EverSpace vs Pendulums

EverSpace is an algorithmic reverb designed for creating enormous spaces, evolving ambience and richly modulated echoes. Built upon the CloudSeed engine by Ghost Note Audio, its architecture draws inspiration from the algorithmic studio reverbs of the 1980s, where networks of delays and diffusion were used to create spaces that extended far beyond the limits of the physical world. Rather than attempting to recreate a particular room, hall or vintage hardware unit, EverSpace embraces the more creative possibilities of algorithmic reverb. Dense diffusion, multiple delay stages and extensive modulation allow sounds to bloom into vast, constantly moving environments — from spacious reflections and drifting echoes to seemingly endless atmospheric tails. At more extreme settings, EverSpace becomes much more than a conventional reverb. Simple sounds can be transformed into thick evolving pads, shimmering textures and expansive ambient soundscapes, with subtle variations within the reverb network ensuring that the resulting space remains alive and in motion. EverSpace can be used as a traditional send or insert effect, but it is at its best when treated as a creative sound-design tool where the reverb itself becomes part of the sound. Features ○ CloudSeed algorithmic reverb engine — a sophisticated multi-stage architecture inspired by classic algorithmic studio reverbs. ○ Early reflection and diffusion network — control density, delay, modulation and feedback to shape the initial character of the space. ○ Extensive late-reverb section — create everything from diffuse ambience to enormous, slowly evolving reverb tails. ○ Deep modulation — independently shape the movement of diffusion and delay stages for lush, animated spaces. ○ Multi-tap delay stage — adds rhythmic reflections, echoes and additional complexity before the main reverb network. ○ Integrated equalisation — low and high frequency shaping plus filtering to sculpt the colour and decay of the reverb. ○ Seed controls — regenerate the internal delay and diffusion structures to produce different variations from the same settings. ○ Interpolation modes — alter the behaviour and character of the internal delay network. ○ Real-time visualisation — waveform and frequency displays provide visual feedback as sounds enter and evolve through EverSpace.

Twin Double Pendulums is more chaotic than my other engine using 2 Lorenz. This is more suited for Oriental music like Koto. The pendulums are released from the same angle, one of them offset by a ten-thousandth of a degree. Both drive a lead voice. While the runs agree the two voices land on the same note, so you hear a single line over a rolled arpeggio. As the gap opens the twin becomes a real second voice, chords swell in behind it, and once the runs have parted for good the whole band is in. Nothing is pre-recorded and nothing loops. Every release is new. THE SONG FORM IS THE PHYSICS There is no bar counter and no scripted arrangement — the piece is however long the pendulums stay together. Their separation is the structure: IN SYNC — lead and twin on the same note, one line over an arpeggio that re-voices itself every bar. DRIFTING — the twin splits into real intervals; chords enter and swell. DIVERGED — bass enters, the grooves open up, the mood engine wakes. Watch it happen: the gap is plotted live on a log scale with the running Lyapunov exponent beside it, so you can see the divergence coming before you hear it. DRAG THE RIG Pull the pendulum round to set the release angle and both twins let go again, from anywhere in the piece. Release it inside about 79 degrees and the motion stays regular, the twins never part, and the solo holds indefinitely — chaos is a regime, not a law, and the scope draws you that boundary while you drag. Outside it, the twin offset sets how long the solo lasts before chaos takes over, though only logarithmically: each extra decimal place buys about the same few seconds again. CHAOS PROPOSES, A POLICY DISPOSES The far bob's position proposes a region of pitches rather than one note. A Markov policy then decides what actually plays, learning from its own output and rewarding intervals that sound consonant. Slide BLEND to choose who wins. WHAT YOU GET • 26 scales — major and minor modes, pentatonics, blues, whole tone, harmonic and melodic minor, altered, both octatonics, and folk scales from Japan, China, Egypt, Spain, Arabia and Hungary • 8 accompaniment styles — chaos grooves, jazz comping with walking bass, stride piano, Alberti bass, classical, arpeggios, oriental rolls, ambient • Four sampled instruments — grand piano, koto, zither and shakuhachi — plus synthesised voices: soft piano, glass, strings, FM bell, AM organ, pluck, fat saw and more • 6 moods on a valence/arousal plane, from serene to tense, driving chord quality, register, tempo and note length • Expression built into the physics: ring, note bend, grace notes, and octave doubling called by the second bob • Live displays — the rig itself, the divergence chart, the note-probability distribution the engine samples from, and the mood point • Record a take and export it as a standard MIDI file to the Files app — lead, twin, chords and bass on separate tracks TWO AUDIO UNITS Pendulums ships as an instrument AUv3 and as a MIDI processor AUv3, so it can either play its own sounds or sequence any other instrument in your host. Lead, twin, chords and bass each send on their own MIDI channel, so one instance can drive four different instruments. Both Audio Units follow your host's transport and tempo — hit play in AUM, GarageBand, Logic, Cubasis or Zenbeats and it locks to the grid, on the bar line, even if you start mid-bar. One install registers both. CREDITS Sampled piano: Salamander Grand Piano by Alexander Holm (CC BY 3.0). Koto and shakuhachi: FluidR3 GM soundfont by Frank Wen (MIT). Zither: đàn tranh from the Versilian Community Sample Library (CC0).