My buddy Will has published again, and this time it’s his most ambitious song yet. It’s a 10+ minute prog rock epic with changing time signatures, face-blistering guitar solos, and a storyline that reaches into the realms of dreams and prophecies. I’m not ready to call it his Bohemian Rhapsody or his Good Vibrations because I think he still has plenty more in him where this came from.
It all started for me a bit over a year ago, shortly after he published Friday Night, we were talking about his next project. He told me he was ready to tackle a song that he had been thinking about for years. It was based on a prophetic dream he had received when he was a young Christian nearly 20 years before. He covers the story very well on his song page, so I won’t rehash it here.
How it concerns me is that he wanted a prominent and significant piano part, which he wanted me to write. He knew what he wanted for the intro, and had concepts for the rest of the piece, but wanted me to realize them.
About the Song
Alt Mix A
Alt Mix B
When the song was completely written, we sat down with the lyrics and he mapped out his plan. The song clocks in at just over 10 minutes, with a good chunk of that in various instrumental sections. The song has a repeating chorus, but the “verses” between the choruses are completely different (save the first and last). The chorus itself has the same melody but different lyrics each time except for the tag line.
The song begins with an extended introduction based on a simple 4-chord vamp, with all the of various instruments making an appearance and contributions as it builds to a climax a minute and a half later. The first verse begins at two minutes and is a simple four-line affair followed by the first chorus.
The second section moves in to 3/4 time, the first a several time signature changes. This section builds to a driving repeat of the chorus. Now over five minutes into the song, Will introduces the third section of new material. This section again builds into a energetic, distorted-guitar-driven exposè of the darkest hours of his life (the whole song is auto-biographical).
When this section wraps up at 6:50, the fourth and final section of new material is introduced. This is a call and response section with a female vocalist that contrasts the outside with the inside of his life. The whole is written brilliantly – it carries immense meaning without ever hitting the listener over the head with the lyrics. Careful listeners will notice references to many of the songs he has written in the last several years.
At the conclusion of this section is a repeat of the original verse and chorus (again with different lyrics – actually, the verse even has a different melody, too) climaxing in a wall of sound before the final, exhausted, resolution. Total time: 10 minutes, 15 seconds – and what a ride it has been.
Will tracks pretty much everything in his basement studio. I wrote the piano and synth parts at home on my digital piano, exporting the parts via internet so he could hear and critique. It also allowed him to write the guitar and bass parts such that everything worked together. Then, when I hear the guitar and bass, I tweaked my piano parts to fit in. As we talked about Will’s desire to produce a tracking video similar to the one he did for Friday Night, we looked at how we wanted to record the piano – both audio and video. I said I really wanted to figure out how to actually play a real, acoustic, grand piano for the video.
In the end, we found a friendly church and used their second sanctuary all morning on a Tuesday to record 4 takes of audio and video. Will brought a portable recording system and friend Mike Kerby recorded the video. This coincided with my last day in Peoria – in fact, after lunch, we packed up the Chrysler and drove to my folk’s house in Champaign, leaving the next morning for Virginia.
While Will’s mix is the definitive version of this song (and the one featured in his video), I was itching to take a shot at a mix, as well. Once he had tracked all his parts, he sent me a thumb drive with all the files (over 3-gig worth). The first thing I did was strip out all the plug-ins and volume automation so I could start from scratch (I left in the guitar effects because that was an integral part of the sound).
It is a good idea to start a mix like this with the largest, most complex, loudest part, get good, full levels, and then work backwards. In this case it worked out well to start with the end of the introduction, which is pretty much as loud and full as the end of the song (and being separated by over 8 minutes, the listener won’t really have a reference between them anyway to say one is louder than the other).
A quick tour of the screen-shot: the top track is Will’s lead vocals. The second track are Jessamyn’s backing vocals. Next is the acoustic guitar, followed by the bass. The two purple tracks are the double-tracked electric guitars. They are small because I did all the volume automation on them in an aux bus which is at the bottom of the screen. The dark orange track and the empty track below are the drums – I never rendered them from MIDI because I found I needed to make small tweaks here and there to the timing. The two cyan tracks below that are the rendered synth parts – MiniMoog lead on top and Mellotron strings below. The final track is the piano aux. We used three mics on the piano and supplemented certain parts with a piano plug-in to beef up some specific notes. Once I balanced and panned them all, I subbed them to this bus and did the level automation there (that’s also where all the plug-ins are located).
One of the reasons that I wanted to do a mix of this song, aside from the fact that it’s a killer song, is that I wanted a chance to play with some of the concepts and techniques that I’m always reading about in Mix Rescue and various blogs. There are also matters of taste – in this case the amount of reverb on lead vocals in a contemporary pop/rock song. To whit, the first thing Will mentioned when he hear my remix was “The lead vocal is really dry!”
My response – “I know.” I like my lead rock vocals pretty dry. Now, if you were to solo the lead vocal and turn the reverb on an off it I think it would be blindingly obvious, but in the context of the mix the vocals do stand out fairly well. There’s more about what I actually used for reverbs below under the reverb heading. There are also sends to the delays (1/4-note and 1/8-note) that are used during certain portions of the song.
The effects chain is fairly straightforward. This is actually my third iteration of effects – the first two I just went crazy with plugins and did more damage that good stuff. The first rule of mixing is “do no harm”. The chain simply consists of a cutting EQ, compressor, boosting EQ and saturation. The cutting EQ is the stock Sonitus EQ and it is removing all the really low frequencies that are not part of the vocal sound, but can take up sonic space in the mix.
The boosting EQ is the Bootsie plug, the BootEQ Mk II. This is a simple little parametric EQ (variable frequency bands) that is usually my first go-to equalizer for boosting (I use the built-in Sonitus Equalizer for cutting). It also has a pre-amp stage for a little more harmonic saturation. I’ve learned that saturation is best applied in numerous subtle layers for best effect.
The compressor is TL’s Leveling Amplifier. This analog-style limiter is to make sure that I don’t push the levels too much and control the gain I added in the previous steps. This one comes very highly recommended, too.
I took a slightly different approach with Jessamyn Luong’s backing vocals. This is her second appearance in a Willsong production – the first being as a tambourine player in Friday Night.
Jessie’s vocal chain has many of the same elements as Will’s, with some important differences.
First in the chain is compress/saturation, provided courtesy of Bootsie’s brand new Thrillseeker LA. Bootsie is experimenting with Stateful Saturation, and is integrating it into all his new plug-ins. This non-linear saturation seems to be more pleasing than the static, linear saturation found in most plug-in emulations. So far I like it, too.
Next in the chain is the PushTec EQ, gently boosting the highs. Last in the chain, I inserted DeSample’s Glaceverb plug-in. This is a good room simulator (although capable of much more creative effects) and I use it to create some space around Jesse’s voice. I like this plug for this purpose because with just a few controls I can adjust the size and qualities of the room.
The next track is Will’s Taylor 414. The second thing he asked when he heard my mix was “how did you get that tone from the acoustic?” A large part of that answer is the Plektron WTC Comp. While it is advertised as a compressor for the two-buss (whole mix), the name Plektron made me think of guitars. Wow, it makes a great sound out of the guitar. It only has a couple of controls and works everything else internally. Can’t recommend it highly enough.
Before the compressor in the chain is the Sonitus Equalizer, cutting the low frequencies out (don’t need those subsonic rumbles muddying everything up) and gently boosting the highs. This boost is another piece of the amazing guitar sound I found.
At the end of the chain is my constant friend, saturation. Here it is courtesy of Bootsie Ferric TDS. It adds a little more gentle compression while filling out the harmonic spectrum. I think it turned out well.
The bass is one of the simplest tracks, processing-wise. This is because it is really well recorded. All I did was roll off the lowest frequencies, then inserted a multi-band compressor. This is a unit that splits the incoming signal into 5 different frequency bands, and then applies compression separately to each band. Bass seems to respond especially well to this sort of processing.
The final plug-in is the BookEQ Mk II, rolling off whatever is left of the subbass so it doesn’t suck up my energy, and a light boost to the low mids. Then this provides some nice saturation via the tube amp stage.
Here’s a secret – if you want to hear one of Will’s significant inspirations for bass, check this out: http://www.youtube.com/watch?v=BsRdT9hwqGs…
While the bass was pretty easy to mix, the electric guitars practically mixed themselves. This is not so much because they were so well recorded (which they are), but because the part is so well arranged. When they should be in the forefront, they are – level boost not needed. When they need to be in the background, they go play nicely in the corner. Let this be a lesson to aspiring song writers, producers and band members. I cannot emphasize enough – LET THIS BE A LESSON!
So, what did I do? Not much. The delays shown in the screen shot are Will’s originals. I didn’t change them since the delay is an integral part of the sound of the guitars. I also used his Marsh Spring guitar amp impulse response for the same reason. I’m not a guitar player, and every guitar player I know is insanely obsessed with tone! So, I didn’t mess with it. Anyway, I liked it.
I did play with a little level automation in an arranging sense. I have one more part in my mix than he does (Minimoog synth lead), so I duck the guitars occasionally to let the two play off each other.
The last track recorded by Will is the drum track. He uses Native Instruments Battery for the sounds, and plays the part in on a keyboard. He uses a highly customized drum kit, and he’s one of the best keyboard drummers I know of – gets really good tracks. He also worked with a friend who is a professional percussionist who gave him some tips.
The chain is fairly simple, consisting of two equalizers, a compressor and a compressor/ saturation stage. The Sonitus Equalizer is at the bottom of the screen shot and the curve shows that it is in it’s usual role of only cutting frequencies. There’s a roll-off and then a dip at 500 Hz. This is a tip I got from Graham Cochrane and it helps clean up even a sampled drum sound like this.
Following the EQ is Bootsie’s TesslaPro. This fantastic little plugin provides smooth saturation that responds to transients (the attack of the sound) and at the same time provides analog-style compression and warmth. I love what this plugin does on drums and other percussive sources.
The final compressor is a stock Sonitus set to gently boost the entire signal between the EQ and the saturation. I had some trouble with getting the drums loud enough to be heard in the mix. I could have inserted a gain plug-in that just adds level, but this way I can add more without blasting through the volume ceiling on loud transients. Then the TesslaPro works its magic on this transient-rich material.
Now it’s my turn. I added two synth parts to the mix. The first was a Minimoog clone, the Minimogue. I’ve been using this free synth for years and its one of my favorites. I did run into one limitation this time, though, that I hadn’t before – the LFO (for vibrato) has only a few predefined speeds. I really wanted a smoothly variable rate from zero on up to “unusable warble”. Despite the limitations, I really like how the lines turned out.
At one point in the production, Will commented that he thought the arrangement could use strings. I said it needed the gritty, bitey strings from a Mellotron. This wonderful device from the 60’s used audio tape recordings of strings and other sounds – one per note – and played them back, making it the first sampler keyboard! There are a number of free and pay computer versions of the Mellotron with varying levels of control and customization. The one I tend to use the most is the freeware Tapeworm, which uses digital records of an actual Mellotron. This one doesn’t have the additional controls that the more expensive versions have, but it does have one thing the original Mellotron didn’t – envelope controls. This allowed me to extend the release (how the note dies away after the key is lifted) to match the song better.
As mentioned earlier, we used multiple mics to record the piano. We used my ribbon mic over the low strings, Will’s large-diaphragm condenser over the high strings, and a small-diaphragm condenser on a high stand about 10 feet back from the open lid. I balanced the levels between the three microphones, panned them and used eq to cut out frequencies from each mic that I didn’t need.
There were specific notes within the piano arrangement that Will wanted emphasized as an effect. We couldn’t bring them out strongly enough in the audio recording, so Will programmed a MIDI track with just the notes that he wanted to turn up. In my version those are rendered with Garritan’s Authorized Steinway.
Everything was mixed down to a stereo bus and I applied level automation to that so I could mix the piano into the song. All of the effects were also added to the bus, rather than the individual tracks. The compression and eq settings were based on a tip from Bobby Owsinski’s Mixing Engineer’s Handbook. In one of the interviews the engineer for Elton John’s Honky Chateau revealed the recipe he used for Elton’s shimmering piano that sat so nicely in and on a mix. I used the Antress Modern Painkiller to emulate the LA-2A compressor, and dialed in the eq setting on the ubiquitous BootEQ Mk II.
Some ambiance was added via PSP Pianoverb – a specialized reverb that is tuned to the notes of the scale, making it ideally suited for keyboard instruments. Saturation is provided with CDSoundmaster tape emulator.
Let’s finish this article off with the reverb plugins that I used. Each reverb was on a separate aux track, with individual sends from each channel. This way I can balance the reverbs to push various elements further back into the mix. I even automated the send from the lead vocals, and there are three places where I add a lot of reverb to make the sound either pushed back or drift back during a sustained note.
The ambiance is the short, dense reverb and is supplied by the Acustica Nebula plug-in, running an Old Plate reverb patch. This plugin was actually a last resort when I couldn’t get either of my other two IR reverbs to work, and I liked it enough that I kept it in even when I got the other two working.
The other reverb with the long tail is courtesy the Kjaerhus Audio Classic Reverb. This is generally my first call algorithmic reverb, although if I hadn’t already dialed in the sound I wanted, I would have tried out the Epicverb by Bootsy. I keep meaning to use it but the KA works so well I never think of it.
In the end I’m pretty happy with my mix. I could continue tweaking and tweaking (the posted mix is number 11 – no, 12, see note below), but I’ve learned that a mix, like all art, is never completed, just abandoned. It ebbs and flows and breathes like I think it should, and the parts play nice together. I can’t wait to hear what Will has in store next, and I’m looking forward to working on some of my own stuff, as well.