Noise School: Ableton Live Sets + Uboa - 2023 - 8 - 23 (Empathy Shield outtake 2)

Here's the other outtake I found buried in an Albeton folder. It's longer than the original Empathy Shield and probably my favourite outtake. I'll make this a noise school too and describe some aspects of how I set up live sets and record from them.

All these outtakes are in the "Set" are recorded directly from ableton by recording from the master bus output (basically, making a new track adn setting its "in" to "Main". "out" can also be "main" - it won't loop infinitely and break anything. Now whatever you play on ableton while this track is activated will be instantly baked down in real time. It is VERY useful.

The original empathy shield was recorded from the desk of the radio station, hence it lost a bit of quality (which I don't entirely mind).

Likewise, you can jam a session to multitrack. I often do this when I want to "seriously" release something. Sadly ableton does not get down everything - it does not record if you solo a channel, and if you use generative synthesis and/or randomisation it will be different every time unless you bake them down.

---

Here is the Ableton set of Empathy Shield, its outtakes and last year's show in Perth, using the "Session mode", which is played in real time. It's using this setup.

Not all of these are samples. Here are the tracks from left to right.

1 - "BlownOutEtherSlicer" is the instrument for the cut-up slicer, controlled by the grid. It has two midi loops for when I want something more rhythmic/predictable or when I dont have the hands to do noise in a section. This has a lot of randomisation, so recording it for a release I always bake it down. Mixing to something that changes every time is a nightmare. It's made of a bunch of samples made with a combination of highly distorted, "blown out" field recordings and noise taking from the ether.

2 - "Industrial Machine" is another MIDI sampler. instead of play noise slices, it plays a random combination of dark ambient drones and field recordings. I was always dissatisfied that "industrial music" didn't literally sound like a factory, so they have that theme. the samples have a slow attack and even slower release, so I can tap the buttons and it goes while I do other things. I play these with the "computer MIDI keyboard" function, as they are atonal. There is also a MIDI randomiser that makes sure that a wide variety of samples are possible from a single key press.

3 - the "gore machine" - similar to the "BlownOutEtherSlicer" except gore sounds, similar to A Puzzle (which was being or had been re-worked around this time). I always was dissatisfied that the genre of "gore noise" didn't literally sound like gore. These are one-shot rather than sliced samples, with a bit of compression, granular delay and distortion. They are also controlled with the grid, and I switch between the slicer and the gore with an A/B switcher on my APC. Some of the samples I made can be found here, as they are recycled from A Puzzle (and it was mixing this track Tig and I had the idea for a gore machine), the other samples (theres 128 of them) are from sample packs and old video games.

4, 5, 6, 7 - ignore the titles, they are basically leftovers from previous sets (I usually build my sets on previous ones, saving work on mixing it). These are samples, looped and otherwise. They are colour coded - brown (and pale violet) for melodic (different keys use different colours), cyan for textures or atonal sound design, blue for loud textures or sound design, orange for rhythmic (say the big drum sample in this outtake). Basically, similar colours have similar uses, but might vary slightly. This all helps immensely as mnemonics so jamming is more intuitive rather than squinting at unreadable text.

Yes, theres catgirl sounds, the drone from Jawline, the drone from the "Mars" section of Yume Nikki and some drones from the sample CD that was used for Silent Hill 2. Also some from the ableton library (Spectral Textures), youtube, sample packs and splice. Again, sampling is one of my favourite things to do.

8 - The input group channel for the digitone, direct into computer via USB / Overbridge to save on cables and setup time. each channel is a different synth in the digitone, usually 1 noise synth, a low drone synth, a pad synth and a 'melody' synth (or second pad synth). All of these either have a very slow release or infinite sustain, so I can press a note and then use my hands for something else.

9 - Vocal input channel, from mic into the VE-20. I don't always mix my vocals through the laptop - often there is a very slight delay which can interfere with singing, it can be difficult to get specialised foldback in the monitor and the reverb can easily feedback given I usually don't know what PA or space a venue has in advance. However, when it does work, it means I can put all the normal ableton effects on the voice to make it sound almost studio-quality and make it blend into the mix better. The various loops are just dummy loops, they are just all coloured green for mnenomic reasons on the APC.

10 - "Noise" - this is hybrid noise. I've gone into this elsewhere. The loops are specific sounds (rather than just input from pedals/contact mic/synth) that can be fed into the FX circuit to vary the noise a little, although it cuts out the original signal. Often these are momentary, almost like the grid.

11 - "Ms Distortion" - a second vocal channel specific for screaming. I don't really do this anymore. It uses a second "screaming" mic, and has a totally different set of effects to make the screams sound bearable in the mix (i.e. less reverb, saturation, compression) without causing too much feedback in the monitors live. It can also double as a specialised loop/sample track, with specific pre-recorded screams and even some radio-like recordings so I have another source of noise that I can do hands free. (really wish I had more hands sometimes).

Theres other tracks to the side, but they are all dummy tracks for recording - one for baked-down recordings of the Digitone, another for the master, and a dummy track for the Digitone input.

Master bus:

Finally, theres obviously a lot going on here.

To make sure I don't have (unwanted) clipping in the master bus - nor horrid unwanted dynamic jumps, I use both compression/limiting in the masterbus. I also use heavy compression sidechaining between tracks and samples - if two sounds are competing for the same frequency space and one of them is more important (say vocals, bass synth, cut-up noise) then I sidechain the other a little bit to give more breathing room. Channels with large dynamic range and transient spikes are compressed and limited (say the slicer, or a clanky sample)

there is also a minor eq (and a hi-pass) and some multiband compression to make sure the mix sounds glued together and to add some more consistent low end (and remove infrasound that might cause issues with the PA). Finally I limit things to -6 DB, giving plenty of headroom to the engineer of the night so I don't blow up their speaker. Basically, it is a basic real-time mastering job, hence why these recordings sound semi-mastered. They are not fully and properly mastered, I wouldn't put these on bandcamp or a physical like this, but most of the work is done. If you don't do an obvious mistake and the mix is already sounding good, then you can basically go straight from playing the set to mastering. However, it's complex, and mistakes (like accidentally muting a track, or admittedly singing a bum note) can be easy.

---

Hope any of the above interests you. If you have any other questions about how I do live sets put them here. Sadly can only upload mp3's of this song, as it is too long.