ANATOMY OF A MIX RESCUE

I recently completed a mixing/mastering project for a band from Los Angeles called Flaamingos. I thought it would be a good task to discuss it here in the studio blog. The purpose of this piece is not to dis anyone's work, but rather explain the mix process a little bit, demonstrate common mixing problems, and lastly, validate the importance of having an engineer that understands the content they'll be mixing.

A little background on the project: The band's label, Felte, approached me to work with Flaamingos on a few songs, as the band thought their original mixes could be nicer.  Even though their songs were recorded at a proper studio with top of the line gear, there was something missing. The band had mixed this track we will be discussing at their home studio in-the-box, i.e. entirely mixed in a Digital Audio Workstation (DAW), without a mixing console or any outboard gear. There's of course lot of discussion these days about the pros and cons of mixing ITB - I've myself talked a little bit about it on this website. I'm a firm believer in analog summing, so I'm biased towards this technique. 

Before I started to work on any of the material, I listened to the band's demos. They also sent me some mix references, which consisted of many things I've listened over the years, from Gang of Four to The Jesus And Mary Chain. I'm personally very fond of post-punk and so-called darkwave music, having spent most of my teenage years listening and obsessively collecting those kind of records – think anything on labels like Factory, Rough Trade or 4AD. Thus, I'm highly familiar with how this music should sound and how to creatively make use of studio tools to not just emulate a style but rather form part of a collection of music that still inspires me today.

Let's start the dissection by listening to the original master and looking at the mixdown audio file.

 

The first thing I noticed right away: this was mixed with a limiter on the master bus. Notice how the mixdown file peaks are all identical, the file itself resembles a square wave. In my humble opinion, a good mix – one that has dynamics – should never look like that. Here we have barely any dynamic shift between the verses and choruses, they all look (and sound) rather similar, making the song frankly quite flat and unexciting. It also leaves no room for an engineer to fine tune it when it gets to the mastering stage.

Now let's listen to what the Black Knoll Studio master sounds like and take a look at the mixdown file.

 

Notice how in this version, we now have many peaks and valleys with defined dynamic shifts. The choruses for example get louder and sound bigger than the verses. We also have plenty of room for the mastering stage, which means we can sweeten the track and get a descent final audio level. 

Now let's breakdown the mix by its elements. In this section,  I'll discuss and explain some of the decisions I made while mixing down.

BUILDING THE FOUNDATION: DRUMS AND BASS

The first thing I work on whenever I'm mixing a band, it's the drums and bass. The snare particularly can make or break a rock song. Recently a very savvy computer geek decided to replace the iconic snare drum on Metallica's “Master of Puppets” with the crappy snare sound of that forgettable (and embarrassing) Metallica album. The difference of course, is enormous. I was crying laughing the first time I heard it – it reminded me how Hollywood keeps ruining all of my childhood memories with remakes! Anyway, back on subject...

Whenever I'm mixing, as a rule of thumb, I try to leave myself plenty of headroom to work in the mixer. I also make sure to leave plenty of headroom on the files themselves. For the most part, I usually work with an RMS of -18db on each track. That's my starting point. Side note to all you wonderful folks recording too hot: it's NOT necessary – recording hot won't make your track louder, so stop doing it.

The first drum kit microphones I selected to shape into form were the overheads and snare mics. Before touching a single knob I listened for phasing issues. Once I've addressed those, I can start working on the drum sound itself. Luckily, drums were tracked properly, so I moved to the fun stuff. I'm from the mixing school that tends to get the majority of the snare sound from the overheads, as opposed to from the snare mic alone. I actually use the snare mic to just add punch. This song didn't use cymbals, but it used toms quite a bit, so I used a Distressor-style compressor to tame those sounds bleeding into the overhead mics (along with the hi-hat and snare). Once I finished working the snare sound, I moved to the kick drum and hi-hat microphones, which I blended with the overheads, EQ'd and sculpted the meat and bones of the drums. Finally, I worked on the toms. I usually apply gates to all the drums mics, except for the overheads. I find it's easier to work on drum separation when all the pieces are properly gated and there is little bleeding from one mic into the other. The band didn't track any room mics for the drum kit, as they were going for a closed-mic sound (similar to what you hear on those 70's punk albums by The Ramones), so the overheads provided the little ambiance and depth.

Once I had all my drum kit pieces balanced, I crosschecked the meters and worked on my pan pots. I usually pan my drums from the drummer's perspective, that is with the hi-hats panned slightly to the left channel and toms ending with the floor tom on the right channels. I know this is a big point of debate among mix engineers. My philosophy is simple: if I want audience perspective, I can just go out and catch a live show. That said, I do not pan drums too wide and try to keep them fairly centered by not hard panning any individual drum mics (except obviously the stereo overheads, which are panned hard L/R). One last thing while talking drums: any stereo pair I will bus to a stereo channel and EQ together as a L/R signal, as opposed to EQing each microphone. Why you may ask? Well, in order to conserve the phase relationship between the stereo miked pair, it is better to EQ  it as a whole, rather than individually. *This is another one of those long subjects, so I will revisit on another blog posting*

On the analog mixer, the drums peaked at -12db's after applying EQ's and compressors. That's a good amount of headroom for me, so I move to the bass guitar.

Stylistically, I added a chorus effect to the bass and tighten it with an 1176 compressor to keep it consistent throughout the song. I made it glue together with the kick drum by correcting issues with their respective EQ's, separating the kick and bass guitar frequencies so they wouldn't clash but rather complement each other.

As I played back this first subgroup of drums and bass, I worked on the snare sound a little bit further, using a Pultec EQ to make it sound more interesting and inside the mix. I also decided to send the snare mic through an AMS reverb to get that classic post-punk sound ala The Cure.

Finally, to gel the drums and bass, I used the classic SSL G-Bus compressor in parallel mode.

LAYING DOWN SOME MELODIC TILES: SYNTHS AND VOCALS

My next mix element to focus on were the synthesizers, which usually add color and melody - in this particular track, it carries a lot of the song's midrange frequency information. The original mix had them sounding too clean. They almost sounded like a VST-synth. Thus I dirty ‘em up quite a bit with some valve hardware; the Culture Vulture really added some nice saturation and grit to it. The EL-7 FATSO did some wonders too, making them sound warmer, with a specific tonal character reminiscent of those hazy sounds from goth rock's heydays . After I ran all the synth tracks through the hardware and captured again in my DAW, I EQ'd each so they would sit better in the mix. I also added some creative use of effects like reverb and delays, again, to evoke some variation and excitement within the song.

Finally, I worked on the vocals after I had a fairly solid instrumental mix going. The original mix's vocals felt a bit lifeless. They needed some EQing to make them cut through the mix. Even though they now sound much more present than in the original mixdown, I did not boost the high end at all. I simply EQ'd out problematic frequencies in the low and low-mid range, then I ran through a Teletronix LA-2A limiter to even them out. The end result is they sit better in the mix and you can hear them more clearly.  I had to correct some tracking problems, like harsh consonants sounds like K's and S's – I ended up using a de-esser to address the sibilance. After correcting some more tracking issues, I leveled the vocals with the snare drum, just sitting a hair on top. One trick I love doing is making all the percussive consonants and breathing of the vocal track match the level of the hi-hat, so I worked on that for a bit.

Creatively, I went for an early-80's Andrew Eldritch vocal sound. I added the tape echo effect prominent on the Sisters of Mercy's classic Temple of Love. Another cool trick I did on this particular track was placing a tape machine simulator at the end of the vocal reverb. I set the wow & flutter parameter so you would hear the notes detune as the reverb decays. Yes, I blatantly lifted that production trick from the great Martin Hannett - but as Picasso said once: good artists copy, great artists steal.

At this point, the mix started to sound more like a record. Once I had a rough mix, I went back to address any leftover technical problems and side-effects from the mix process. I checked all my effects for phase problems, making sure nothing was getting thrown out-of-phase. I EQ'd some of the effects and made them fit inside the mix. I felt the track needed a bit more spacialization, since the original mixdown felt too centered when it came to the panning. I ended up panning the multiple synths a bit and positioning them in different areas of the stereo field with the purpose of creating layers and a sort of 3D image. I'm a big fan of the multi-layered and spacious mix style of British bands produced by the aforementioned Martin Hannett. I had some room for creative use of delays and reverbs in the song, so I also played around with some automations. In the end, I added my very own aesthetic: mixing is not just a technical job afterall, it is a creative one.

Once I was happy with the first mix, I printed it. I let it sit for a day or two, having a little break and freshen up the ears. Then listened again in the studio, in my headphones, and in the car. The idea with this multiple sound system listening exercise is to hear which elements are more dominant than others. I listened to the song and made some notes, all without the saturation of a 10-hour mixing session. After making some adjustments, I was satisfied with the mix and sent it to the band and label, then awaited for their feedback. Of course, there were some small changes requested: a synth here or there needed to be a hair louder in the chorus, the bass guitar in the verse a hair lower. That sort of thing is normal – it's not my own music and the band's opinion is what ultimately matters. But overall, they were very pleased with how it sounded. I made the adjustments requested and voila, we called it good. I then printed all the stems, printed the 2-track mix and created my backups.

All said and done, listening back for the purpose of this blog entry, I'm still enjoying listening to this mix. And I'm not the only one enjoying it: the specialized music press has enjoyed it as well.

But what do you think?