Firefly Secret Mixter Remixes!

How I Did It

by Quarkstar
Recommends (22)
Sun, Mar 23, 2014 @ 12:14 AM

Tools I Used

Drum Kit1: Ableton Drum Rack - By Quarkstar
Drum Kit2: Ableton Designer Drums - Nervous (clip Nervous A)
Drum Kit3: Soniccouture Konkrete Breaks - (clip Chiq)
Main Bass: Dune - Chorus Bass RL
Sub-bass: Ableton Analog - By Quarkstar
Icelandic Pads: Dune - Icelandic SynPad
Ice Bells: Alchemy - Metal Plucks
Clavinet: Native Instruments Razor - Space Clavi
Koto: Synapse Dune - Koto
Beeper: Ableton Analog - By Quarkstar
Shy Trumpet: Ableton Analog - By Quarkstar

Reverb - Airwindows Space
Console: Airwindows Console (Analogue mix desk))
Channel Effect: Sonimus Statson Channel (Models channel distortion)
Distance effect: TDR Labs Proximity (Distance emulation)
Instrument reverb: Airwindows Pocket Reverb (Spot effect on instruments)
Master Limiter: Airwindows ADClip4
Master frequency analyser: Voxengo Span

Samples I Used

Kick: Thomas Penton - kick91 (attack is reduced)
Hihat: Thomas Penton - hihat-606-closed


Tune structure
I loved Beluga’s Emergence of Chance tune, but wanted to change the subtle electro edge into something different. This meant changing the sounds. I deconstructed all of Belugas samples, turning them into midi using track to midi converters to extract the drums and beeper. I was happy with the original 120 BPM. I followed Beluga’s structure of the piece, intro of 4 bars, verse 32 bars, breakdown 32 bars, verse 32 bars, end.

I created an outline composition from the drums, bass, beeper and piano lead. These drums included the Nervous Drums and Chiq drums from the Ableton library.

I wondered why the little beep noise was in-tune with his piano, but mine wasn’t although I was playing the same notes. Playing the two beeps side by side I found that Beluga’s beeps were micro-tuned +27.3 cents and that was bringing them into tune with everything else. Once I retuned my beeps, they all fitted perfectly. Turns out the 27.3 cents was tuning the notes to the Pythagorean 7th (thanks Wikipedia) so it was intentional and I assumed, showed a deep knowledge of music theory by Beluga.
The beeps are subtly swung 1/32 off the beat. There are several other places where the notes are swung off the beats.

Chords, tempo and beats
Looking at the notes in the Piano Solo by Beluga [A, C, D, E, G, B] the Key is C major. Together with the Bass, the chords seem to be Am7, D7sus4, Am and F7. Tempo is 120BPM, time signature is 8/8 and the stress is on beats 1 and 5 for most bars, a few are on 1, 4 and 6. There are also a lot of 1/16 note timings in the bass and lead.

I didn’t bother much with that information as I took all the instrument parts from Beluga’s composition, simply removing notes or adding notes that sounded right. If I were writing new parts, I would certainly be looking at that information.

Choosing vocals
The Tempo is 120BP and the Key is C major; there quite a few vocals in that key to choose from in ccmixter, except that beats 1 and 5 are stressed. Its unlikely that you will find vocals with beats 1 and 5 but you will find many with a beats on 1 and 4. It would be possible to warp the 4th beat onto the 5th, just by dragging the warp markers, but the notes are also on 1/16 and the warping is going to sound odd.

The easy way is to use a spoken word, or to chop up lyrics and use only a few repeated phrases, Beluga took the latter approach. When choosing lyrics or spoken word, I try to pick something that I can feel or sympathise with.

Once I had the major pieces in place, I searched for alternate vocals, ones that meant something personal to me. I found snowflakes spoken word "May we join hands” and edited it to be more upbeat. I do feel guilty when I change the meaning that a songwriter has given them.

Having chosen spoken word I cut the vocals into sections and placed them into bars and sections. At this point the spoken word is not fitted to the tune and its very important that the words are at their normal speed so you can hear the phrasing, i.e. no warping.

Composing for the vocals
At this stage I had a tune outline with Drums, Vocals, Bass, beeper and lead Piano. I started sorting out the sounds of each part to match how the vocals “felt” to me. Then I cut up the vocals, fitted them to bars and beats and tided up using little warping as I could, preferring to chop samples where I could. For spoken word, its better to cut the vocals up and keep the natural speed and intonation.

I listened many times till I started to hear the accompanying melodies and sounds. Then I began finalising the instrument sounds and got the vocals sounding right.

I started adding the extra melodies and parts to help the tune along and to fit the vocals. For example snowflake pleads with us too join hands and under that I wrote an uplifting melody.

I did move some vocals around to get a better flow and contrast.

The words set the tone and I started changing the sounds and instruments to something more tribal, old, ancient and evolving. To help with that visualisation, I gave the instruments meanings; the bass was earth, the refrain spirit, there was water, ice, rain and air in there too and then wrote the melodies and sounds to fit together into a soundscape with snowflakes vocals.

To fill out the sound after the first verse, I doubled up the bass in places, adding sub-bass, Icelandic SynPad and Metal Plucks playing the same bass line.

For the extra instrumental parts I took Beluga’s piano solo and, using the same notes, wrote the Shy Trumpet part by removing notes from the solo. The first Koto solo was written by taking the beats from the Nervous Drums and then turning those into notes. The second Koto solo copies the Shy Trumpet part and plays improvisations over the top. The piano was changed to a Clavinet, which gives a shorter, richer in high harmonics percussive sound.

There are few melodies in the composition and several instruments play the same melody or variations at the same time. The four melodies are by instrument, the Beeper, Bass line, Beluga’s Piano melody and the Koto playing the drum melody.

Sound engineering
The various instrument sounds complement each other and are chosen not to conflict with each others frequency bands. Examining the mix…

The Bass is the main part of the mix and goes from below 30Hz to 10KHz so it pretty much occupies all of the low and mid frequencies. The Bass has to be heard despite other instruments playing over the top so all the instruments playing in the mid frequencies are short duration and percussive. This leaves the tail of the Bass to be heard after the other instruments have decayed away. Where I double up the Bass part I use Bells and Ice pads, both of which are rich in high frequencies and can be played over the bass.

I was careful to keep the sounds separate in space. Instruments playing in the same frequency ranges are panned left and right away from each other. To further separate the instruments in space I used a plugin called Proximity by TDR Labs, that emulates the effect of distance. This is used on the Icelandic Pads, IceBells, Koto, Clavinet and Shy Trumpet. It adds, a filter that emulates the fall off of frequencies with distance, delay, reverb, sound spread, etc.

Vocal treatment
There was background noise in the vocals which I removed with a gate. This also removed snowflakes intake of breath which I replaced with samples from another song. Its obvious when breath intakes are missing, but you are so used to hearing them that you don’t notice when they are there.

Each clip has clip gain automation (done by hand) to keep the same volume, followed by a De-Esser, a Gate to remove noise, Saturator, Compressor (New York Style), Chorus and Equalisation. Echo is done on the return track.
I like a little bit of saturation or valve distortion on some vocals, to my ears it helps the vocals sit in the mix and warms them up.

I repeated the phrase “The world is changing” eight times but changed the phrasing of each one by warping the timing and altering the pitch so they sound different.

To help stop muddiness in the bass, the Bass line is side-chained from the drums and uses a compressor to duck the bass out the way when the kick drums hit.

I used a shelf equaliser to reduce the bass frequencies between 200 and 2,000 Hz so that snowflakes voice would shine through. I also added some distortion (saturation) to snowflakes voice so that it was richer in high frequencies and would sit above the bass and mids.

To get a stereo spread from a mono signal, add a Delay with 1ms on the left and 5 to 12ms on the right, then follow that with width reduction 80% to reduce anti-phase.

I put reverb under the hi-hats but only when they are playing. Leaving a long reverb tail on hi-hats makes them lose impact in the mix and adds unwanted noise and confused sound. I use gated reverb which follows the envelope of the hi-hats and serves to thicken up the hi-hat sound. The gating only leaves a short reverb tail so the sound is still tight. At one time I used my own effects rack for this, but found that Airwindows PocketReverb does a better job.

I add a little Saturation to my kick drums, it increases the mid and high harmonics, leaving the bass untouched. The new higher frequencies are loud and would muddy up the mid frequencies so I add an Equaliser and shelved the sound above 200Hz about -3dB.

Master channel has an Airwindows Buss, equaliser, a compressor, gain and a limiter, with Voxengo Span frequency analyser. I used the Ableton equaliser, the Ableton Mix Gel Compressor and an Airwindows limiter.

Each channel has an Equaliser, Sonimus Statson analogue desk emulator, Voxengo Span frequency analyser and an Airwindows Console. Each channel is set to output -12dB at 0dB slider setting, so the Voxengo Span is set to K-12C metering on each channel.

So what do the Airwindows Console and Buss do? Basically sound magic, they improve the 3D Sound image and the difference with and without is significant. Tom Middleton mentioned Airwindows console in an interview and he said he uses it in all his mixes. I describe it as taking all the instruments and eroding away the gaps between them so each stands out. The overall sound is slightly thinner but each instrument stands out and is clearer in the mix.

I used a clean reverb, Airwaves Space. It is set to reduce the bass so it doesn’t rumble and the high frequencies are reduced so it doesn’t hiss or catch sibilance in the vocals. I have not tried other reverbs in this mix, I sometimes use a convolution reverb. Its quite possible another type of reverb would sound better, but I stayed with Space because it did the job and I was happy with the sound.

Testing in Mono
You should check the track in mono to make sure that there are no anti-phase sound between the two channels. If you switch to mono and some instruments become thin, the volume goes down or instruments disappear then there is a problem.

OK you got me here, I haven’t fixed the phase problems in the bass and vocals, but it sounds so good. I mix with Headphones so I don’t hear phase problems unless I switch to mono.

There are a few places where the mix goes from mono to stereo or back, I use this to indicate section changes.

Sub-bass is added at section transitions for extra depth.

At the end there is some delay on snowflakes voice, it fades so the tune does not end abruptly.

White noise wipes your musical memory, I used the sound of a rain storm.

Gated verb on some high pitched sounds to add depth (described elsewhere).

Automated filters to bring in the Clavinet and fade it out again.

Distortion on the voice so it cuts through the mix without increasing the volume.

Other Notes

Transitions and Transformations
I always love the Secret Mixters and themed remixes. I learn from them and they stretch my abilities. I learnt a lot from Beluga especially unusual beat timing and micro tuning, thanks for the great tunes and samples to pick from.