I recall reading an interview where they described a rather creative use of reverb, where sounds/passages would be fed through a sequence of varying spaces — to me, it really intensifies this feeling of plasticity/elasticity.
@Olivier they are just using Altiverb and running a reverb into another reverb.
The bulk of their older self titled record was mostly Waldorf Blofeld, microKorg, DSI Tetra, hand mallets, Line 6 multi fx and some other random studio bits.
Reassemblage onward is almost entirely Ableton. They still use Blofeld controlled by a Korg micro sampler for MIDI. A lot of GRM / IRCAM type VSTs for effects rather than generic filter / chorus types. Their sound sources are Kontakt or Ableton sample VSTs, or samples cut from old recordings. A lot of the voice generation is from a program called Infovox Voice Manager, which lets you generate text to speech with an accent. They also use Ableton’s Operator VST. AFAIK the physical modeling VSTs were used on the newest album, from the company Swam.
Parts of this could be done in hardware but it’d involve a lot of convoluted signal chains and buying extra gear to do something in 15 seconds with a computer. Its intentionally working with the strengths of modern computer music. While you can get complex with CV control, the specific ways in which Ableton automation and MIDI devices bounce off each other has a unique feel.
SOURCE: I’ve worked with them on events and one of them sold me weed for awhile.