In the olden days, keyboard players had no choice but to change the sounds on their instruments 'live', twiddling Minimoog knobs or adjusting organ drawbars with one free hand ready for the next part of the song. It was part of the performance, in many ways. As technology advanced in the '1980s-'90s, it became standard for keyboard instruments to have storable 'patches' that could be set up and then recalled exactly during a gig. This is great for ensuring consistency of sounds, and freeing up the keyboard player to focus on playing rather than knob twiddling (although adjusting sounds 'on the fly' is still great, for that live, spontaneous effect - it's just nice not to have that as the only option). So the next stage of handling of sound changes involved keyboard players dialling in pre-programmed patch numbers on each of their keyboards as the songs progress (e.g. patch 76 might be "Smooth strings" for one section, patch 52 might be "Phased electric piano" or whatever). For more complex songs with more sound changes, keyboard players soon hit a few problems...

• The next sound required is on the same keyboard, immediately after the previous sound. You sometimes see keyboard players tackle this problem by just stopping playing half way through the last bar of the last sound, to dial in the next sound ready for the new section, which is pretty undesirable.

• The next sound required is on another keyboard, immediately after the previous sound. So the keyboard player can probably get the other keyboard cued up ready with the new sound in advance, if he has a spare hand at some point. But he still has to leap over to the other keyboard in what is effectively zero time, and hit the right notes/chords without it sounding like a crash landing.

• The sound required is on the same keyboard, but it can be switched easily (e.g. by a foot switch, or single button-press). However on many digital synths, it takes a short time for the new sound to load up - half a second or more is not uncommon on some keyboards, meaning the previous sound is muted and you can't play the new sound for a short time. Again, not desirable if you need to be playing notes on the old sound right up to the new section.

What's the solution to all this? The approach I've found relies on the fact that controller keyboards can send out on different MIDI channels, and can also send out program changes to control other equipment. In my case, playing primarily on a single keyboard (my trusty Kurzweil PC3x), rather than playing multiple keyboards, simplifies this. When I'm about to play a song, I dial up a pre-programmed patch on the controller keyboard which sends out program change messages to all my other synths and synth modules, initialising them with the various sounds and volume levels I'll need from them for the song in question. Where a single synth or module will play multiple sounds, I place these on different MIDI channels. The end result is that all my equipment is ready to go, with all sounds pre-loaded and ready on different MIDI channels. Then all I need to do is to configure my controller keyboard to send to different MIDI channels for each part of the song. I do this by setting up consecutive patches on the controller keyboard that direct different zones on the keyboard to the various MIDI channels, and also volume control information for each sound, and then step through them incrementally using a foot switch. The result is that to get to the next sound on whatever keyboard/module it's on, I just need to hit the foot switch. No need to even lift my hands off the keyboard - they're already in place to play the next part, so no need for any diving/stretching/acrobatics onto another keyboard. No need to remember patch numbers, find a spare hand to select patches, and no need to suffer delays while the sound loads. In fact the Kurzweil PC3x excels here, because it tracks note information across patch changes - what this means is that if I'm holding notes down when I hit the foot switch, those notes will continue to sound on their respective MIDI channel, while any newly-played notes will route to the new MIDI channel (and thus the new sound). So sound changes can be super smooth, and I can even hold notes/chords over the transitions. Some controller keyboards will lose track of held notes, so the MIDI "note off" message is lost, and the held notes will hang forever. Others defeat this by forcing notes off when you change patch, so sounds are clipped (not ideal, but at least you're safe from the dreaded hanging note).

The other benefit here is that any other foot pedals, such as switches for effects (e.g. Hammond organ leslie speed) or expression pedals (e.g. for foot-controlled volume changes) can be assigned to different MIDI channels for each part of the song also. So my expression pedal might control reverb amount on one sound on one part of the song, and then volume for a fade-out of a different sound at the end of the song. If you've ever seen an aerial shot of Rick Wakeman's keyboard rig, not only will you see about a dozen keyboards, but each keyboard will have at least 3-4 foot pedals/switches....so it's really nice to be able to just have a couple of generic foot switches/pedals and have them re-assigned as I need during each song.

Of course, there are some songs that require more than 16 sounds (the number of channels supported by MIDI), and the Kurzweil PC3x only has a single MIDI OUT socket (something I've protested about to Kurzweil several times now!). The work-around is to accept that you're going to have to send some program changes at some point during the song, so you can re-use MIDI channels. But usually you have a lot of flexibility in when this happens; for example if the song intro uses a sound that's never used again, you can send a program change out on its MIDI channel any time after the intro and start re-using that channel. Not ideal, but it works, and still avoids glitches/gaps.

So on my controller keyboard, a song typically has a series of patches that I step through incrementally. This can be anything from 2-3 for a song with few sounds such as IQ's "Speak My Name", to maybe 25-30 for a lengthy complex track like "The Seventh House". To arrange a setlist for a gig, I don't mess around moving these blocks of consecutive patches around; I just note down the starting patch number on my setlist and dial it in before each song, and the first patch sends out program changes to configure everything ready. Observant IQ gig audients might notice as the lights go down after each song, I use my wrist glow-stick (essential gig item!) to illuminate the buttons on my keyboard so I can type in the number of the next song.... :)

More updates soon, as I set about re-building my synth rack with a totally different and exciting configuration!

So I decided to rebuild it all into a single larger (16 unit) rack case, so that the interconnections were all neatly hidden away inside, and cable-tied down to prevent anything getting loose or tangled. The resulting rack has served me well over the last year, and has travelled with me all around Europe to a whole load of gigs.

For those of an inquisitive disposition, the units in the rack, from the top down are:

• Yamaha Motif Rack ES
• 2 x Muse Receptor 2+ Pro
• 2 x Behringer ADA8000
• 2 x MOTU Midi Timepiece AV
• 2 x Nord G2 Modular Engine
• 2 x MidiMan Fineline mixer

Most of the units appear in pairs, one being a backup in the event of failure. Surprisingly, although most of these units are built like tanks, things do fail, especially in the harsh conditions of the stage environment, or bumping around in the back of a truck. Having a backup MOTU Midi Timepiece AV saved the day at the recent IQ gig at the Melkweg in Amsterdam, when the primary one failed (due to a dead internal CMOS battery). I don't have a backup of the Yamaha Motif Rack ES simply because it really is built like a Panzer tank, and I actually only use its sounds in a few minor places.

Most of my live sounds come from the Muse Receptor, and the Nord G2 Modular Engine. The Receptor is essentially a rack-mounting computer which runs synth plugins such as Omnisphere, Kontakt, Ivory II and so on, making it just about the most flexible synthesis module available. It connects to the ADA8000 module via optical cable, which in turn provides 8 analogue outputs, allowing me to send 4 different stereo feeds to the venue PA system for different types of sounds (pianos, Mellotrons, leads etc), and thus allowing the sound engineer some flexibility in adjusting relative levels during a gig.

The Nord G2 Modular Engine is the modern day equivalent of the giant Moog Modular systems Keith Emerson used to haul around the world; essentially an infinitely flexible modular synthesizer. It has no external controls apart from a power switch; you devise sounds on it using a computer, which provides an immensely powerful interface where you can stack up synthesis components such as oscillators, filters and so on, and wire them together in weird and wonderful ways. Want more oscillators? Just drag a few more into the rack! Once you're happy with the sound you've created, it can be stored and recalled instantly during a gig, without the need for an attached computer. The result is a mind bogglingly capable, eminently giggable, rich-sounding analogue modular synth, which I use for many of my lead and polysynth type sounds.

Inside the rack, all of the modules are linked together using nice short cables to avoid any tangling, making it easy to make modifications, re-route signals and whatever else I need to do.

The rack served me well, but if you were to ask the opinion certain members of the IQ band or crew, you might get a somewhat colourful response. It was the most evil thing imaginable to lift! The Receptors alone are 7.5kg each, and all in all it weighs about the same as a person, which makes it not entirely fun to lift up to neck height onto a stage, or up slippery metal steps to a rehearsal room.

As well as the weight, there are other drawbacks with the system. The MOTU units route MIDI from my three keyboards to the synth modules in a very static fashion. What this means is that a range of MIDI channels is dedicated to the Receptor, some to the Nord G2 Modular Engine, and others to the Yamaha Motif Rack and the physical keyboard synths I use on stage. Consequently I'm restricted to about 6 MIDI channels for the Receptor, which then limits how many plugins I can have set up ready for a song, without the need to load other plugins. The Receptor provides features to combat this, called Snapshot Multis, that allow you to switch between different banks of plugins almost instantaneously. However it does limit how smoothly you can make such transitions within a song, and makes managing song setups more fiddly.

And so after lots of careful thought, I've come up with a plan for a completely revised synth rack, which I think will weigh in at a little over half the weight of the current one (well, maybe two thirds), and will provide complete flexibility in which MIDI channels are assigned to which physical modules and software plugins. This means I'll have more freedom to use exactly the sounds I want, for each part of each song, without compromise, and hopefully without too many headaches (or back aches from the weight). I haven't worked out all the details yet, but I shall be posting to this blog as I make each step towards the next working evolution of my synth rack. Stay tuned!

My keyboard setup has evolved through many variations over the years, and along the way I've learn many tricks, learned from many mistakes, and slowly I've managed to tailor it to my needs, and overcome its limitations. I'm hoping perhaps in making a blog about this, it might be useful for other people following a similar path, who knows?

Another reason I decided to start this blog was that I'm about to embark upon an ambitious rebuilding process, effectively replacing much of the technology that has been serving me well for the last few years, and so I thought documenting that process might prove to be interesting, at least for me!

If you find it useful or interesting, please chip in with comments or questions, that's what it's all about.