Just got back from the mini-tour with Mark (Mosher), and am kind of going through my memories of the performances that took place. The overriding thing that surprised me was that my Korg ESX-1 was completely underutilized. Not only did I not use the sampling engine, I never even turned on the sequencer's transport. It served one sole purpose: acting as a series of buttons that would stimulate the MIDI port on the modular!
Now, I love the ESX (I recently found an SD card version, which made it much more lovely...), and was sure that it would become an integral part of my performance rig. I custom created a bank of sound that would be conducive for the sort of scary-ambient soundscapes I generally make, and I made a number of banks that had well-tweaked effects and channel settings that I could depend upon during performance. I rehearsed with the machine so that I could immediately hear a problem and fix it in the fly - the actual hard work of learning a new instrument.
So why wouldn't it work for my performances? Here are a few reasons I've come up with:
- Within the fluid manipulation of analog synthetics, the use of samples sounded way too prepared. I want people to understand that I'm hand-managing the things that they are hearing, and there just isn't that much you can do with the samples in the ESX that isn't bloody obvious.
- In comparison to the sounds of the digital part of my modular, the sample playback of the ESX sounds lame. I've got three digital sound sources in my modular: an MI Braids Swiss Army Oscillator, a Qu-Bit Nebulae granular player, and the Music Thing Radio Music sample switcher. Each of these creates a much more dynamic digital soundscape than the ESX. The Korg is fun on its own, but it gets kind of swamped in these environs.
- There's no way to do real-time resampling. Or, at least, there is no way to do it that doesn't make me want to take out my eyes with a spoon. Loop selection, for example, is a nightmare.
- The use of repetition, even the 8-bar variety, doesn't seems to fit. I'm not sure why, but even using the long-form sequencing of the ESX-1 sounds really mundane in this context. I applaud the ESX-1's 8-bar pattern sequences, but the inability to have different length tracks means that your ear immediately latches on to the repetitive elements - unless you are hands-on the controls at all times. But I'm hands-on the modular, so I can't pull it off.
But rust never sleeps!!!