don’t have Vocal Rider, then using level automation can achieve similar results, but it takes rather longer to do. I also added a little overdrive on the copied track, just to ‘warm’ it up a bit, but it was nothing too obvious. A little left-right panning and I had a nicely energetic ADT vocal that only needed a touch of reverb dialled in via a send to bring it to life.
If you listen to my isolated example of the treated vocal part (see the Audio Examples box), you can still hear hints of popping and the upper region sounds quite aggressive and messy, but somehow it sounds different sitting in the final track. That just goes to underline that final EQ and processing decisions should always be made when listening to the whole mix.
For the doubled vocal parts that came in on the chorus, I took a slightly more aggressive approach and followed gentle pitch correction with a limiter (lopping 5dB or so off the peaks) and then some fairly aggressive high- and low-cut EQ to give me more of a dirty mid-range sound with a bit of a ‘bark’ to it. A touch of reverb and that was done too. There was one other vocal track on which there was a bit of a growl, following each chorus, so I just added a generous helping of reverb and mixed it in with the rest.
The female vocal part, which just sings the word ‘Honey’ once during the song, was treated to some low-cut EQ and a touch of reverb. Alex’s wah-wah guitar part worked perfectly well as it was, mixed in with no further processing. I did, however, allow myself a couple more luxuries on the final stereo mix bus, as I wanted to create an ‘on the verge of pumping’ effect, to give the track some old-school rock credentials. This was achieved by using a UAD 1176 plug-in, set to a ratio of 8:1 and adjusted to give a fast release time and a maximum gain reduction of around 7dB. A Slate Digital VTM analogue tape emulation came next, to further cement the ‘analogue era’ character, with a UAD safety limiter bringing up the rear to trim another couple of dBs off the peaks. I did try a very mild overdrive instead of the tape emulation and it worked reasonably well in creating a vintage rock sound, but the tape emulator sounded better so I stuck with it.
Overall, I was pretty happy with how this mix turned out, as while the song itself worked well enough, the original mix sounded like there was too little energy in the captured performance, which seems common in home-studio recordings. The type of song and its style of delivery meant that it was never going to sound ‘polished’, but by leaning in the opposite direction, to help create a sense of rawness, energy and power, I think it worked to give an almost live-recording feel. I was also happy to achieve this working mainly with the plug-ins that come bundled with Logic, as it shows what can be achieved on a modest budget — although the handful of Waves, UAD and Slate plug-ins did make life easier!
Alexander Kamburov, who hails from Houston, Texas, told us about his track and what he was looking for from this Mix Rescue: “My songs usually take longer to write, but I finished ‘Hussy’ in less than a day. I wrote the main riffs one evening and woke up next morning with most of the lyrics — though my wife deserves credit for helping me iron out some of the more problematic phrases and rhyming! I recorded and mixed everything on my portable recording setup based around an HP i7 laptop running Windows 7, Mixcraft and Studio One DAWs, an Edirol UA-4FX USB audio interface, an AKG D3200 dynamic mic. I tracked in Mixcraft 5, using its loop library for the drums, but mixed in Studio One. Unfortunately, the Mixcraft project got corrupted, and I lost the DI’d guitar parts, but I had versions with amp simulation printed on, so I didn’t bother to re-record.
“The mix I submitted was my third attempt. When Paul asked what I felt could be improved, I explained that I felt the drums should have more power but that I’d exhausted my ideas on how to squeeze more kick and snare from a stereo loop. I thought the guitars, all tracked using the same setup, sounded too similar. Also, the bass needed to be a touch fatter, as my attempts didn’t quite get it where it needed to be, and I wanted to hear another mixing approach on the vocals.”
Alex: “Considering that my goal was a song with a ZZ Top, Motörhead and AC/DC vibe, Paul got very close on the first mix, which sported a ’70s ZZ Top feel. I could probably have called it ‘job done’ if I didn’t feel that some of the levels needed a slight adjustment, but I did request that he take a more in-depth look at the drums, to see if he could bring out more kick and snare in the mix. Paul addressed the drum and vox issues with mix number two, and with a couple of tweaks to dip the bass and vocals — to bring the guitars slightly up — all the balance issues were sorted in the third mix, which proved to be ‘the one’.
“In Paul’s mix, the drums have more sizzle and pop, the bass line is simply a monster, and the rhythm guitars’ tone separation is more pronounced. Paul also added a nice touch to the lead phrases, courtesy (I’m guessing, from what I hear) of a bit of a delay and stereo widening. He also managed to make the vocals stand out more, and yet somehow sit better in the mix.
“I’d like to thank SOS for facilitating this Mix Rescue, and to give an especially big thank you to Paul White for lending his time and expertise. I can’t wait to read how he did it!”
I created a number of audio examples to show what I did to the various sounds to make them work in this mix. You can find them, and full ‘before and after’ mixes, on the SOS web site.