Some time ago I was asked to "narrate a mix" and put emphasis on the thought process and the workflow: keeping focus, handling revisions and generally making note of everything that brought me to the final mix. I realised more and more people wanted to know how something is finally branded as "completed" and how one can survive self-criticism. I thought this was a great idea and I told myself "one day I'll write a mix chronicle of something".
So what better occasion than a Zelab mixing contest, which is something most of you have already downloaded, mixed and listened to? Let's go!
2. THE SONG
At the time of this article, we just released 'A Chaque Fois' by Clement Jacques for the new mixing contest. I am still waiting for Fab's mix to put on the page and I have not listened to the album release: it's the perfect time to work on it without influences. The song is raw, energetic, with a blend of noise and alternative rock. It doesn't use special effects or the weirdest sounds to achieve the kind of aggressiveness some industrial tracks have, instead it creates edge and grit through guitars, vocals and one synth-y bass sound. Great. I like the attitude. This kind of personality is rare. I make a note to purchase Clement's record.. but only after having mixed this song.
This song is from his album "Indien", produced by Greg Bonnier, mixed by Fab Dupont and released in 2014 by Music Mansion. It's available on iTunes, here.
3. DECISIONS
My studio is in the middle of another mixing session, so my choices are limited. I can't just move the hardware and routing I've made for the scheduled band (I don't want to get killed) so I decide to make the entire mix in-the-box while still using analog summing and a couple pieces of outboard for the final touch. By listening to the track, I notice the song has two main moments: one with lead vocals singing, the other without. I immediately make the decision to turn the part with lead vocals into something that is angled and edged forward, centered and propelled at the listener. The part with background vocals is instead going to be quite wide and floaty, more watery and deeper. Then there is a third part which we could identify with the solos, but if I fail at making those two main parts interesting, no one will care about the rest of the mix.
4. DRUMS
I see two bass drum microphones and immediately blend them into a single 'ALL KICKS' Bus. Same for 'ALL SNARES'. I start blending the two bass drum microphones and decide to ditch the Atlantis and keep the D112. I have a sound in my head and I know, from past times, that the beautiful, subby, full sound of the outside microphone will only give me problems for that. But..it's better to have it than miss it later, you'll see. My idea of drums for this song is to have something concentrated, not wide, that stays behind the vocals and keeps pumping rhythm, without being too aggressive.
The bass drum is there but I need weight. No sub, just discrete and constant weight. For starters, I put an L2 maximizer on the inside of the kick and tweak it so I get a constant, 'no-matter-what' hit from the kick. Something I can count on as a foundation when things get busy. In other situations I would have never put an L2 straight away on the kick but this time I feel it's on par with my vision. Now, what about the outside of the kick? I decide to use it for all the air and the high frequency contour. This way I get the nice whistles and vibe of the Atlantis outside without interfering with the punch and transient of the D112 inside. Now, since the L2 brings up the whole kick, the bleeding goes up too, along with the tail. To keep on making things unnatural but bold, I gate the kick pretty hard (again not something I would do in other situations), then remove some of the low-mids and put a UAD Helios on it and take more stuff out, until I finally make it even bigger with some classic UAD Pultec EQ trick. And lots of 16k to add!
Bass drum (D112): Waves L2
Bass drum (Atlantis): PT EQ
Combined Bass drum:
UAD Helios 69 >
PT EQ >
UAD Pultec
Snare drum, once I blend the 451 and i5 microphones, I am happy. The trick this time is on the bottom mic, which gets an L2: I need this guy to be there at all times: the bottom is going to give me that HF content that will help define the snare and help it go through the mix. After the L2, I prolong the snare tail with some UAD AMS RMX16. I find this plugin amazing for this sort of tricks. I love the general vibe of the snare this way.
Than, more gating on the blended snare and a touch of Scheps 73 to give it more character.
Snare drum (bottom mic):
Waves L2 >
UAD RMX AMS 16
Combined Snare drum:
Waves Scheps 73
Overheads and room now need to be tailored specifically for the kick/snare pocket. Sometimes I go the other way: overheads first and then integrate kick and snare..but not with the sound I have in my head: I need the two accents strong and then the overheads will add cymbals and overall air, while the room will provide the coherence and glue for the whole kit. As you can see, it's a simple EQ and more 1176. The one on the overheads doesn't even compress, but I need that add grit (something the Waves CLA-76 really has). I narrow the overheads pan and change the levels so that the two microphones are centered and not left- or right-heavy. I find that most of the drums I like in records are actually more compact than super wide spread.
Then, I add a good blend of room and for now, I mute the toms. I want to tune into the main groove. The tamborines get a rough EQ treatment (highs only!) and that's it.
Overheads:
PT EQ >
Waves CLA-76
Drum Room:
PT EQ >
Waves CLA-76
5. COMPRESSION BUSSES
I tend to set some parallel processing busses straight away. One is mono and has a DBX 160 VU (the real one was being used in the 'real' session, but luckily I have the one from UAD). This is used to compress both kick and snare, in parallel. The other one is two separate mono channels. I don't want the left to know what the right is doing. Like with hardware, I tend to never use 'Link' for stereo units. One plus of analog gear is the width and depth it creates through correlation not being spot on. It's..'there', but that smidgeon of error adds a magic to the stereo image. So this bus is made of two 1176 compressors. I usually use Distressors for this but..guess what? Yes, taken. I set the 1176 to 'All buttons mode', slowest attack and fastest release. I didn't come up with these tweaks: they are just the easy way to an amazing 1176 sound. For now this bus will get mostly drum overheads and the room but with time it will be used by many other tracks.
(as of Day 1. More busses were added on Day 2 - see below)
Mono Parallel Compression:
DBX 160 VU
Stereo Parallel Compression:
PT EQ >
UAD (Dual Mono) 1176LN
(the EQ is used as a sidechain filter, a missing feature in the original 1176)
Mono Parallel Bass Compression:
UAD DBX 160 VU
6. BASS
I really like the sound of the bass, I don't know if it's a real bass super processed or a synth bass. In this case, I basically leave the sound as it is, merely enhancing some aspects. What I want to make sure is that kick and bass don't fight in the lows. I don't want the kick drum to be super low but I still need to make some space and make sure the bass didn't mask it. I also added some RenBass to add some 90Hz to give it more density.
Bass:
PT EQ >
Waves RBass
7. GUITARS
On guitars I used the API Vision Channel by UAD. The saturation model in the input stage is just phenomenal. By driving the input harder and making it up on the output stage, this plugin creates a buoyance that I hardly find in other emulations.
Then I move to the HP and LP filters, then shape the sound via the compressor. I carve some details here and there in the main guitars and pan the main one slightly to the left. I don't want it centered because it's gonna be in the way of my vocals/kick/snare/bass. For the chorus guitars I noticed there were room guitar microphones, and that is great. I've actually started to use room mikes when recording guitars only recently, and I now regret not having done it before. They add a depth to guitar-oriented tracks that I was always lacking. For this track I want the chorus guitars to open wide left and right. Once I do that, I feel I needed to re-balance the room mikes. I solo the guitars and make sure I like the general soundscape, then I add drums and bass back.
Guitars:
UAD API Vision Channel
(only one shown, each track has slightly different settings)
8. REVERBS, SPACE AND EFFECTS
Reverbs can ruin a good mix in no time. I know what I want so I go ahead and open an UAD EMT 140 and a 250 plates. I set the 140 to be a blatant short reverb and give it a pretty good predelay to make it slap. It's gonna make the reverb appear roughly after most transients have passed. The 250 is going to be my long, thick reverb. As with all reverbs, I tend to find the settings that sound the best to me for the vibe of the song, with a pretty high amount of volume on the return track. If I can make it sound cool when super exposed, it's going to blend perfectly once lowered. After getting the best possible result with the reverb alone, I carve it with an EQ to make it less apparent. Then I take it back to the level I like and start pushing sends into it.
Next in the FX chain are an exciter and something like a spread. The Waves Aphex Aural Exciter works great, I use it to give things more high frequencies and bite without the need to EQ them. For spread I've been using UAD Roland Dimension D. Once set, I start adding sends to pretty much everything I feel like effecting. Setting reverbs, exciters, spread and other effects to the whole session - by experience and instinct - has helped me save time and avoid using compression/equalization to achieve what I want. Some unusual things I remember in this session is that I added the 140 to the outside of the kick only, a lot of it. If you remember that's only HF content from the kick. It really helped put the kit in a definite space, rather than having the bass drum stay in your face, which was not what I wanted.
Effects:
UAD Plate 140
UAD Plate 250
Waves Aphex Aural Exciter
UAD Roland Dimension D
9. BACKGROUND VOCALS
I didn't want to consider the background vocals as separate entities. To me, in this song, they sound like a pad. So I took soe lows and low-mids out and then processed them together. A Fairchild 670 is great for gloomy compression and buoyance, plus it has a way of adding some kind of aura to everything.
Backing Vocals:
Sonnox EQ
(same settings for each of the vocal tracks)
Combined Backing Vocals:
UAD Fairchild 670
10. 2-BUSS PROCESSING
I don't think you can achieve what 2-buss processing does with anything but..2-buss processing. When forced to do in-the-box (remember there was an actual session going on in the studio, so most of the hardware was being used), I always use the UAD 33609. I don't use the limiter, but the character at 1.5:1 ratio and 100-400ms always does it for me. 3-4 dBs of compression.. it must be a trick that gets passed with generations. After that I always insert some kind of 'analytic' EQ to rebalance things. Since the Brainworx bx_digital came out I've been a fan and now that is in UAD form, it's my favorite. This is where I do most of my analytic EQing. I don't like soloing mid or side, I'd rather keep things in context and decide while looking at the whole picture. Here I end up brightening the side and adding punch to the mid. I pretty much always play a bit with the stereo width. I know it can mess up the imaging but I tend to like it in moderate amounts. After that, I 'put a smile on the track' as the veterans say. I usually use the Dangerous BAX EQ but this time the UAD Pultec sounded better in the context of the song. More lows, more highs. This saves me from adding bottom and highs to the whole mix and again, the list of sound engineers doing that is just endless. I wish I had invented that, though! After that, a Limiter. Just to keep the peaks from turning red. My demo period for the Fabfilter Pro-L was still in effect so I used that. Limiting at -0.02dBfs and that's it.
2-Buss Processing:
UAD Neve 33609 >
UAD Brainworx bx_digital >
UAD Pultec >
Fabfilter Pro-L
11. HANDLING WHAT IS 'AROUND' THE MIX
If things work, at this point I usually start handling not what is 'in front of me' but rather, what is 'behind' the obvious tracks. I keep the kick/snare/bass as a reference for my center, then add guitars. The whole sense of "around the mix" sums to as sending tracks to reverbs, exciter, spread and that stereo buss compression I set at the beginning. The fact that different things go to the same processing unit helps the coherence of the mix rather than clogging it. The stereo bus compressor helps giving sustain to the back of the mix without altering the original sounds: the beauty of parallel compression.
12. LEAD VOCALS
Once happy with the general balance of the track and the atmosphere behind it, I decide it is time for the lead vocals. The package is formed by the lead plus two added tracks, one is a sort of falsetto and the other one is a distorted part. I tweak the lead vocal in solo to get it to sound light and natural (Oxford EQ this time) and then I make the falsetto part nothing but high frequencies, while the distorted parts provided me the guts of the vocals. I blend levels and add a new aux channel, where I use another 1176 to super compress the vocals, in parallel. I go really hard on that compressor, something like a constant 20/30 dBs of compression.
(Lead Melody)
Lead Vocals (Lead):
Sonnox EQ >
Waves L2
Lead Vocals (Lead) FX Send:
Waves Doubler
Then, I make the falsetto part stereo and phasey with the Waves ADT. I end up distorting the part a bit and panning it mostly left and right. This adds gloom and a nice aura to the whole vocal part.
For the distorted vocal part, I ended up carving most of the low-mids and mids, but I I took those choices while listening to the whole vocal package. Next, an L2 limiter to keep things aggressively in level.
(Falsetto double)
Lead Vocals (falsetto):
Sonnox EQ >
Waves L2 >
PT Delay >
Waves ADT
Lead Vocals (falsetto):
Lead Vocals (Distorted):
Sonnox EQ >
Waves L2 >
PT EQ
13. TIME AND FOCUS
At that point, more or less 2 hours have passed. I tend to go fast when mixing because I know in 2 hours from the beginning my rational part kicks in and I start second-guessing myself, loosing focus and taking decisions I regret the day after. So with that in mind, I decide to print the track to take a snapshot of where I am *now*. Experience taught me that the first 2 hours of work on a mix are the most productive, for me. It's when I can really work for the sound of the record, straight to it. The second mixing session (usually on day 2) is when I go all specific on things and try and get the record to sound more expensive by fine-tuning some aspects that I created in Day 1. Day 3 is even more about smoothing the edges. However, nothing works if I wasn't creative and focused on Day 1.
One thing you will notice is that I didn't cut the background noises in the intro. The reason for this is that - when listening to the whole song before mixdown - I found that the FX in place made the initial noises sound super cool! So instead of making the song start on the first bass drum hit, I used it as an intro and also took parts of this noisy fx to the middle of the song, during the breakdown. You'll hear it. This was totally random but I think it sounds pretty unique! (in the final stages of the mix I shifted some accents to get a tighter feel).
So, this is the mix as of Day 1:
14. DAY 2
The next day, I listen to the mix and.. I like it. However, I can now pinpoint some parts of the mix that need improving.
I am missing some excitement in the mids. So I create another Aux with a Thermionic Culture Vulture. I wanted to use the real thing but guess what? Yes, taken by the scheduled session. Luckily again, UAD makes the plugin version. I know this unit very well now love it: I use it for making sounds interesting during a recording session or (like in this case) to add grit and saturation to a mix Nothing does it like the Culture Vulture. Once the Aux is ready I take some of the center content off and leave this kind of sandy distortion for the side channel. Waves Center is great for this. I start sending vocals to it, background vocals and even some guitars. Now I need to counterbalance the 'back' of the mix by adding more guitar room, tweaking something here and there..but the general mix is there. I am pretty happy with it.
New effect bus:
UAD Culture Vulture >
Waves Center
Solo guitar:
Soundtoys Little Alterboy
Lead Vocals (Melody):
Waves Kramer Tape
Then I go to the guitar solo track. Ok, it's there, it's nasty, but it could be nastier. Soundtoys has recently released a free plugin called "Little Alterboy", which really comes in handy. Just a bit on it, tuned up one octave and distorted to give the guitar more sustain and make it easier to feel in the mix. I still need something more to give it the feature it deserves. So I create two sends in which I put two different compressors. The first one gives the track more weight, the second one is there to create some moving, panning air. I achieved that with an LA-2A send to the SoundToys Little PrimalTap to create a phasey duplicated mass around the main guitar..then into an EQ that takes only the highs and puts them into a Waves MondoMod (basically an auto panner).
Solo Guitar Parallel Processing:Waves CLA-76 >
PT Delay (Haas)
Solo Guitar Parallel Comp 2:
UAD LA-2A >
Soundtoys Little PrimalTap >
PT EQ >
Waves MondoMod
Next, vocals. They still sound a bit done in a hurry. Need more mass. I use the Kramer Tape on an FX send. Absolutely fantastic for this kind of things. I am getting a fierce lead, an airy doubled falsetto and a grindy distorted bottom. Ok..but I want more, I need a bit of silk to push the details back in the mix. I end up using the Waves Doubler plugin on a separate FX send.
Lead Vocals (Melody):
Waves RDeEsser
Lead Vocals (falsetto):
Waves S1 Stereo Imager
Everything is getting ready, I like the vibe. I add some de-essing and open the falsetto vocals a bit. For de-essing I end up using two plugins: the Fabfilter Pro-DS (putting the demo period to good use) and the Waves Ren DeEsser. One is working on the the Ss, the other one on the extreme air that might be a bit too much at times.
The next steps were a bit technical. I tuned the lead vocals with Melodyne (just the melody track), I remembered I had left the toms off, so I put them in the mix (nothing extremely important about them). Then it is automation time (see if you can catch some cool moving parts in the background of the final mix!).
15. FINALIZING THE MIX
Finalizing:
Dangerous Master
Rockruepel Comp.One
Given the nature of this project, I wasn't really going to master the song. At this point it's easy to overdo things and think you are "mastering" your track while in reality you are just ruining everything. All I did was route the output of the Dangerous 2-bus into the Dangerous Master. I put my Comp.one on it, no compression, amp only mode and then I went back into my UA 2192 A/D converter. I liked a +1 on the S&M width and that's it. Notice that this was happening before my actual 2-bus processing chain, in Pro Tools. But it's okay, I handled the output in analog, went in harder on the A/D and lighter on the 33609. Bounced to track and here it is, my final Mix03.
As usual, this is not a bible on how you should mix your song, but I think it's nice that the chronicle of a mix - with the multitude of breadcrumbs I left in my train of thought - is now on paper. Well.. virtual paper, at least!
GOING IN-DEPTH
I thought it would be a good idea to let you listen to the various stems to give you a better understanding how how things sounded by themselves. I picked 30 seconds in-between the first verse and chorus, plus the first guitar solo produced these with no analog processing on the 2-buss (straight from analog summing to the A/D). My approach to the mix and makes it impossible to have separate stem sound sound exactly as they do in the final mix. The main reason comes from the fact that the sound of most auxes and busses is derived from the combination of multiple tracks being fed to the same processor, for example: in the final mix the stereo buss with 2x 1176s is influenced by everything going into it - while in the separate stems it is influenced only by the sound of that specific stem.
Well.. I think this is the best compromise to let you hear the separate stems. It won't be 100% as in the mix, but I hope it is still cool for you!
Last but not least, here is a screenshot of my Pro Tools session, at the end of the mix. I am sure I forgot some information here and there, and I don't want to make excuses for hiding something that you might find interesting or useful...hence: here is the full session (most of the bypassed stuff is for automation).
My Pro Tools Session at the end of the mix
(click here to download the full-size image)
Alberto Rizzo Schettino is a pianist, synth head, audio engineer and part of the Puremix family since 2013. He is the owner of Fuseroom Recording Studio in Berlin and also works as a product specialist for some of the best brands in high-end audio gear.
Fuseroom Recording Studio: http://www.fuseroom.com
Alberto's Web Site: http://www.albertorizzoschettino.net