There has always been this mysterious picture about mixing music. We all know what happens during a recording session. It has been seen various times on TV and engineers and artists talk very openly about this part of the (pre)production process. However when we switch over to the mixing phase of (pre)production it all because very mysterious. We have heard certain things and to an extend we know what goes down during a mixing session in a very basic sense. Mixing engineers appear to close up when asked, instead of opening up, sharing, and starting discussion about varies techniques. Luckily this is slightly changing nowadays and information, thanks to the internet becomes more accessible. Thanks to websites and magazines like Sound on Sound and Waves.com. The last one in particular is offering free plugin presets recently for Waves product owners.
Why be secretive about what we do, what tricks of the trade we are using and applying when we are, nine out of ten times, ‘re-inventing the wheel’ anyway. Besides the music industry is a billion dollar industry with plenty of artists and producers out their in need for mixing engineers. I can’t be there to do it all, neither can be any colleague of mine. Why not hire interns with potential and loads of talent and above all the right attitude who eventually will, and be able to, fulfill a position within your company? Get more clients, give back to the community, and on top of that earning money doing this. To me this only seems like a Win -Win situation, but I can be wrong.
Recently I was over at Larrabee Studios on Lankershim Blvd. in Studio City, a block away from where I live. I met up with colleague and master-mixer Dave Pensado discussing these very same issues. Dave, who runs two rooms within the Larrabee facility (B & C) simultaneously, occupied by his second engineer and his assistant, and an additional set of runners, has that structure I’m talking about. I also need to acknowledge that Dave taught me a lot of tricks among another guy -dear friend and colleague Richard Furch (emixing.com/ tribase productions) and I have always been supported to the fact of sharing, discussing, and publishing mixing techniques. However I never like my mixes anyway. I always think everyone’s mixes always sound better to me than my own. A lot of times I think the rough mix sounds better. I break out in cold sweat a lot of times when another major project comes forward. A project they booked me for as one of the mixers, especially now this is no longer a hobby for some time, this is serious business and the client pays a big amount of money for my services. The thought coming up eases me down. A&R managers and artists, and producers don’t choose you as one of the mixers if they don’t like your sound. It’s like they say about magicians, they know how the trick gets executed, the public simply only sees the trick – the magic. On other words, the magician can never see his own trick like the audience can. The audience sees his assistant getting cut in half, but the magician sees the actual trick.
Mixing requires both halves of the brian, left and right or in technical words creative and technical; Law vs freedom, science vs. breaking the rules. So whenever I don’t feel creative for whatever reason that day, I start working on the technical side and eventually my creativity kicks back in. No matter what happens I always let the client leave with a good mix.
Any aspiring engineer, audio engineering student, teachers and professors, and engineering colleagues are free to openly discus techniques and views on this and techniques in general by emailing me on royalcollegeofsurgeonsinc @ gmail . com
Enough talk… it’s time I get to the point of this blog. I’d like to share one of my mixing sessions with fabulous and very gifted UK singer/ songwriter Keisha White to the public.
SONG: “WRONG AND RIGHT”
ARTIST: KEISHA WHITE
Producer & Arranger: Harmony for H-Money Music
Writer: K. Nicole White
Vocal Producer: Harmony
Recording Engineer: Harmony
Recording Studio: H-Money Productions, Tottenham (UK)
Mix Engineer: The Music Surgeon for Royal College of Surgeons, Inc.
Mix Studio: Jasonair Music Studio, London (UK)
Mastering Studio: Metropolis Studios, London (UK)
Executive Producers: J. Dias, K. Saint, and J. Ferguson for J&J Artist Management
Executive A&R Manager: K. Saint for J&J Artist Management
Publisher: J&J Artists
Label: J&J Artists
Distribution: iTunes (mp3 only)
Pro Tools Version: 7.3.1 HD running on a Apple Mac Quad Pro G5
Time Signature: 4/4
Key: F Major Accidentals: C/ D
Duration: 3:40 min.
Console: Digidesign D-Command 24
Monitoring: Yamaha NS10’s, PMC TB2A’s, Genelec 8050 A’s, and Genelec 7000 series subwoofer
Waves Bundles: Diamond Bundle TDM and SSL 4000 Bundle TDM
Intro: Mixing “Wrong and Right” for UK R&B sensation Keisha White wasn’t a hard task. It’s her latest release after a two year period of media silence. I have much respect for her management team for putting out this track all by themselves. The track is a straight forward R&B song. The BPM speed was quite interesting. The song is produced and recorded on a rather fast speed of 122 in terms of R&B music, but it was composed at half the speed (61 BPM). The reason, make it easier and more natural sounding for the remixer of the song. Which in my opinion is genius! The producer and recording engineer of the song Harmony did a great job getting the emotion of the singer, Keisha White right. He prepared the session for mixing above expectations. The song was mixed at new studio facility Jasonair Music Studio in London UK at the end of February of 2008 entirely “within the box” and mastered at Metropolis Studios in London UK. Both studios are located in London’s famous West End area.
During the session all my dynamic processing and sound design happened “post-fader” as I don’t want any thing to change on the input when adjusting leveling (gain). On another note I inserted the Waves PAZ Analyzer a few times a cross the board to find annoying “ring” sounds so I could easily and time efficiently eliminate them. Audio analyzers are an extremely powerful tool, both during recording and mixing. You can basically see what is going on exactly on that channel in terms of sound, from good to bad and be able to fix it quick!. However you need to know how to read them correctly and don’t depend on it only. Please depend on your ears, cause you are an enginear and not an engineye.
I also ran out of the available Pro Tools Voices (HD 3 system) quickly. My assistant engineer flew the vocal tracks around to free up some Voices by letting some vocal tracks share voices with each other when not played simultaneously at the same time. We needed to downgrade the session half way through the mix session too, from 88.2 kHz to 44.1 kHz which is an extreme sacrifice however it’s still CD quality, to gain additional voices and more processing power, which was absolutely needed, but we contained 24 Bit rate resolution for mastering.
At the end of the day I invited the producer of the track Harmony down to the studio to come and revision, tweak the overall leveling and panning of the song with me, as the producer holds the emotion of the song and I like to maintain those.
VOCALS: Waves Renaissance EQ, Waves Renaissance Compressor, Waves C4 Multi-Band Compressor, Waves L1+ Maximizer, Waves Renaissance Reverb, Waves MetaFlanger, Waves Doubler, Waves SuperTap 6, Antaras AutoTune, Digidesign Dynamics III.
The song consists of a few instrumental tracks, but a lot of vocal tracks (harmonies, add libs).
The vocals were my main focus when I started mixing. The harmonies existed of Keisha White’s own harmonies and some male harmonies to give them more range and power. The male harmonies were provided by one of the singers from the new R&B male group Fundamental’03 (Geffen Records).
I splitted nearly every single track in the session at least once, but in most of the cases twice – the so called parallel processing chain. On the vocal group I used the C4 Multiband Compressor to keep certain areas in control. As the C4 Multiband Compressor doesn’t sound much like a compressor, Renaissance Compressor followed to emulate an old classic LA2A before hitting the L1+ Maximizer brick wall limiter to flatten out the vocal group completely, yet leave some dynamics. On the EQ part I added a few dB at 1,300 cycles and 10,000 cycles to open up the vocals a bit, and made a few cuts at 247 cycles and 2,200 cycles for a more transparent sound. All the vocal parts got threaded differently to give them all their unique signature and position in the mix.
For effects on the vocals I used Renaissance Reverb to give a nice background blend to them. MetaFlanger was inserted to make them sound less static and a bit moving around the centre image like more of a chorus effect with the tape option activated to make them sit more solid in the mix. Doubler 4 functioned as a pitch shifter/ stereo widener to make the vocals sound larger than life and the singer like a super star. SuperTap 6 was used three times in the mix, one was used to give the vocals more depth and blend, and a second one was used (offline) to emphasize some key words in the lyrics phrases as quarter note delays (ducked). I ducked the vocal effects to achieve a clear sound, as delays and reverbs tend to mash out the vocals sound when it becomes almost inaudible. The trick is to place the compressor after the delay or reverb. You set-up another stereo Aux Input track and send the dry vocal group track to the Aux Input track, and do another send from the dry vocal group track to the compressor’s external side-chain input. It’s a great trick but take in account it will only work on dedicated individual effects rather than global effects.
An extra Aux Input track (pre-fader) was created to act as some sort of “exciter” effect. I heavy compressed the incoming (dry) vocal group signal (-30 dBFS), followed by Renaissance EQ (post compressor the EQ to avoid some low end rumble coming through when its filtered off) with the high pass shelf set to around 770 cycles and large boosts at 5,000 cycles, 7,800 cycles and 10,000 cycles. This trick is also known as MoTown EQ (not the actual device!), and its descent and discovery comes from the legendary 1960’s MoTown era. SuperTap 6 acted as the vocal “exciter” with extremely short delay times set over the entire stereo spectrum of the mix. Its phasing effect complements the vocals.
DRUMS: Waves SSL G-Master Buss Compressor, Waves SSL E Channel, Waves SSL G-EQ, Waves TransX Multi, Waves TransX Wide, Waves C1 Compressor, Waves Renaissance EQ, Waves Renaissance Reverb, Waves C1-sc Sidechain Compressor/ Gate, Waves Renaissance Bass, Waves MaxxBass, DigiDesign Sound Generator, Digidesign Dynamics III, Digidesign PitchShifter.
Before I started to clean and polish the drums I set up a few Aux Input tracks to act as instrument Group Tracks. On all of the Group Tracks I inserted the SSL G-Master Buss Compressor and TransX Wide plugins to act as overall compressor and transient emphasizer/ gate. I mixed the drums through a main Aux Input track acting as the drums main Group Track with, yet again a SSL G-Master Buss Compressor inserted to blend them together and give them that nice punch of the classic analogue SSL consoles.
All the drum sounds came of the classic Roland TR-808 drum machine. I splitted nearly all of them twice, with the first one EQ’d only – acting as the ORIGINAL fader, the second one heavy compressed (-12 dBFS) and EQ’d – acting as the MUD fader, and a third and final one extreme compressed (-30 dBFS) and a high pass shelf EQ with a lot of high end acting as the ATTACK fader. Blend together they formed the sound of the drums. To emphasize the transients on all of the tracks I inserted TransX Multi at the end of the processing chain to attenuate the low and high end content of the instruments timbre. On its Group Tracks I inserted TransX Wide to act as overall sound shaper and gate. The wonderful thing about the TransX plugins is that they don’t only design transients but also add harmonics to the instrumental content and on the TransX Multi you can do it frequency selective. On the original kick I inserted MaxxBass to make the kick audible on the smallest and most common multi-media speakers (TV, Radio) available on the consumer market, as the 808 kick only exist out of a pure sine wave, I created a fourth mono audio track and added ProTools own “signal generator” (offline) running through the entire length of the song (50 cycles sine wave, like an oscillator on the drop frame of old analogue consoles) to reinforce the kick drums low end content. ProTools own Gate/ Expender III acted as gate, and kick ORIGINAL as trigger (pre-fader). I inserted Renaissance Bass on this signal to add additional harmonics and more power to the low end regions. I dipped around 300 cycles to give space for the keyboard sounds and 600 cycles to lose some of the kick drums energy so it wouldn’t overpower other instruments in the mix. Boosts were set at 64 cycles, 130 cycles and 3,500 cycles and 4,000 cycles so the ear can easily find the kick. I didn’t split the OH instruments.
During the blending of the drums I noticed that the snares/ claps overpowered the drums, so I added C1 Side-Chain Compressor/ Gate with very fast attack and release times on both of the Group Tracks. The snares and claps wound up to 500 cycles to give the kick more ..breathing.. space and the same for the OH group. I put C1 Side-Chain Compressor/ Gate with a large dip at 700 cycles to give the snare and clap tracks more room to ‘breath’. I added a nice Renaissance Reverb on the snares and claps to bring them more to the back of the mix, just behind the vocals. An additional Renaissance Reverb was used lightly on the kick Group Track to make the feel of the song more big and mystic, yet keep the kick clear enough to be upfront and hitting your stomach and chest.
I felt that the snares/ claps sounded a bit weak so I created 2 additional mono audio tracks and placed a copy of the snare on one, and a copy of the clap on the other. ProTools own PitchShifter plugin (offline) was used to de-tune the channels information with about 7 cents to give the overall blended snares and claps more body and strength. This trick basically works on every transient sound that sounds a bit to ‘thin’.
BASS: Waves S1 Imager, Waves Renaissance EQ, Waves C4 Multiband Compressor, Waves MaxxBass, Waves Renaissance Bass, Sans Amp, Digidesign Dynamics III.
The bass of the track carries a lot of low and low-mid energy and lies directly below the kick. Another problem was the phasing of the bass. I used S1 Imager (offline) to make the bass mono. I needed to find place for both, the kick and bass, within the mix without fighting each other for the same space; therefore the bassline got placed below the kick drum. The bass got splitted once, and I used the original fader to EQ the bass only, with a low pass shelf around 3,000 cycles and some boosts at 72 cycles, 980 cycles and cuts at 64 cycles to give the kick some space, and 734 cycles to give the snares and claps some space. The splitted channel acted as the MUD channel similar to what I did on the drums to contain more control and phatness.
On the Bass Group track I added C4 Multiband Compressor to control its low end so it wouldn’t jump in the song. To distinguish the bassline from the kick drum I inserted Sans Amp for some distortion on the same Group Track. I tweaked it until I got the desired effect I was looking for.
On the original EQ’d only fader I inserted MaxxBass to add some harmonics and make the bassline audible on the small multi-media speakers of consumer audio products. On the MUD fader I inserted Renaissance Bass for additional punch, low end content and harmonics.
ProTools own Compressor/ Limiter III acted as ducking device on the Group Track to push the kick through whenever it kicked in during the song to avoid the kick getting eaten by the bass. The ORIGINAL kick drum fader acts as the trigger (pre-fader). I put the bass group on mute when I continued mixing the song. The reason why? Cause the bass contains so much energy it can literally screw up your mix. I blend the bass back in as the final instrument of the song and tweaked it further to make it fit without interfering with the instruments.
KEYBOARDS: Waves Renaissance EQ, Waves Doubler, Waves MetaFlanger, Waves Enigma, Waves MondoMod.
The only tracks left are keyboard based sounds and didn’t need much treatment. EQ served to emphasize the instruments timbre and a boost around 14,000 cycles to gain some air and make them sound less digital and more analogue like the old classic analogue synths from the 70’s and 80’s.
The groups’ chain of effects made the keyboard sounds really stand out. I used Metaflanger to emulate an old analogue Ampex 440 tape machine. Enigma acted as a nice flanging effect and MondoMode made the entire instrument group move around as some sort of a chorus effect. Doubler 2 functioned as pitch shifting and widening tool to give more depth and spectral placement of the keyboard sounds.
FX SOUND: Waves MaxxBass, Waves Renaissance Reverb.
The entire song only carried one single effect sound, a closing door. Because it was a great picked and clean sample I did not use any EQ or compression. I wanted to make the sample sound more cinematic with some phatt low end. MaxxBass with loads of low end and harmonics created that sound I was looking for.
Renaissance Reverb inserted directly onto the channel to bring the sample to the background as if the door is closing behind you.
MASTER BUSS: Waves SSL G-Masterbuss Compressor, Waves S1 Shuffler, DuY DAD Tape, Focusrite D2 EQ, Waves L3 MultiMaximizer.
The masterbuss contains the SSL G-Master Buss Compressor to gain additional punch and glue the mix together with gentle and subtle gain reduction. S1 Shuffler was used to give the mix a nice wide 3D radio ready sound.
When widening the stereo image the human brain perceives a loss in mid-high and low end frequencies. I made up for this by adding some dB at 1,088 cycles – the in your face frequency, and a 1 dB bass boost. DuY’s DAD Tape functioned as _” analogue tape emulator at 30 IPS, I added a D2 EQ from Focusrite to shelf about 1,5 dB high end back into the mix to make up for the high end loss during the “transfer” to analogue tape.
Lastly and finally I added L3 MultiMaximizer with the output set to -5 dBFS to gain some headroom for the mastering engineer before we went on printing the four common versions; main mix, instrumental, tv version and the acappella version of the mix.
SIDELINE: Additional Tricks.
Lead Vocals:Split Reverb technique is one of my favourite techniques to apply on lead vocals, to give them more depth and more of a stereo effect.
Create two Aux Input tracks and load the same reverb plugin and the same settings on both tracks. Pan them hard left and hard right. Send the lead vocal to both tracks and adjust the decay time of the reverb on one of them. Tweak until you got a desired ‘stereo’ effect.
On automation: To give your song more of a ‘live band’ feel and thus more emotion, ride the kick drum group fader during the song a little by changing the level a bit ever 2 to 4 bars. On the chorus put more of the attack fader in (about 1 dB), so it sounds like the drummer is hitting harder during the most emotional part of the song. Put the snares group fader up randomly throughout the song.
On the final chorus try to insert another EQ (1 band) on the vocal group track. Ride the MHF range with a little boost during that section of the song to really make the vocals ‘scream’.
Sample Replacement: Sample replacement is usual used as a tool for mixing engineers to replace the performance of a bad drummer. I usually use this trick to ‘enhance’ sampled drum patterns, therefor I rather call it ‘sample enhancement’ then replacement.
You can easily enhance drum samples by ProTools own Beat Detective plugin. I won’t dig deep into it but for those who are interested drop me a line at my email given before.
Another trick to reinforce TR-808 snares instead of de-tuning the parallel sound source is to enhance it with a short sampled TR-909 snare, which has more body. Blend the TR-909 snare in BUT keep it low so it won’t mess to much with the original TR-808 snare sound. You can do the same trick for the kick drum(s). Put a thumb -low end kick drum sample (needs to have information below 40 Hz), that suites the sound of the other kick drum(s), on a new channel and blend it in until it sounds phat.