Digital Album: Let's Build Mecha (Remastered)

 

Let’s Build Mecha (Remastered)

2020 marked the 10th anniversary of Let’s Build Mecha! so it seemed like a good excuse to revisit it. I’ve always been happy with the sound quality, and I found the mix reasonable considering my lack of studio headphones or near-field monitors when I made it in 2010. At that time I was using a Logitech 5.1 system with a very clean – but thin – sound, and some Sony earbuds. It was also early days for mastering on my part. I believe I’d discovered the wavehammer compressor in Sound Forge by then, but none of the multi-tracks were summed to a master bus, and the title track was mixed-down with hard clipping instead of compression.

In 2012 I made some alterations – Clockwork Robots was made to run seamlessly into Autonomic, for example – but no real ‘remastering’ was done. The biggest hurdle, ten years later, was accessing the original multi-track files. They were recorded in Sonar 2 and 4 and saved as bundle files: single files containing all the data. Newer versions of Sonar don't play well with bundle files and prefer a project file pointing to a folder containing all the stems. A good comparison would be a zip file vs a folder. So I had to open the files on my old Windows XP machine, ensure all the samples and plug-ins were present, and export them.

A couple of songs needed no tinkering whatsoever, beyond clean mix-downs on a superior DAC and some mastering compression. Others required more care. The e.p. has always sounded clean and crisp to me, but I had to wonder why I'd chosen to EQ so many bass frequencies out. I’ll put it down to inexperience.

The songs that required the most attention were the title track - the most layered song on the album - and the other full vocal track: Production Line. The latter has always fallen a little flat, so I wanted to address that. Another challenging piece was Rust and Bones. The bass guitar always had a ground hum that stuck out like a sore thumb. Having some new tricks up my sleeve, I was able to address it. Another issue with that piece was the sampler I was using at the time, a clunky old horse called VSampler, which I could not resurrect. Each instance had to be replaced with Grace, my current sampler of choice, and the attack/release settings carefully tweaked by ear to match the original.

In most cases, the generic Cakewalk EQ plug-in was replaced with the superior Sonitus Equalizer. Heretofore unnoticed digital hiss on certain samples was carefully tamed. Vocals were brightened, and suppressed bass frequencies were restored. One tiny edit was made on the vocals of Production Line, where a slightly off-key word was replaced by an alternate take from the original session. The final step was to check the dynamic range of each track and adjust as necessary.

 

A little history. In 2010 my good friend Stevie K. Farnaby was producing a Doctor Who audio drama for Brokensea.org. He had an upcoming episode called Mechalution, which featured a race of sentient machines. He’d heard the music I’d been making with circuit-bent Speak and Spells, etc... and thought the sounds would be perfect for the show: both as music and sound effects. It just so happened that I had a song called Let’s Build Mecha, inspired by Japanese anime, and it fit the concept perfectly. With the script for Mechalution to hand, I set about writing the rest. 

 I don’t recall any difficulty or writer’s block. The subject matter was so in tune with my aesthetic that I could easily have carried on. In fact the Mechalution storyline was intended to carry on for the rest of the season, but for various reasons Stevie had to halt production. I wrote four more pieces for subsequent episodes (three of which can be found on The Mechanicals e.p.) and two more demos for what would've been the season finale. As of this writing, Stevie is upgrading his studio: so there’s a very real chance there’ll be more Mecha-inspired music forthcoming.

To coincide with the remaster, I've released a promo video for Production Line that was filmed in 2014 but never completed until now.

Cover song: My Brother's Time

I finished something? Wonders will never cease. This project was started January 31st and completed May 30th, but it's not like I worked on it full-time over the four-month period. It was done in fits and starts and sometimes languished for weeks at a time. But I'm very pleased with the end result. Here's a behind-the-scenes look at what went into it.

When I learned there was to be a fan tribute to Gary Numan's 1981 album 'Dance,' compiled for the website Numanme.co.uk, I thought it would be great to participate. The entire album, including b-sides and demos, would be recreated by fans from around the globe. I decided early on that the song 'My Brother's Time' inspired me the most to create a cover version (although my favourite from the album is actually 'A Subway Called You').

For those of you who haven't heard it, here's Gary Numan's original:




And here's my version, The Manitou featuring Jimmy Aaron:


My Brother's Time by themanitou

After formulating some ideas, I started recording the various parts as MIDI data. This was initially linked to the EastWest Orchestra plugin's Steinway piano, and Shortcircuit sampler loaded with a music box sample. The concept was to replace the piano and music box by degrees using custom-made synthesizer sounds. The distinctive melody on the chorus was added right away. This, like most of the other synthesized sounds, was realised on the Novation K-Station.

Percussion elements were also added early on. I decided on a bass drum and snare combo I came up with for my (as yet unreleased) soundtrack project. The bass drum came from my Alesis SR-16 drum machine, and the snare from my Yamaha MR-10 rhythm box. All percussion was loaded into instances of Shortcircuit so I could create a complete drum track for the song within the Sonar workstation software. Ride and crash cymbals and some massive tom-toms also came from the SR-16.

In Numan's original, there are percussive sounds which mimic the ticking of a clock. I believe they were produced by recording claves through an effects pedal. I opted to keep the clock motif but use sounds from a real clock, in this case an all-metal alarm clock I recently acquired (I used to dismantle them when I was a kid!) Another percussion element, which punctuates the breaks between vocals, is actually the sound of ice shattering. As a bonus it happened to sound clock-like.

Again dipping into my recent soundtrack experience, I called upon an effect I'm fond of, which involves running an electric guitar through the Roland SP-404's 'lo-fi' effect. This effect mimics lowering the audio bitrate. If you've ever heard a really badly-encoded mp3 you'll know what that's all about. For the layman: high quality digital audio is generally recorded at 24- or 16-bit. Old Atari game systems etc... used 8-bit audio, and a fair few of us 'children of the 80s' understandably have a penchant for that kind of sound, hence the invention of the 'lo-fi' effect. You can dial in as much lo-fi as you want until all you hear is bubbly white noise. Needless to say, I go for somewhere about middle. The result is a strange, buzzy, almost chime-like sound somewhat like a grandfather clock chiming in a rusty shed.

Next to go on was a simple square-wave lead made on the Yamaha CS01 synthesizer. I played a solo of my own devising in place of the saxophone which plays out the original song. The solo proved problematic when interacting with other elements: it clashed with some notes I came up with to accompany the original piano parts. I decided to leave the solo intact and remove a few of the offending notes from the accompaniment. Creating an almost atonal synthesizer patch also helped the sounds 'play nice' with each other, as much as that sounds counter-intuitive.

A descending melody, played on Crumar Performer string synthesizer, was initially going to accompany the choruses. They were busy enough as they were, so I saved the strings to follow the solo which ends the original version. Once the tom-toms were added, it quickly built up to a powerful send off for the song.

Next was bass guitar. The original features the inimitable talents of the late Mick Karn on fretless bass. The only part of his performance I hinted at was a piece of melody on the chorus, played on synthesizer. Aside from a couple of notes during the verses, my bass parts were saved for the outro. I wanted a little more bottom end without competing too much with the piano. Throughout the track I left the piano bass notes in. The music box is also still present where I thought it sounded good.

Early on, I envisioned my friend Jimmy Aaron singing harmony with his distinctive voice on the chorus. I asked him if he was up for it and he said he'd give it a go. Jimmy is one for experimenting with vocal ideas and improvising. Not only did he sing on the chorus for me, but he came up with a simply incredible vocal track for the outro as well. My intention was to leave that part instrumental, but after adding the vocals it didn't sound the same without them. We also toyed with the idea of singing 'aahs' to accompany the solo, but decided against it in the end.

In Jimmy's words: "I'm glad you liked that outro idea, I like it too! :-)) Those bells and the word innocence go so good together. And then 'what's done is done,' a reminder it's the end of the song..." A little glimpse into his thought process when choosing the lyrics. Having worked on a track together in the past, I know my voice compliments Jimmy's best if I sing low. But when it came to following his melody, I realised I couldn't quite hit the low notes if I stayed in key with the first line. I toyed with the idea of singing higher and pitch-shifting the result, and even leaving the first line out altogether, before coming up with a vocal melody I could manage while staying in the lower register. Usually my vocals require a lot of editing: I choose the best parts from the best takes. In this case, the two takes you hear on the outro are unedited.

Rather than follow Numan's original vocal melody for the verses, I decided to wing it. The odd line came out much the same, simply from being ingrained in my memory, but the bulk of it is my own spin on the words. The line that proved trickiest was the first line of verse 2. No matter how you look at it on the lyric sheet, it doesn't fit the music! My vocals are far from perfect but I'm happy with the overall result.

I'm still waiting to hear whether this will be included on the compilation, but as soon as I know, I'll post the details here. Until then, enjoy the song. If you click through to the Soundcloud page, you can download the high-quality mp3 (it's free!) and also check out two remixes I did last year. One is another Gary Numan song, the other by John Foxx - two artists who've been highly influential to me.

Tales Music: Turned To Stone

Another transition piece for Tales Episode 1, 'Turned To Stone' marks the demise of the Troll Tree Removers after they're petrified by the sun's rays. I wanted it to have a haunting quality. The first idea that came to mind was the sound of tingsha (Tibetan prayer cymbals). These are small cast bronze cymbals on a leather strap which produce a high-pitched ringing when struck together. They're more like bells in that respect. Each cymbal is supposed to sound identical, but possibly mine are slightly out of whack because they produce some harsh harmonics. This just adds to the character though, making them all the more suitable for this track.

I started by recording several takes into my CAD mic and selecting the best ones. If the cymbals wobble about too much you get a wobbly-sounding recording, so after some experimentation I got some clean takes. You can be sure I kept the wobbly ones too, though - you never know when they might come in handy!

I loaded the sample I was happy with direct into Sonar (my multi-track software) and repeated it every couple of bars. I applied reverb to soften the sound a little, but had to tweak it quite a lot. Certain frequencies were bouncing around and sounding horrible. This served as my guide for adding other instruments.

Next I added a plucked spring sound, sampled from my 'Spring Thing' which is nothing more than two dishwasher springs mounted on a board. But hooked up to a contact microphone it sounds fantastic. This already had a reverb all its own, but I added extra software reverb and it totally stole the show.


For the third element I turned to the K-Station synth, with the intention of adding a slow, bright, metallic arpeggio. My tweaking produced a detuned sound - not unlike bells that have been really messed up - more suited for melody, so a melody it was. In an attempt to fit melody to tempo, I threw in a bass-drum sample to play to. That lead to the mimicked sound of a slow hearbeat, which echoes the plight of the poor trolls and further enhances the track.

Oh yes, I almost forgot about the growls. I took a sound from my modified Yamaha PSS-140 FM keyboard and pitch-shifted it way low in shortcircuit. To me it sounds like the growl of an alien animal. No real reason for adding it except that it adds some spookiness.

By then the track was pretty much finished, to my mind. But there's always the temptation to add more, and later on I thought of some string parts and had to try them out. I called upon my trusty Crumar Performer analog string synthesizer for this. I dropped the bottom and mid-range out via the on-board EQ, and modulated the pitch. The latter is a trick I used on the Doctor Who: Mechalution soundtrack / Let's Build Mecha e.p. to give the strings a wailing, melancholy quality, not unlike tape-wobble.

The final element I added was a Chinese woodblock (also called a 'wooden fish'). This was again recorded using the CAD. It sounded too bright in my first version so I pitch-shifted it about an octave lower and EQ'd out some high-end.