This Monday I had my first gig with Terra. I feel like I learned a lot about Terra while preparing for this performance, and I took some notes. I thought that maybe someone here would be interested in that, so here it is.

For the context: I played solo, with Terra as the main instrument, Octatrack for looping, sequencing drums & resampling, and an iPad for some virtual instruments. Most sounds were coming from Terra itself, and I also used Korg Gadget (drums sequenced by Octatrack and played live from Terra, as well as acoustic bass and piano, controlled by Terra).

General notes:
– Many people were intrigued by both the looks and the sound of Terra. It was a “live music & sketching” event and most people in the audience were drawing me, so the visual aspect was appreciated. Especially that many of the ‘expressive’ controllers out there have… questionable easthetics.
– Terra is small and light and that is awesome for transport and performing on small stages.
– Gibraltar GEMS stand is great for Terra. Depending on how wide and tall you set the legs of the GEMS, you can wobble it with intention and control Terra’s gyroscope for vibrato.
– Because of the distinct layout of the sensors 1-12, it’s possible to play even complex phrases with one hand and/or without looking at Terra, and switch something on my Octatrack with the other, or look at the audience, or just close my eyes. This sets Terra apart from other ‘expressive’ instruments/controllers like Flux, Continuum, Linnstrument in my opinion.
– Note down the presets; I use groups A-D, 1-6, 7-12  as logical “sub-banks”. I need between 2 and 6 presets per song. Some of them are only for MIDI instruments, so the presets carried the tuning, G sensitivity, the MIDI velocity mode (which depends on the engine).
– Sometimes loading a preset/tuning or shifting the tuning by an octave (fork + ^H + pitch-shifter) doesn’t register; I’m 90% sure it’s user error, but since there is no immediate feedback, this is a bit of a problem for me. I’ll see if it gets better with more practice.
– The default pitch-shifting tuning is not for me. In most cases I need 1 octave up/down, so this is what I put on the ‘outer’ pitch-shifting sensors. My favorite factory preset is preset 10-110.
– I didn’t really need to use external fx; it sounds great on its own. I did apply effects with octatrack on recorded loops (mostly things like LPF to blend the tracks out of focus), but solo parts were totally fine “dry” (using internal delay/reverb and engine-specific fx).
– Mastering the ^H (timbre-control hold) and understanding its limitations is crucial in my opinion. I use it most often for locking a sweet-spot for one or two sensors (allowing me to shift focus to control of the other parameters), or when I need to free the left hand to do something on the Octatrack.
– I often change the G sensitivity during performance. e.g. 1 (the lowest) when I’m about to loop something that I want to stay strictly in tune; 3-4 for most solo parts; the highest 2 when doing something really weird like holding a note with H^ and tilting the Terra with my both hands. G also applies to MIDI control, so you might disable it for some types of instruments.

MIDI-related notes:
– I think I have ‘abused’ Terra as a MIDI controller with virtual instruments that are clearly not a good fit (because they were not designed with something like Terra in mind). So the list of my gripes below is not a critique of Terra’s design, just some things to be aware of if you want to use Terra this way.
– Terra’s sensors are physically not great for controlling initial velocity, and therefore not optimal for instruments like piano or drums. Still, since I really wanted to keep my setup small, I used Terra for exactly these instruments and it’s still workable.
– It works OK as a finger drumming MIDI controller. I use it without velocity. I find it the most convenient to set kick to sensor 4, snare to 5, hi-hat to 8 (closed) and 9 (open), toms to 10 and 11, and 6/7/12 on different parts depending on the kit/song. Without velocity It’s easy to accidentally trigger a sound by just barely touching the sensor, though. Regular finger-drumming pads are of course better, but with Terra it’s OK enough that I didn’t have to add another device (and a stable surface) to my setup just for finger-drumming. Maybe there is a way to set some parameters in the controlled instrument to make it work reasonably with Terra in velocity-sensitive mode, but I need to dig into it more. I used Korg Gadget (gadgets Gladstone and Recife).
– I used it as a MIDI controller with velocity for some piano and acoustic bass virtual instruments (Korg Gadgets: Madrid for bass, Module for piano).
– Lack of MIDI program change (either sending or receiving would be enough) is problematic; this means that I had to often change something both on Terra and another device at the same time. This is the biggest problem I have with Terra. In some songs I need to switch between Terra’s sound to external MIDI control and that requires:
1. Loading a preset on Terra
2. muting/unmuting Terra (I’m using a volume pedal for this)
3. Enabling/disabling MIDI
4. Changing the MIDI channel (depending on the setup)
– I guess I will have to leave MIDI on and use a single channel and do some clever programming to ameliorate these issues, but it’s not feasible with every setup; e.g. I can’t automate muting channels on my mixer. Or I might just have to adjust my arrangements to give myself enough time to do all the steps above
– But the fact that it doesn’t process most MIDI messages allowed me to skimp on an extra MIDI thru box in my setup, and connect everything with just one Kenton MIDI merge box, because I can have Terra in a MIDI loop without issues
– …but if you do have Terra in a MIDI loop, you must stay clear of tuning of Terra sensors while any sequencer is sending notes. Because during the tuning, MIDI notes are received by Terra on all channels.
– When using Terra as a MIDI controller for most „normal instruments” (that were not designed or adjusted for expressive instruments) you might want to disable CCs corresponding to the sensors A-D (CC 16-19); otherwise touching and releasing any of them will set something to zero, and often mute the virtual instrument.

Notation for the ‘pyramid’ sensors:
– For noting down the presets, engines, notes etc. I use a notation that can be entered in text and read quickly by using “1” for a lit sensor, “0” for unlit, and in the shape of X-XX-XXX, so e.g. 1-10-001 means that the top sensor is lit, the first sensor in the middle row is lit, and last sensor in the third row is lit.
– In some contexts, some rows can be omitted, e.g. for engines you only need XX-XXX, and for preset banks only one row is needed (and additionally a sensor A-D / 1-12).
– example: the engine “Electrons singing” is 11-011
– example: Preset in the last bank + sensor A is 001-A

Overall, it was great and I’m looking forward to performing more with Terra! I turned out to be totally possible to have it as the only instrument/controller for the entire set (around 1h 20m).

On a more general note: some design decisions in Terra that baffled me somewhat at first make more sense in the context of performing (as opposed to e.g. sound design, or optimizing for use by beginners). For example the tuning and preset management is cumbersome, but almost any operation that I need to do during the performance can be done very fast and often without the need of visual feedback.

Photo of my setup: