# OVNI Audio — everything in one file The long version of https://ovniaudio.com/llms.txt: the catalog, the page of every released plugin, the full manuals, how to install, which one you need, the questions, the price and the licence. Generated from the same data as the website. No images. Other versions of this file: [llms-es.txt](https://ovniaudio.com/llms-es.txt) (Spanish) · [llms-full.txt](https://ovniaudio.com/llms-full.txt) (everything in one file) · [llms-full-es.txt](https://ovniaudio.com/llms-full-es.txt) (everything, in Spanish). ## The catalog OVNI Audio is an indie plugin label from Argentina, free and open-source. Three kinds: spatial FX that move sound around your head and open spaces, a visual synth (SUPERNOVA) that turns sound into image, and an analyser (TELESCOPE) that measures your mix and writes what it found. VST3 + AU on macOS; Windows VST3 where noted. ### Spatial FX Effects that move a sound around your head, open a space for it or spread it across the field — for producers and mixers. #### ORBIT — Free · v0.3.0 - What it is: A binaural movement engine: place a sound around your head, orbit it or fly it past. - Who it is for: For producers and mixers who want a sound to move around the listener, on headphones or speakers. - What it does: Sets angle and distance on a radar, bends the pitch on the fly-by with real Doppler, and keeps it mono-safe with IN PHASE. - Formats and systems: VST3 and AU on macOS 11+ (universal); VST3 on Windows 10+ (64-bit, unsigned ZIP) - Page: https://ovniaudio.com/orbit #### PULSAR — Locked · coming soon - What it is: Auto-pan that throws your sound into orbit around your head, driven by chaos instead of an LFO. - Who it is for: For producers who want movement that doesn't sound like a sine wave. - What it does: Morphs between Orbit, Pendulum, Lorenz and Rössler paths, adds Doppler on the fly-by and head-shadow width; IN PHASE folds it to mono. - Formats and systems: VST3 · AU when it lands - Page: https://ovniaudio.com/pulsar #### NEBULA — Locked · coming soon - What it is: A reverb of impossible, infinite spaces that breathes and freezes. - Who it is for: For producers and sound designers who want a tail longer than any room. - What it does: Eight delay lines you sculpt by dragging a cloud: size, and decay up to 12 s — at the maximum the tail freezes. IN PHASE keeps it mono-safe. - Formats and systems: VST3 · AU when it lands - Page: https://ovniaudio.com/nebula #### DUST — Locked · coming soon - What it is: Binaural echoes: reflections of your sound, scattered as bubbles around your head. - Who it is for: For producers who want echoes with a place, not a diffuse wash. - What it does: Each echo sits in its own direction: RATE spaces them, DENSITY grows a few taps into a cloud, and DUCK makes room when you play. - Formats and systems: VST3 · AU when it lands - Page: https://ovniaudio.com/dust #### HALO — Locked · coming soon - What it is: Shimmer that orbits: a reverb whose octaves climb while it circles your head. - Who it is for: For producers and sound designers who want an endless pad rising under their sound. - What it does: Feeds a long reverb back through an octave and a fifth, carries it around the head free or synced, and FREEZE holds the cloud. - Formats and systems: VST3 · AU when it lands - Page: https://ovniaudio.com/halo #### HORIZON — Locked · coming soon - What it is: Spectral freeze with a pulse: hold an instant, then fire it again on the beat. - Who it is for: For producers who want a held pad or a rhythmic stutter out of any moment. - What it does: Captures one spectral frame and holds it; RATE and SYNC re-trigger it in time, from 1/16 to 2 bars. Its latency is reported to your DAW. - Formats and systems: VST3 · AU when it lands - Page: https://ovniaudio.com/horizon #### AURORA — Locked · coming soon - What it is: Spectral panning: every frequency goes to its own place in the stereo field. - Who it is for: For mixers who want width that still holds up in mono. - What it does: Gives each band its own position by per-bin phase de-correlation — no Haas delay — and MONO SAFE keeps the lows in the centre. - Formats and systems: VST3 · AU when it lands - Page: https://ovniaudio.com/aurora ### Visual synth Sound turned into image — for anyone who wants the picture to move with the music: producers, VJs, streams. #### SUPERNOVA — Free · v0.4.0 - What it is: An audio-reactive visual synth: an image or a video becomes particles that move with your sound. - Who it is for: For producers, VJs and streamers who want the picture to react to the music. - What it does: 262,144 GPU particles deform your image with the audio. As a plugin it passes the sound through untouched; the app listens to your whole Mac. - Formats and systems: VST3, AU and a standalone app on macOS 11+ (universal); macOS only - Page: https://ovniaudio.com/supernova ### Analyser Measurement you can read — for anyone mixing or mastering who wants the number behind every finding. #### TELESCOPE — Free · v0.1.0 - What it is: An audio analyser that measures your mix and writes down what it found. - Who it is for: For anyone mixing or mastering who wants each finding with its number and its rule. - What it does: Thirteen lenses — loudness to ITU-R BS.1770, dynamics, spectrum, stereo, tonal balance and VERDICT, which writes each finding. 0 samples of latency. - Formats and systems: VST3 and AU on macOS 11+ (universal); VST3 on Windows 10+ (64-bit, unsigned ZIP) - Page: https://ovniaudio.com/telescope ## The released plugins ### ORBIT — MOD·00 / Primary signal https://ovniaudio.com/orbit · Free · v0.3.0 · [Manual](https://ovniaudio.com/orbit-manual) The binaural movement engine the whole family is born from. Place a sound around your head, orbit it, or fly it past with real Doppler — backed by physics. #### The deep engine. ORBIT is the flagship: a full binaural spatial engine. Place a sound anywhere around you, set it orbiting, or fly it past with real Doppler. Headphones or speakers — mono-safe with IN PHASE. #### What it does, control by control - **The radar** — Drag to place the source in 3D — angle and distance. The whole field on one screen. - **Doppler** — Bends the orbit eccentric: the source flies close then far, pitch rising and falling. A real fly-by. - **Rate / Speed** — Free, tempo-synced (1/4–1 bar), or Fixed placement. The orbit follows your track. - **Chaos** — Erratic, alien motion — darting speed and azimuth wobble. - **Width / Room** — Ear-shadow width plus decorrelated early reflections that push the sound outside your head. - **Output / In Phase** — Headphones or speakers (crosstalk-cancelled). IN PHASE guarantees mono. #### Real binaural, measured. ORBIT is the conservative, pure binaural core — no tricks, just acoustics. - **Binaural HRIR** — A variable-position HRIR spatializer: two bracketing voices crossfade with fractional-delay ITD, so the source moves without clicks. - **Near-field** — Under a meter, Duda–Martens proximity: bass bloom and asymmetric ear-shadow. The source gets intimate. - **Real Doppler** — Pitch emerges from propagation delay — no pitch-shifter: the pitch is the propagation delay moving. - **Externalization** — Early reflections (Room) lower interaural coherence, so sound leaves your head. - **Speakers (RACE)** — Crosstalk cancellation for speakers — that stage adds no latency of its own; IN PHASE drops it for mono. - **Peak-held output** — A stereo-linked sample-peak limiter holds the internal peak at −1.4 dBFS; OUTPUT and the filters sit after it (measured true-peak −1.2 dBFS with pink noise, up to −0.05 dBFS with full-scale broadband material). #### Specs - Formats: VST3 · AU - Systems: macOS · Windows - Build: Universal (Apple Silicon + Intel) - Engine: lightweight · 1.7 % CPU - Latency: 424 samples (8.83 ms) at 48 kHz · reported - License: AGPLv3 · open-source #### Download - ORBIT macOS (double-click .pkg installer, VST3 + AU universal, macOS 11+): [OVNI-ORBIT-v0.3.0.pkg](https://ovniaudio.com/download/OVNI-ORBIT-v0.3.0.pkg). Signed and notarized by Apple (Developer ID): double-click, no Gatekeeper warnings. - ORBIT Windows (VST3 x64, Windows 10+): [OVNI-ORBIT-v0.3.0-Windows.zip](https://ovniaudio.com/download/OVNI-ORBIT-v0.3.0-Windows.zip). Unblock the ZIP (Properties → Unblock) and copy the .vst3 folder into C:\Program Files\Common Files\VST3. - Upgrading ORBIT: Coming from 0.2.x — the plug-ins are now called **OVNI ORBIT.vst3** and **OVNI ORBIT.component**. On macOS the installer retires the old *ORBIT* for you, but only if it is ours (*com.ovni.orbit*): someone else's *Orbit* is never touched, and nothing is deleted — it is moved to *~/Library/Application Support/OVNI Audio/replaced//*. Then run *killall -9 AudioComponentRegistrar* and rescan in your DAW. On Windows, delete *C:\Program Files\Common Files\VST3\ORBIT.vst3* by hand before copying the new folder in. Your saved projects keep working: only the file name moved. ### SUPERNOVA — MOD·07 / Visual https://ovniaudio.com/supernova · Free · v0.4.0 · [Manual](https://ovniaudio.com/supernova-manual) · [App manual](https://ovniaudio.com/supernova-app-manual) An audio-reactive **visual** synth. Drop in an image or a video and a GPU particle engine deforms it with your sound — it breathes, electrifies, **explodes on the kick** and re-forms. Runs as a plugin **and** a standalone app that hears your whole Mac. #### Your sound, made visible. SUPERNOVA turns any image or video into a living cloud of **262,144 GPU particles** that move, glow and burst with the audio. The differentiator: it ships as a **VST3 + AU** plugin inside your DAW **and** as a **standalone app** that captures your Mac's **system audio with zero drivers** — point it at Spotify, YouTube, Serato or Ableton and it just reacts. No BlackHole, no routing. And the audio itself is **bit-exact pass-through**: SUPERNOVA never colours your sound. #### What it does, control by control - **Load & Worlds** — Drag in a photo, a photo sequence or a video, then browse **50 worlds** — presets that swap the physics, geometry and colour in one click. - **Intensity** — The master **life** fader. One knob takes the field from a calm shimmer to a full storm — how hard the whole system reacts. - **The four domains** — Every control lives in one of four groups — **Movement** (flow, chaos, gravity), **Matter** (size, trails, plexus, glyphs), **Camera** (3D depth, rotate, orbit) and **Color**. - **Color Lab & 3D** — 17 curated look ramps baked in OKLab, plus a real 3D cutout of your subject you can rotate — order-independent additive, no z-sort. - **Triggers** — MIDI notes fire events — a note explodes the field or shoots a directional ray. Play the visuals from your keyboard, in time. - **LFOs** — Four LFOs breathe any of **12 targets** — hue, size, chaos, orbit, trails. **21 beat divisions**, each with its triplet and dotted, or a free-running **Hz** mode that keeps moving with the transport stopped. Per LFO: bi/unipolar, phase and retrigger. - **Output & Syphon** — Send the frame to a window, fullscreen on any display, or straight into OBS / Resolume via **Syphon** (macOS). - **Export with sound** — Record your visuals to **MP4 / H.264** with the audio muxed **in sync** — four aspect presets and any length from **1 to 600 s**. What drives the clip is the last **30 s** you played: a longer export repeats that window, picture and sound together, crossfaded over 10 ms so the loop point does not tick. #### A real engine, not a filter. SUPERNOVA is a GPU synthesizer for light — HDR compositing, filmic tonemapping and per-preset physics — wrapped around an audio path that stays untouched. - **262k GPU particles** — A quarter-million particles sample your image and move under real forces — advected by an analysis of the audio (onsets, bands, RMS), every frame, on the GPU. - **HDR linear + ACES** — The field is composited in **linear additive HDR** and resolved through an **ACES** filmic tonemap — highlights bloom and roll off like film, never a flat clipped white. - **Physics per world** — Each of the **50 worlds** is a different physical system — flow fields, vortices, gravity, radial blasts — coupled to geometry and a colour look. Presets that change how it **moves**. - **3D, glyphs & plexus** — A 3D cutout of the subject (order-independent additive, no z-sort) you can rotate, SDF **glyphs**, velocity **trails**, and GPU nearest-neighbour **plexus** links. - **Driver-free system audio** — The standalone app captures your Mac's own output — zero drivers, zero routing, the BlackHole dance gone. On **macOS 14.2+** it uses Core Audio taps, so the permission macOS asks for is the small, exact one: **System Audio Recording** — never the screen, and nothing is recorded. (macOS 13–14.1 fall back to ScreenCaptureKit and its screen permission.) - **Bit-exact pass-through** — As a plugin the audio is a **read-only tap** — out equals in, sample for sample. SUPERNOVA analyses your sound to drive the visuals and never touches the signal. #### Specs - Formats: VST3 · AU · Standalone app - System: macOS 11+ (Metal) · Apple Silicon + Intel universal · System Audio in the app: macOS 13+ - Engine: 262,144 GPU particles · HDR linear + ACES - Audio: bit-exact pass-through (read-only tap) - Capture: driver-free system audio — asks for System Audio Recording (Screen Recording on macOS 13–14.1) - Modulation: 4 LFOs · 21 beat divisions (triplet + dotted) · free Hz · 12 targets - Output: window · fullscreen · Syphon (macOS) - Export: MP4 / H.264 · in-sync audio · 1–600 s from a 30 s window, looped - License: AGPLv3 · open-source #### Download - SUPERNOVA macOS (.pkg installer: VST3 + AU plugins and the standalone app, universal, macOS 11+): [OVNI-SUPERNOVA-v0.4.0.pkg](https://ovniaudio.com/download/OVNI-SUPERNOVA-v0.4.0.pkg). Signed and notarized by Apple (Developer ID): double-click, no Gatekeeper warnings. SUPERNOVA is macOS only. ### TELESCOPE — MOD·08 / Analysis https://ovniaudio.com/telescope · Free · v0.1.0 · [Manual](https://ovniaudio.com/telescope-manual) An audio analyser that also **concludes**: one engine, **thirteen lenses** and a rules engine that writes what it found — with the number and the rule next to each line. Bit-exact pass-through. #### Twelve lenses show. One says. TELESCOPE puts **thirteen lenses** over one analysis engine: loudness to the broadcast standard, dynamics, spectrum and sonograms, notes and key, stereo and pan, and your tonal balance against a reference. The thirteenth, **VERDICT**, turns what was measured into sentences — each one with the number that produced it and the id of the rule that fired. It is **not a mix assistant**: it measures, it doesn't judge. **Measurement, not taste.** And it never touches your audio: **0 samples of latency, bit-exact pass-through**. #### What it does, control by control - **LOUDNESS** — Integrated, short-term and momentary LUFS, LRA and true peak — also per channel. Against ITU-R BS.1770 gating, EBU Tech 3342 for LRA and the literal FIR table of BS.1770 Annex 2 for true peak. - **DYNAMICS** — How much headroom is left: PSR over the last 3 s, PLR since reset (AES TD1004), a 61-bin short-term histogram and clip events on a 10-minute timeline. - **SPECTRUM** — FFT from 1 024 to 32 768 points with Hann, Blackman-Harris or Kaiser, and free, ⅓-octave (ISO 266) or Bark bands — every bin in dBFS referred to a full-scale sine, with an exact coherent-gain correction. - **SPECTROGRAM** — Level over time: 512 log rows from 20 Hz to 20 kHz, 10 / 30 / 60 s of history and level as colour — the instantaneous power, never an average. - **WATERFALL** — The same data in depth: level as height, time receding from now. A 2.5D oblique projection drawn in software — no GPU — with an occlusion proven identical to the painter's algorithm. - **CQT** — The spectrum by note: constant-Q at 24 bins per octave from A0, a chromagram and a key estimated against the Krumhansl & Kessler profiles — never shown without its confidence and its share of time. - **SPIRAL** — The same constant-Q coiled, one turn per octave: equal notes line up on one radius, so the octave structure of what is playing reads at a glance. - **SCOPE** — Correlation, width, balance and mono loss over 100 / 300 / 1000 ms, as Lissajous, POLAR SAMPLE or POLAR LEVEL: a half-circle with one ray per degree, where what is out of phase folds onto the base and prints its percentage. - **BAND CORRELATION** — SCOPE's figures per ⅓-octave band, so mono bass and open highs stop averaging into one number. Its band-wide figure, computed from the STFT bins, matches the time-domain meter to 0.00076 in correlation. - **STEREO SPECTROGRAM** — The sonogram with phase as colour — red out of phase, green wide, white mono — and level as brightness. A cell with no energy is black, not dim red. - **FIELD** — Energy by pan direction × frequency on a 64 × 96 grid, with decay and a trail, in a 2.5D view. The label under its axis cannot be switched off: **energy by pan direction · not localisation**. - **TONAL BALANCE** — Your mix against a reference track you load, each normalised by its own integrated LUFS: tilt at equal loudness, so the delta sums to zero — checked band by band to 0.006 dB. - **VERDICT** — 21 deterministic rules over a per-second history: within range, how it will feel, where it translates, and what to check and where — every line with its number and its rule. **Measurement, not taste.** #### Measured, with the test next to it. Every measurement quoted here comes from a test in the plug-in's repository, run against a synthetic signal — and what is not measured is not listed. - **Bit-exact pass-through** — It reads L and R, pushes them to a lock-free bus and writes nothing back: **0 samples of latency, 0 tail**. Deterministic noise through all 13 lenses, with bypass on and off, in blocks of 1 / 7 / 64 / 4096 samples and from a mono source — **0 mismatches**. - **True peak, errors included** — The literal 4 × 12 polyphase FIR of BS.1770 Annex 2, and a K-weighting recalculated per sample rate, checked against the published 48 kHz table to 8.9·10⁻¹⁶. The known 4× under-read is written down, not rounded away: EBU test 3341-17 reads −6.3160 dBTP against a real −6.0. - **File = live, to the bit** — Drop a track and it is measured offline with the same classes: integrated, LRA, true peak, the 30 measured ⅓-octave bands and the per-second history come out equal to the bit. 60 s of audio in 0.24–0.63 s on an M4. - **Not localisation** — FIELD and POLAR LEVEL show energy by pan direction: no HRTF, no ITD, no azimuth. Recovering where a sound was from a finished stereo mix has no unique solution, and TELESCOPE does not pretend it has one. - **Rules, not a model** — VERDICT is a threshold table with the source of every number — no model, no network, no telemetry: the same signal gives the same report, word for word. Its per-device checks are generic, with no speaker curve inside. **No findings is not the same as finished.** - **A paint budget, with its conditions** — Every lens is timed against 4 ms median / 8 ms p95 at size L — 6 / 12 ms at Retina scale — and published with the load factor k of the machine that measured it: k 0.92–1.05 on the house M4. #### Specs - Formats: VST3 · AU (macOS 11+, universal) - Latency: 0 samples · bit-exact pass-through - Loudness: ITU-R BS.1770 · EBU Tech 3341 tests 1–5 & 15–19 · Tech 3342 tests 1–4 - True peak: BS.1770 Annex 2 · 4× FIR below 96 kHz - Channels: mono or stereo in → stereo out · stereo only - Languages: EN · ES (reviewed) · PT · FR · DE · IT (pending native review) - Licence: AGPLv3 · open-source - Windows: VST3 x64 · Windows 10+ · unsigned ZIP #### Download - TELESCOPE macOS (double-click .pkg installer, VST3 + AU universal, macOS 11+): [OVNI-TELESCOPE-v0.1.0.pkg](https://ovniaudio.com/download/OVNI-TELESCOPE-v0.1.0.pkg). Signed and notarized by Apple (Developer ID): double-click, no Gatekeeper warnings. - TELESCOPE Windows (VST3 x64, Windows 10+): [OVNI-TELESCOPE-v0.1.0-Windows.zip](https://ovniaudio.com/download/OVNI-TELESCOPE-v0.1.0-Windows.zip). Unblock the ZIP (Properties → Unblock) and copy the .vst3 folder into C:\Program Files\Common Files\VST3. ## Manuals ### ORBIT — Manual https://ovniaudio.com/orbit-manual #### What ORBIT is ORBIT is a **binaural spatialiser**: it takes a mono point of sound and flies it on a real path around your head — a circle, an ellipse, a spiral that breathes in and out, a pendulum. It is not a stereo widener and not a reverb. The position is computed from head-related filtering (HRTF) plus decorrelated early reflections, so the source actually sits *outside* the head, at a distance, moving — with optional real Doppler on the fly-by. It shines on synth leads, plucks, arps, FX hits and risers, on percussion you want orbiting the head, and on whole stems via a send. Built for headphones, but it stays mono-safe and has a SPEAKERS mode (crosstalk cancellation) so the spatialisation survives on a real PA. #### Quickstart 1. Drop ORBIT on a mono source — or a stereo one, it sums to a mono point before spatialising. 2. Pick a **SHAPE** (Ellipse is the musical default) and a **RATE**: leave it on **Free** and dial the **SPEED** knob, or lock to **1/4 · 1/2 · 1 bar** to ride the tempo. 3. Set **RADIUS** for how far the source sits, then open **WIDTH** until the movement reads on headphones. 4. For a real fly-by, raise **DOPPLER** — the source rushes close at the front and falls away behind you, pitch and all. 5. Add a touch of **ROOM** to push it fully out of your head, and flip to **SPEAKERS** output if you are not on phones. #### Signal flow The input is summed to a single **mono point source** and stage-gained by INPUT. The **Orbit brain** turns SHAPE · RATE/SPEED · DIRECTION · RADIUS · HEIGHT · SPREAD · CHAOS · DOPPLER into a live azimuth + distance trajectory, synced to the host transport when RATE is a musical division. That trajectory drives the **spatial engine**: an HRTF/ITD pair places the source by ear, a 1/r distance law and DOPPLER modulate a propagation delay (the pitch bend on a fly-by is real, not a pitch-shifter), and ROOM adds decorrelated early reflections that drop the interaural coherence so the sound externalises. WIDTH scales the spatial axis of the wet, MONO SAFE forces the lows to mono, and SPEAKERS mode applies RACE crosstalk cancellation so the binaural image translates over a stereo pair. A limiter with no lookahead holds the peaks, MIX blends wet against dry, and OUTPUT trims the final level. The whole chain reports its latency to the host — 424 samples (8.83 ms) at 48 kHz, constant — and the dry path is delayed to match, so any MIX setting stays phase-aligned. Chain: INPUT → IN GAIN → BINAURAL HRIR (ITD + near-field shelf) → ROOM (early reflections) → DOPPLER → OUTPUT MODE: Phones / Speakers RACE → LIMITER (0.85) → OUT GAIN → OUTPUT Highlighted: the binaural HRIR engine — the stage that puts the sound outside your head. ORBIT reports its latency to the host, so delay compensation lines every other track up with it: 390 samples at 44.1 kHz, 424 at 48, 775 at 88.2 and 844 at 96 — the count grows with the sample rate, the delay itself stays around 8.8 ms. It is constant: DOPPLER does not move it. The dry path inside the plug-in is delayed by the same amount, which is why a MIX between 0 and 100 % no longer comb-filters. #### Every control **The field** - **Orbit Radar** — The big left display — the source flying around your head, seen from above. How: Read-only visualiser. Shows the live path, the current azimuth/radius, velocity, Doppler %, height and mono status in the four corner telemetry readouts. Tip: Watch the dot to feel what SHAPE + SPEED are doing before you commit — the AZ and RAD readouts confirm exactly where the source is. **The engine rail — path & output** - **RATE** — How the orbit's timing is driven. How: Selector: 1/4 · 1/2 · 1 bar · Free · Fixed (default Free). The three musical divisions lock one full lap to the host tempo; Free runs the lap off the SPEED knob; Fixed parks the source at a static angle (no movement). Tip: Use a division for rhythmic motion glued to the groove; use Free for slow cinematic drifts; use Fixed to park a source at one spot around you and leave it there. - **FIXED ANGLE** — Parks the source at a fixed azimuth instead of orbiting — static placement. How: −180…+180° (default 0° = front). Available when RATE is set to Fixed. Tip: Nail a sound to one spot — rear-left, front-right — and automate the angle for cinematic jumps. - **SHAPE** — The geometry of the path the source travels. How: Selector: Circle · Ellipse · Spiral · Pendulum · Pendulum Back (default Ellipse). Circle is a steady lap; Ellipse squashes it by SPREAD; Spiral breathes the radius in and out (a vortex); Pendulum swings front; Pendulum Back swings through behind you. Tip: Ellipse is the most musical default. Pendulum Back is great for sounds you want to pass behind the listener's head. - **DIRECTION** — Which way the orbit turns. How: Selector: CW (clockwise) · CCW (counter-clockwise), default CW. Reverses the direction of travel; on Pendulum it flips the swing's lead side. Tip: Pick the direction that moves *away* from your busiest element so the motion stays legible in the mix. - **OUTPUT (Phones / Speakers)** — The playback target the binaural render is tuned for. How: Selector: Phones · Speakers (default Phones). Phones is the pure HRTF/ITD image. Speakers applies RACE crosstalk cancellation so the out-of-head placement survives on a stereo pair. Tip: Leave it on Phones for headphone mixes; switch to Speakers when checking on monitors so the spatialisation doesn't collapse. **Trajectory // path** - **SPEED** — Free-running orbit speed, in laps per second. How: 0.05–8 Hz (default 0.5), skewed for fine control down low. Only active when RATE is set to Free; otherwise the lap follows the host division. Tip: Under 0.3 Hz gives a wide, slow cinematic drift; push it up for tremolo-like spin. For tempo-locked motion use a RATE division instead. - **WIDTH** — The spatial axis of the wet image. How: 0–100 % (default 50). 50 % is the natural binaural width; toward 100 % the panning gets exaggerated and 'flying', built on ITD/HRTF — not a fake widener. Tip: Open it past 50 % for obvious head-circling motion; pull it back for a subtler, contained orbit. Check high settings in mono. - **RADIUS** — The base distance of the source from your head. How: 0–100 % (default 60). 0 is right at the ear, 100 is far away; it scales the 1/r direct level and feeds the externalisation, so far sources sound quieter and more 'out there'. Tip: Higher RADIUS + some ROOM is the recipe for a sound that sits clearly outside the head rather than panned across it. - **HEIGHT** — Up/down tone cue — a tilt, not a real change of height. How: −100…+100 % (default 0, bipolar). A high-shelf of up to ±6 dB above about 4.5 kHz: negative darkens the tone as a downward cue, positive brightens it as an upward cue. It is a spectral tilt, not a true elevation — the HRIR ring ORBIT places the source on is horizontal, so nothing actually moves above or below your ears. 0 leaves the tone untouched. Tip: Pair a negative HEIGHT with a Spiral and high DOPPLER for a sound that reads as descending as it rushes past — classic fly-by. - **CHAOS** — OVNI randomness — erratic speed and wobble. How: 0–100 % (default 0). Injects jitter into the orbit's velocity and a small azimuth wobble, so each lap drifts off the perfect path. At 0 the motion is clean and repeating. Tip: A little (10–25 %) makes the movement feel organic and alive; a lot turns ORBIT into an unstable, drunken flight. **Space · tone // filter** - **DOPPLER** — Real fly-by pitch shift on the orbit. How: 0–100 % (default 0). Makes the distance eccentric — close at the front, far behind — and modulates a propagation delay, so pitch rises on approach and falls on retreat. It does not change the latency ORBIT reports: that stays constant whatever DOPPLER is doing (measured error 0.00 samples at 0, 25, 50 and 100 %), and the dry path is delayed to match. Bypassed at 0. Tip: A little (20–40 %) adds life; high settings get cinematic whoosh. The wobble is physics from the moving delay, not a pitch-shifter — lower SPEED or DOPPLER to tame it. - **ROOM** — Decorrelated early reflections for externalisation. How: 0–100 % (default 30). Adds short, ear-independent (velvet-noise) reflections that lower the interaural coherence — the cue that pulls a sound out of your head. Not a reverb tail, and it adds no latency of its own — though ORBIT does declare a 0.104 s tail to the host, so a bounce keeps its last reflections instead of being cut short. Tip: Even 20–40 % is enough to externalise. Push higher for a sense of space, but it stays an early-reflection field, not a long reverb. - **SPREAD** — Eccentricity of the elliptical path. How: 0–100 % (default 35). On Ellipse it sets how squashed the orbit is (0 = near-circle, 100 = a long thin lap, eccentricity up to 0.9). Shapes the front/back vs left/right balance of the travel. Tip: Raise SPREAD on Ellipse for a wide left-right sweep with a fast front/back cross; drop it for an even circular feel. **The gain bar** - **MIX** — Dry/wet blend of the spatialised signal. How: 0–100 % (default 100), bottom rail fader. At 100 % you hear only the moving binaural source; lower it to keep a dry anchor in the centre. Tip: Keep 100 % on a send/aux; on an insert pull back to 40–70 % to keep the dry punch while the wet orbits around it — the dry is delayed to match the wet, so a partial blend stays phase-aligned instead of comb-filtering. - **INPUT** — Pre-engine input trim (drive). How: −24…+24 dB (default 0), bottom rail fader. Stages the level going into the spatial engine before any processing. Tip: Trim hot signals down a touch so the internal limiter isn't constantly catching peaks on fast, close fly-bys. - **OUTPUT** — Final output level trim. How: −24…+24 dB (default 0), bottom rail fader, post-limiter. Sets the level leaving the plugin; the clip LED warns if you push past 0 dBFS. Tip: Use it to match the wet back to your dry reference; watch the clip indicator if you raise it on dense material. - **IN PHASE / MONO SAFE** — Forces the low end to mono for translation. How: Toggle (default off), on the bottom rail. Collapses the bass to a mono-compatible path so nothing cancels when the mix is summed. Tip: Turn it on for anything that has to survive a club PA, a phone speaker or a mono broadcast — it keeps the low end solid while the highs still orbit. #### Tips & tricks - On a **send/aux bus**, keep MIX at 100 % and automate SHAPE/SPEED — you get a moving spatial layer without touching the dry track. - For tempo-locked motion, set RATE to a division (**1 bar** reads as obvious circling) instead of Free. - The externalisation recipe: higher **RADIUS** + a little **ROOM** + WIDTH around 50–70 % puts the source clearly outside the head. - **Fixed** RATE turns ORBIT into a placement tool — park a backing vocal or pad just behind the listener, off to one side, and leave it there. Not above: the HRIR ring is horizontal, and HEIGHT is a tonal cue, not a move. - Build a riser with a **Spiral**, a negative **HEIGHT** and DOPPLER climbing to 80 % — the sound reads as descending and rushes past at the drop. - A small **CHAOS** (10–25 %) keeps repeated laps from sounding mechanical; great under long sustained sounds. #### Troubleshooting - **I updated to 0.3.0 and my DAW still shows the old ORBIT — or shows two of them.** From 0.3.0 the plug-ins are called **OVNI ORBIT.vst3** and **OVNI ORBIT.component**: macOS matches plug-in file names without regard for case, so our *ORBIT.component* and another maker's *Orbit.component* were the same file and overwrote each other. On **macOS** the installer retires the old bundle for you — only if it identifies itself as ours (*com.ovni.orbit*) — and moves it, never deletes it, to *~/Library/Application Support/OVNI Audio/replaced//*; run *killall -9 AudioComponentRegistrar* and rescan for your DAW to notice. On **Windows** nothing is removed for you: delete *C:\Program Files\Common Files\VST3\ORBIT.vst3* by hand before copying the new folder in, or the DAW sees two plug-ins claiming the same identity. Saved projects keep working either way — the Audio Unit identity and the VST3 unique ID did not change, only the file name. - **It still sounds like it's inside my head, not around me.** Externalisation needs distance and decorrelated reflections. Raise **RADIUS** and add some **ROOM** (20–40 %), and listen on headphones. WIDTH alone only pans — it's the 1/r distance plus ROOM that pushes the source out of the head. - **The spatial effect collapses on my monitors / phone.** The pure binaural image is tuned for headphones. Switch **OUTPUT** to **Speakers** for crosstalk cancellation on a stereo pair, and turn on **MONO SAFE** so the low end stays mono-compatible when summed. - **I hear a pitch wobble on fast orbits — is that a bug?** No, that's the **real Doppler**. Pitch emerges from modulating the propagation delay as the source flies toward and away from you. Lower **DOPPLER** or **SPEED** to reduce it; it's physics, not a pitch-shifter. - **The pan isn't following my tempo.** Set **RATE** to a musical division (1/4 · 1/2 · 1 bar). On **Free** the lap runs off the SPEED knob in Hz and ignores the host; on **Fixed** the source doesn't move at all. #### Specs - Formats: VST3 · AU - Systems: macOS · Windows - Build: Universal (Apple Silicon + Intel) - Engine: binaural (HRTF + ITD) · lightweight - Doppler: Real, delay-line based - Latency: 424 samples (8.83 ms) at 48 kHz · reported · constant - Output: Headphones · Speakers (RACE crosstalk) - Presets: 40 factory - License: AGPLv3 · open-source ### SUPERNOVA — Manual https://ovniaudio.com/supernova-manual #### What SUPERNOVA is SUPERNOVA is an **audio-reactive visual synthesizer** — a VJ instrument for macOS. You drag in an image or a video and a GPU engine samples it with **262,144 particles** (a 512×512 grid) that move, glow and burst with your sound — it breathes on the pads, electrifies on the hats, **explodes on the kick** and re-forms. It runs both as a **VST3 / AU plugin** inside your DAW and as a **standalone macOS app**, and the audio it reads is a **bit-exact pass-through** — a read-only tap that never colours your sound. It shines on live streams and VJ sets, on music videos and cover art, on a second monitor at a party, or just as a beautiful listening companion. Feed it your own photos, a logo or a loop — anything with an image becomes a living, sound-driven visual you can play, record and route into your show. For the standalone app and its driver-free system-audio capture, see the companion **App manual**. #### Quickstart 1. As a plugin: drop SUPERNOVA on any track or bus — its audio is bit-exact pass-through, so it's even safe on the master. As the app: just open it (see the App manual for audio setup). 2. Drag an image, a folder of images (a **photo sequence**), or a **video** onto the window. The particles take its shape and colour. 3. Open the **Worlds** browser and pick one of the **50 presets** — each one swaps the physics, geometry and colour at once. 4. Raise **INTENSITY** until the field feels alive with the music. It's the master "life" fader — the one knob that drives everything. 5. Send the picture where you need it: a window, fullscreen on any display, into OBS / Resolume / VDMX via **Syphon**, or record a clip straight to **MP4 with sound**. #### Signal flow SUPERNOVA taps your audio **read-only** and runs a lightweight analysis every block — onsets (the kick), frequency bands and overall RMS. That analysis advects **262,144 particles** (a 512×512 grid) on the GPU: each world defines a force field (flow, vortices, gravity, radial blast) that pushes the particles, which sample the colour of your image or video. The field is composited in **linear additive HDR** and resolved through an **ACES** filmic tonemap, so highlights bloom and roll off like film instead of clipping to a flat white. The final frame goes to the on-screen view, any fullscreen display, and — if enabled — a **Syphon** server for OBS / Resolume / VDMX; it can also be recorded to an MP4 with an in-sync audio track. The audio itself passes straight through, sample for sample: out equals in. Chain: AUDIO IN (read-only tap) → ANALYSIS (onsets · bands · RMS) → GPU PARTICLE SIM (262K) → HDR LINEAR ADDITIVE → ACES FILMIC TONEMAP → BLOOM + CINEMATIC OPTICS → DISPLAY · FULLSCREEN · SYPHON This is the picture pipeline, not an audio one: the sound only enters as a read-only tap and leaves untouched — out equals in, sample for sample, at 0 samples of latency. #### Every control **Media, worlds & the master fader** - **Source (Load)** — What the particles are made of. How: Drag a single image, a folder of images (a timed photo sequence), or a video file onto the window. The particles sample its colour; a video plays and re-samples live. Tip: High-contrast subjects on a plain background cut out best (see Cutout). A slow video loop makes an endless evolving backdrop. - **Worlds (Preset)** — The 50 curated worlds — the factory presets. How: Each world is a whole scene: it swaps the physics (how particles move), the geometry and the colour look in one click. Browse the grid and step next / previous. Tip: Treat a world as a starting point — pick the closest one, then push INTENSITY and the four domains until it's yours. Five of them — **Satellite, Parallax, Lantern, Meridian and Vantage** — share one idea: your picture stays your picture. They add depth, parallax, a slow orbit, a cut-out or a light, but they never mirror it or break it apart. Reach for those when the image itself is the point. - **INTENSITY** — The master "life" of the whole field. How: 0–100 %. Scales how hard the entire system reacts to the audio — from a calm shimmer to a full storm. The single most important knob. Tip: Automate it (or MIDI-learn it to a fader / mod wheel) to make the visual build and drop with the track. **The MEDIA session — cue, reorder, relink** - **The MEDIA strip** — Your whole session, in order, under the picture. How: As soon as media is loaded a strip of **numbered thumbnails** appears between the visual and the knobs — every photo and video of the session, the current one outlined with a small bar counting down to the next change. **Click** a tile to cue it (**←** / **→** step through it, and the number keys **1**–**9** / **0** jump straight to one), **drag** to reorder, **✕** to remove. Dropping files or a folder *onto the strip* inserts them where you drop them; **▦** in the top bar hides it. Tip: Removing, reordering, rotating and clearing are all undoable with **Cmd + Z** — and the session is saved with the project and with your user presets, so reopening brings it back. - **Canvas FORMAT · FIT / FILL** — The shape of the canvas, like a project format. How: **AUTO** follows your **first** item — a vertical photo gives **9:16**, a square **1:1**, a landscape **16:9** — or pick FREE / 16:9 / 9:16 / 1:1 / 4:5 / 4:3. The same menu chooses **FIT** (the whole image, letterboxed) or **FILL** (cropped to fill). The editor, fullscreen output, the Syphon feed and the EXPORT default all follow it. Tip: Shooting for Reels? Set 9:16 first — EXPORT then lists Vertical at the top, marked *canvas*. - **SEQ — clock, order & transition** — How the sequence moves on. How: **Seconds** (the classic slideshow), **Beats** — 1 · 2 · 4 · 8 · 16 · 32 · 64, changes landing on the bar — or **Kick**: a new photo on every kick the analysis hears, never faster than a minimum gap (0.25–8 s). Order is **Loop** or **Shuffle**; the transition is **Dissolve** or **Burst**. **Dissolve** is always on and has nothing to set: every photo melts into the next, however the change is fired — the clock, a click, a number key, an arrow, a MIDI note or dropping a photo on top of another one — and the exported clip dissolves the same way. Its length comes from the clock itself (45 % of the interval, capped at 0.7 s and never under 0.1 s), so at the usual pace it takes about **0.7 s** and on Kick with the minimum 0.25 s gap it closes in **0.11 s**, before the next kick. **Burst** explodes the old photo first, and what blows apart re-forms already wearing the new one's colours. Two things still cut, on purpose: CLEAR, and opening a session onto an empty canvas — there is no picture to melt out of. Tip: In **Beats** a cue waits for the next bar — the armed tile shows a dashed outline — so the change lands *on* the beat. Hold **Shift** to fire it right now. - **Missing media & Relink** — A file that moved is marked, not dropped. How: If a photo or video isn't where it was, its tile stays in the session marked **missing**: a red **!** instead of the thumbnail, the name greyed out. Nothing cuts to it — the sequence, **← / →**, the number keys and the MIDI cues all step over it. Right-click for **Relink…** (that one file) or **Relink folder…**, which finds every missing item that used to live beside it in one go. Tip: If the file simply comes back to where it was, the tile heals itself the next time the session refreshes. **Movement — how the field moves** - **MOVEMENT domain** — How the particles move — the first of the four domains. How: Motion picks the flow field (Contours · Matter · Wave · Vortices · Radial); Chaos, Curl Scale, Home Pull, Gravity and Momentum shape the forces; Speed time-warps the whole simulation (50 % = ×1). Tip: Vortices + a little Chaos gives a galactic swirl; Radial + Home Pull snaps the image back together between hits. - **Blast** — The kick response — a radial shove on every onset. How: 0–100 %. When an onset is detected the particles are pushed outward, then the field re-forms. This is the "explode on the kick" behaviour. Tip: Pair a high Blast with Home Pull so it detonates and re-gathers cleanly on every beat. - **Breathe & Pump** — Slow RMS breathing and a visual sidechain. How: Breathe swells the whole field with the loudness envelope; Pump ducks it on transients like a sidechained compressor (≈30 % is the classic feel). Tip: A touch of Pump locked to the kick gives that pumping, club-visual throb without any MIDI. **Matter — what the particles are** - **MATTER domain** — What each particle looks like — the second domain. How: Size and Glow set the grain and bloom; Density draws more or fewer of the 262k; Scatter morphs from a crisp image to a loose cloud; Shimmer adds sparkle. Tip: Low Density with big Size reads as chunky embers; high Density with small Size reads as fine dust. - **Shape (Glyphs)** — The sprite each particle is drawn with. How: Choose from Dot · Disc · Ring · Dash · Tri · Quad · Spark — SDF glyphs rendered crisp at any size. Tip: Dash + Trails gives calligraphic streaks; Ring reads as bokeh; Spark reads as fireworks. - **Trails** — Velocity streaks behind each particle. How: 0–100 %. Feeds a fraction of the previous frame back, so fast particles smear into light-paths. Tip: High Trails with low Density turns motion into flowing ribbons of light. - **Links (Plexus)** — GPU nearest-neighbour connections. How: 0–100 %. Draws lines between particles that are close together — a live constellation / network mesh that reacts as the field moves. Tip: Keep it subtle on dense fields; on a sparse field it becomes a clear plexus web. - **Cutout (3D)** — Isolate the subject from its background — as a live knob. How: 0–100 %. Uses Vision saliency + subject cutout to blend between the full image and a mask of the subject only, so the background falls away and the subject floats — a 3D cutout you can rotate in particle-space. Tip: Best on a photo with a clean, flat background; combine with Depth to sculpt and spin the subject in space. **Camera — where you watch from** - **CAMERA domain (3D)** — The viewpoint on the field — the third domain. How: Depth lifts the field into real 3D volume (order-independent additive, no z-sort); Rot X / Rot Y tilt it; Rotate and Orbit spin the view automatically; Kaleido mirrors it into 2/4/6/8 segments. Tip: A slow Orbit with some Depth gives a hands-off, always-moving 3D sculpture — perfect for a stream. - **Figure & Form** — Morph the image into a geometric body. How: Figure picks a target shape — Image · Sphere · Spiral · Rings · Grid · Helix — and Form fades from the flat image (0 %) to that 3D geometry (100 %). Tip: Ride Form up on a drop so the picture collapses into a spinning sphere or helix, then back. **Colour — the look** - **COLOR domain — Color Lab** — The look ramp (Palette) and how it's applied — the fourth domain. How: Palette selects one of 17 curated looks baked in OKLab (Thermal, Magma, Vaporwave, Black Gold…); Color Amount mixes the gradient-map in; Background sets the look's paper. Tip: Palette 0 (Original) keeps your image's own colours; the curated ramps re-grade the whole scene by luminance. - **Saturation / Hue / Hue Cycle** — Global colour trims. How: Saturation 0–100 % (50 = neutral); Hue rotates the whole palette; Hue Cycle drifts the hue slowly over time. Tip: A gentle Hue Cycle keeps a long stream visually evolving without touching anything. - **Variation** — Curated, seeded randomisation. How: A master Variation plus per-domain Movement / Matter / Camera / Color Var nudge a world into a fresh but still-musical variant — deterministic from the preset's seed. Tip: Turn Variation up a touch to make a familiar world feel new, without going off the rails. **LFOs — four modulators, recalculated every frame** - **LFO card — ON · TARGET · RATE · SHAPE · DEPTH** — One card per LFO, four in all. Open the panel with the LFO button. How: **ON** enables the slot and the target starts moving. **TARGET** picks one of **12 parameters** — hue, saturation, intensity, size, glow, chaos, rotate, orbit, scatter, trails, links, depth. **RATE** under **SYNC** offers every division from a sixteenth to four bars, each with its **triplet** (·T) and **dotted** (·D) — **21 in all**; under **FREE** it becomes **Hz** (0.05–20) and runs off the clock instead of the tempo, so it keeps moving with the transport stopped or with no host. **SHAPE** is a strip of **6 icons that draw their own curve**: Sine, Triangle, Saw, Square, Ramp Down and Sample & Hold. Tip: At **100 % DEPTH** an LFO sweeps exactly the target's range and is clamped to it — several LFOs on the same parameter can no longer push it out of bounds. - **BI / UNI · PHASE · RETRIG** — The second line of the card: how it swings. How: **BI** swings both ways around where you left the knob; **UNI** only adds, from the knob upward. **PHASE** offsets where the cycle starts, **0–360°** — set two LFOs apart and they chase each other instead of pulsing together. **RETRIG** restarts the cycle *now*, on the current beat, without moving your PHASE setting. All three are saved with the project. Tip: RETRIG is how you line an LFO up with a drop by hand — and the starting point when there is no host transport to lock to. - **Live output meter** — What each modulator is doing, right now. How: Every card carries a meter that pulses with the LFO's value in real time. Synced LFOs follow the BeatClock — the host tempo inside a DAW, the **TAP** button in the app — and either way the modulation is recalculated **on every rendered frame**, at your display's refresh rate. Tip: That per-frame recalculation is why a fast LFO reads as a wave and not a staircase. Watch the meter to time a peak to the drop. **Scenes, MIDI & output** - **Scenes & User Presets** — Save, recall and undo your own looks. How: Store the full state as a user preset (save / recall) and switch between scenes on the fly. Every move is undoable — full Undo / Redo — so you can push a world hard and step back. Tip: Build a few scenes for a set and jump between them live; nudge Variation inside a scene to keep it fresh without losing the look. - **MIDI (Learn · CC · Triggers)** — Play and automate the visuals from a controller. How: Right-click any knob to MIDI-learn it, then drive it from a CC. Incoming notes fire events — a note detonates the field (Explode) or launches a directional ray whose angle follows the pitch. Works in the plugin (needs MIDI input) and the app. Notes **72–87** cue tiles 1–16 of the media session, **88 / 89** step next / previous and **90** jumps to a random photo (never the one on screen) — so the session stays playable in immersive or fullscreen, where the strip isn't visible. Cues follow the sequence clock; **MIDI cue: now** in the SEQ menu makes every note cut immediately. Tip: Learn INTENSITY to a fader and map a pad to Explode — you VJ the whole field by hand, perfectly in time. - **Output & Syphon** — Where the picture goes. How: The on-screen view, plus fullscreen on any connected display, plus a Syphon server that hands the live frame to OBS, Resolume or VDMX with no capture card. macOS only. Tip: Syphon is zero-copy on the GPU — cleaner and lighter than screen-capturing the window. - **Export (MP4)** — Record the visuals to a video file. How: Renders an MP4 / H.264 clip with four aspect / format presets, muxed **with sound** — a **30 s** real-time audio ring buffer is written to an in-sync AAC track. Choose any duration from 1 to 600 s — but the clip is driven by that same 30 s window, so anything longer repeats it, picture and sound together — and the repeat is crossfaded over 10 ms, so the loop point no longer ticks. **The clip plays back the window**: the same physics (measured in seconds, not in frames of your display), the same exposure at 4K as at 1080p, your photo decoded the way the window decodes it, and the photo changes dissolving as they do on screen. It starts already in motion and keeps the tilt and colour drift ORBIT and HUE CYC have accumulated. What does *not* travel: the knob values are frozen the moment you press EXPORT, so an LFO or a preset morph in flight is not baked in. macOS only (AVFoundation). Tip: Set a custom duration and hit export at the drop — you get the picture and the audio locked together, ready to post. Park the knobs where you want them first: what is on screen when you press EXPORT is what the clip renders. - **Clear** — Reset the field. How: Wipes the particles and trail buffer back to the current image's rest state — a clean slate without reloading. Tip: Hit Clear after a heavy Trails/Blast passage to instantly settle the picture. #### Tips & tricks - As a plugin, put SUPERNOVA on your **master bus** — the audio is bit-exact pass-through, so it sees the whole mix without changing a sample of it. - Load a **photo sequence** (a numbered folder) to tell a story: the images advance on a timer while the particles keep reacting. - For a hands-off stream, set a slow **Orbit** + **Hue Cycle** (or park an **LFO** on Hue) and leave it — the scene keeps evolving on its own. - Big **Blast** + strong **Home Pull** is the money move: it explodes on the kick and snaps the image back together in the gaps. - Use **Cutout** on a portrait with a plain backdrop, then push **Depth** and a slow rotation — the subject lifts off the background into 3D particle-space. - Building a set? Save a few **scenes**, MIDI-learn INTENSITY and a pad to Explode, and jump between looks live — every move is undoable. - Caught a great take? **Export** an MP4 with sound at a custom length — the 30 s audio buffer keeps the clip perfectly in sync for socials, and a 30–45 s piece no longer brings the same bar back. #### Troubleshooting - **It's not reacting to the audio.** As a plugin, make sure signal is actually passing through SUPERNOVA (it's an insert on a playing track) and INTENSITY is up. As the app, you need to select an input and grant permission — see the **App manual** for system-audio capture and its one-time macOS permission — **System Audio Recording** on macOS 14.2+, the older Screen Recording switch on 13–14.1. Either way it is how macOS gates system audio; it is not the microphone. - **Everything blows out to flat white.** Pull **INTENSITY**, **Glow** or **Blast** down. SUPERNOVA composites in HDR with an ACES tonemap, but very high settings still push the core past the roll-off. Lower Glow first for the biggest change. - **The subject won't separate from its background.** Cutout works from a subject mask (Vision saliency), which needs a reasonably flat, contrasting background. On busy photos the mask is weak — lower **Cutout** and lean on **Scatter** and **Depth** for separation instead. - **My kick isn't triggering the explosion.** Raise **Blast** and make sure the onset is landing — a very compressed or quiet source gives weak onsets. You can also MIDI-learn a pad to **Explode** and fire it by hand, perfectly in time. - **Syphon isn't showing up in OBS / Resolume.** Enable the **Syphon** output in SUPERNOVA, then add a Syphon Client source in OBS / Resolume / VDMX and pick SUPERNOVA's server. Syphon is macOS-only and hands the frame straight over the GPU — no capture card, no screen recording. - **Does it change my sound?** No. The audio is a **read-only, bit-exact pass-through** — out equals in, sample for sample. SUPERNOVA only analyses the signal to drive the picture. #### Specs - Formats: VST3 · AU · Standalone - Systems: macOS 11+ · Apple Metal - Build: Universal (Apple Silicon + Intel) - Engine: 262,144 GPU particles (512×512) · HDR linear + ACES - Worlds: 50 presets (physics + geometry + colour) - Color Lab: 17 OKLab look ramps - Audio: bit-exact pass-through (read-only tap) - App capture: system audio, driver-free — Core Audio taps (macOS 14.2+) · ScreenCaptureKit (13–14.1) - Output: window · fullscreen · Syphon → OBS / Resolume / VDMX - Export: MP4 / H.264 with sound · 1–600 s from a 30 s window (AVFoundation) - Modulation: 4 LFOs · 21 beat divisions (triplet + dotted) · free Hz · 12 targets - MIDI: MIDI-learn · CC · note triggers (explosion · directional ray) - License: AGPLv3 · open-source ### SUPERNOVA App — Manual https://ovniaudio.com/supernova-app-manual #### The standalone app SUPERNOVA is an **audio-reactive visual instrument**, and it ships two ways: as a **VST3 / AU plug-in** inside your DAW and as a **standalone macOS app**. This manual covers what's unique to the app — above all its headline trick: it can capture your Mac's **system audio with zero drivers** and react to **anything** playing on the machine — Spotify, YouTube, Serato, Ableton — with no BlackHole and no routing. Everything about the visuals themselves — all **50 worlds**, the four domains (Movement · Matter · Camera · Color), the media drop, the 3D cutout, MIDI and export — behaves exactly as in the **plug-in manual**; this page is only the app-specific layer. The app is how you run SUPERNOVA outside a DAW: for streaming, VJ sets, listening parties, a second-screen visualiser, or reacting to a set you're playing off the decks. Because it listens to the whole machine, it doesn't care which app the sound comes from — point it at anything and the field moves. And because it only ever **analyses** the audio, it sits happily beside your DJ software or DAW with no routing and no doubled sound. #### Quickstart 1. Open the SUPERNOVA app — nothing pops up at you. It starts, the visual runs, and the moment it needs to hear the Mac the top bar shows a notice: **⚠ System Audio needs permission**, with an **ALLOW** button. 2. Click **ALLOW** and *then* macOS asks — so you see what is being requested before you answer. Say yes and the meter starts moving; macOS only asks once. If you said no, the notice becomes **OPEN**, which jumps to the exact pane — **System Audio Recording** on macOS 14.2+, **Screen Recording** on 13–14.1. Turn SUPERNOVA on there and the notice normally clears by itself within a couple of seconds; if it doesn't, **REOPEN** relaunches this copy so the grant takes effect. 3. Pick your **SOURCE**: **System Audio** (whatever is playing on this Mac, no drivers) or a mic / audio-interface input (choose the device in Setup). 4. Play something and raise **INTENSITY** — the field ignites. The app only **analyses** the audio to drive the picture; it never plays it back, so there's no double sound and no feedback loop. 5. Go big: **Fullscreen** (F) on a second display, **Immersive** (Tab) to hide the chrome, and/or publish the live frame over **Syphon** straight into OBS, Resolume or VDMX. #### Signal flow In a DAW, the plug-in reads the audio on its track or bus and the MIDI from the host — and, as a plug-in, it is a **bit-exact pass-through**: out equals in, sample for sample, so the sound is never touched. In the standalone app, **you** choose where it listens. With SOURCE set to **System Audio**, the app taps the Mac's output mix through Core Audio **process taps** (macOS 14.2+) — which is exactly why macOS gates it behind the **System Audio Recording** permission, asked the moment you pick that source. On macOS 13–14.1 those taps don't exist yet, so the app falls back to Apple's **ScreenCaptureKit** and macOS gates it behind the older **Screen Recording** switch instead. With SOURCE set to an **input device**, it opens your chosen mic or interface instead. Either way the audio is only **analysed**, never played back. That analysis drives the same **Metal** engine as the plug-in — 262,144 GPU particles on a 512×512 grid, 17 colour look-ramps baked in OKLab, HDR linear additive light resolved through an **ACES filmic tonemap** — and the finished frame is rendered to the window, to a fullscreen display, and to a **Syphon** server for OBS, Resolume or VDMX. Chain: macOS · SYSTEM AUDIO RECORDING → ALLOW (asked when you pick the source) → CORE AUDIO TAP · 14.2+ (ScreenCaptureKit on 13-14.1) → SYSTEM AUDIO (the Mac's output mix) → ANALYSIS ONLY (never played back) → METAL ENGINE (262K particles) → WINDOW · FULLSCREEN · SYPHON That macOS permission is how the system authorises system-audio capture — it is NOT the microphone, and nothing is recorded. It is asked when you pick System Audio, not when the app opens, and macOS only asks once. On macOS 13–14.1 the same capture goes through ScreenCaptureKit and its older Screen Recording switch. The audio is only analysed, never played back: no double sound and no feedback loop. #### Every control **Hearing your Mac** - **SOURCE** — Where the app listens for audio. How: Two modes. **System Audio** taps whatever is playing on this Mac (Spotify, YouTube, Serato, Ableton) with no drivers; **input device** opens a mic or audio interface for a live source. Pick the device itself in Setup. Tip: System Audio is the zero-setup path — no BlackHole, no aggregate device. Reach for an input device only when you want a specific mic or line-in. - **The System Audio permission** — The one-time macOS grant that unlocks system-audio capture — asked when you pick the source, not when you open the app. How: The top bar runs it: **ALLOW** → macOS asks → the meter moves. Say no and the notice offers **OPEN**, which jumps straight to the right pane in System Settings → Privacy & Security; turn SUPERNOVA on and it normally clears by itself. macOS asks once — every later launch re-checks in silence. Tip: On **macOS 14.2+** the pane is **System Audio Recording**: audio, nothing else. On macOS 13–14.1 that API doesn't exist yet, so the app falls back to ScreenCaptureKit and macOS shows the older **Screen Recording** switch instead — same capture, same privacy, simply the only route those versions offer. It is never the microphone, and nothing is recorded. System Audio needs macOS 13+; on 11–12 use an input device. - **REOPEN** — The fallback for when the notice won't clear. How: Once you grant the permission the notice usually clears on its own within a couple of seconds. When it doesn't, REOPEN relaunches *this* copy of SUPERNOVA — the exact bundle on disk, as a new instance — so the fresh grant takes effect. Tip: You rarely need it now: the app re-checks by itself, and REOPEN no longer hands the job to LaunchServices, which could reopen a *different* installed copy of SUPERNOVA. - **Setup (device & MIDI)** — Choose your input device and MIDI port. How: Opens the audio/MIDI picker: select a mic or interface for input-device mode, and a MIDI input for the note triggers and CC. The app produces no audio output — it only analyses. Tip: You can stay on System Audio and still pick a MIDI input here, so a controller can fire the note triggers — a note explodes the field or launches a directional ray. - **Visual sensitivity (Gain)** — Input trim into the analysis — how hard the audio drives the picture. How: In dB. Raises or lowers the level the engine sees before it reacts. It changes nothing you hear (there is no audio output), only how energetically the field responds. Tip: Turn it up for quiet material; back it off if a hot master keeps the field pinned at maximum. **The window** - **Immersive (Tab)** — Visuals-only mode. How: Press Tab to hide the entire top bar and the knobs, leaving just the picture in the window. Tab again brings the chrome back. Tip: Great for a clean second-screen look when you don't want to commit the whole display to fullscreen. - **Fullscreen (F)** — Throw the picture fullscreen to a display. How: Press F (or the fullscreen button) to fill a connected monitor; Esc returns to the window. Works on any attached display. Tip: Keep the controls on the laptop screen and send fullscreen to the projector or TV. - **Always on top** — Keep the SUPERNOVA window above everything else. How: Toggle so the window floats over your DJ software or browser — handy when you only have one screen. Tip: Combine it with Immersive for a tidy floating visualiser parked in a corner. **Out to the world** - **Syphon** — Publish the live frame to other apps — the clean way to stream or VJ. How: Toggle the **Syphon** server on and SUPERNOVA appears as a video source inside OBS, Resolume or VDMX — GPU-to-GPU on macOS, no capture card and no window grabbing. Tip: In OBS, add a 'Syphon Client' source and pick SUPERNOVA. It's lighter and sharper than screen-capturing the window. - **Export (record)** — Record what you're seeing to a video file — with sound. How: Writes an **MP4 / H.264** in one of four aspect/format presets, and muxes a **real-time audio track** (a **30 s** ring buffer bounced to an in-sync AAC track) so the clip carries the music. Any length from **1 to 600 s** — but that 30 s window is what drives the clip, so anything longer repeats it, picture and sound together — and the repeat is crossfaded over 10 ms, so the loop point no longer ticks. **What comes out is what the window shows**: same physics, same exposure at 4K as at 1080p, same reading of your photo, same dissolves. The one thing that does not travel is a knob in motion — the values are frozen when you press EXPORT, so an LFO or a preset morph in flight is not baked in. Tip: The audio comes from the same signal the visuals are reacting to, so picture and sound stay locked. macOS only (AVFoundation). **Media controls** - **Load · Rotate · Sequence · Clear** — The media controls, shared with the plug-in, on the top bar. How: Load a photo or video and the particles **form it** (Vision saliency finds the subject and cuts it out); multi-select builds a photo sequence; rotate the current media 90°; set the seconds-per-photo for a sequence; CLEAR returns to a bare canvas. You can also just drag media onto the visual. Tip: Everything else — the 50 worlds, INTENSITY, the four domains, the rotatable 3D cutout, tempo-synced LFOs with their live meter, MIDI-learn and presets — is identical to the plug-in. See the SUPERNOVA plug-in manual. #### Tips & tricks - Streaming? Skip window-capture entirely — turn on **Syphon** and add SUPERNOVA as a native source in OBS. It's cleaner, sharper and lighter on the GPU. - One laptop, one projector: keep the knobs on the laptop and press **F** to throw the visual fullscreen to the projector; **Esc** brings it back. - The app never plays audio, so you can run it **alongside** your DJ software or DAW on System Audio — no routing, no aggregate device, no doubled sound. - On a single screen, **Always on top** + **Immersive** (Tab) makes a neat floating visualiser you can park in a corner while you work. - Capture the moment with **Export**: an MP4 / H.264 *with sound* (a real-time audio buffer muxed to an in-sync AAC track), any length from 1 to 600 seconds (longer than 30 s and the take loops, crossfaded so the join does not tick) — perfect for a clip to post. - Wire a MIDI controller in Setup and keep it on System Audio: a note **explodes** the field or fires a directional ray while the music itself keeps driving the motion. #### Troubleshooting - **I enabled the permission but nothing is captured.** Give it a couple of seconds — the app re-checks on its own and the notice usually clears. If it doesn't, hit **REOPEN**: it relaunches this copy so the fresh grant takes effect. Still black? Confirm SUPERNOVA is ticked under System Settings → Privacy & Security → **System Audio Recording** (macOS 14.2+) or **Screen Recording** (macOS 13–14.1). - **There's no permission prompt at all.** Opening the app asks for nothing — the prompt appears when you pick **System Audio** and click **ALLOW**. If the notice reads *Waiting for macOS permission…*, macOS is showing you its own dialog: answer that one. If nothing appears at all, use **OPEN** and tick SUPERNOVA by hand in the pane it takes you to. - **System Audio is greyed out or silent even after I grant the permission.** Driver-free system-audio capture needs **macOS 13 or newer** — Core Audio taps on 14.2+, ScreenCaptureKit on 13–14.1. SUPERNOVA itself runs on macOS 11+, but on 11–12 there's no system tap — switch SOURCE to a mic or audio interface instead. Since **0.3.2** the top bar says so outright — *System Audio needs macOS 13 or later on this Mac — use an input device or MIDI* — with no button to press, because there is nothing that pressing it could do. - **I plugged in headphones and the visual went deaf.** Not any more: SUPERNOVA notices an output-device or sample-rate change and rebuilds its capture on the new output in about a second, without asking for permission again. The same self-repair covers any other way the capture dies — and if you switch the permission **off** while the app is open, the notice comes back instead of the app pretending to listen. - **It's reacting to my mic, not to Spotify.** Set **SOURCE** to **System Audio**. Input-device mode listens to a mic or interface; System Audio mode taps whatever is playing on the Mac. - **Will I hear the audio twice?** No. The app only **analyses** the audio to drive the picture — it produces no audio output, so there's no doubling and no feedback loop, even on System Audio. #### Specs - Format: Standalone app — macOS 11+ (Universal: Apple Silicon + Intel) - Also ships as: VST3 · AU plug-in - System audio: driver-free capture — Core Audio taps (macOS 14.2+) · ScreenCaptureKit (13–14.1) - Permission: System Audio Recording — asked when you pick the source (Screen Recording on macOS 13–14.1) - Sources: System Audio · mic / audio interface - Output: window · fullscreen · Immersive (Tab) · always-on-top - Share: Syphon → OBS · Resolume · VDMX (macOS) - Export: MP4 / H.264 with sound · 1–600 s from a 30 s window (macOS) - Engine: Metal · 262,144 GPU particles · 17 OKLab ramps · ACES tonemap - Worlds: 50 factory presets - Audio: analysis-only — never plays your sound - License: AGPLv3 · open-source ### TELESCOPE — Manual https://ovniaudio.com/telescope-manual #### What TELESCOPE is Thirteen **lenses** over one analysis engine, plus a rules engine that turns what it measured into sentences — each one carrying the number that produced it and the id of the rule that fired. And it **does not touch your audio**: TELESCOPE reads L and R, pushes them to its own analysis thread and writes nothing back — 0 samples of latency, 0 tail, bit-exact output. That is not a claim but a test: deterministic stereo noise through the plug-in with bypass on and off, with each of the thirteen lenses selected, in blocks of 1, 7, 64 and 4096 samples and from a mono source — zero mismatches in all five cases. Analysis runs on its own thread: the audio never waits for a measurement. It does not localise sources — FIELD and POLAR LEVEL show energy by pan direction, **not localisation**: no HRTF, no ITD, no azimuth. It is stereo only, with no speech-intelligibility metric. It runs with no model, no network and no telemetry: VERDICT is a deterministic table of thresholds — **measurement, not taste** — and **no findings is not the same as finished**. #### Quickstart 1. Double-click **OVNI-TELESCOPE-v0.1.0.pkg**. macOS's own Installer puts VST3 + AU where your DAW looks for them and asks for your admin password itself. The package is signed with a Developer ID and notarized by Apple, so nothing lands in quarantine. macOS 11 or newer, Apple Silicon or Intel. 2. Insert TELESCOPE on the track or bus you want to read, wherever it sits in the chain: the audio comes out bit-exact, with 0 samples of latency. 3. Pick a lens on the **strip on the left**. Only the modules the open lens needs are computed — except loudness, which is always on, so the integrated value, LRA, histogram and clip count never have a hole. 4. **Hover** any plot for a readout, always from the data and never from the pixel colour — where there is no data there is no readout. **S · M · L** resize the window, and **LANGUAGE**, at the foot of the strip, switches the whole plug-in. 5. Open **VERDICT** and read the headline first: the count of what was measured, and where the first thing to look at starts. Re-run the analysis after every change. #### Signal flow Inside the plug-in, TELESCOPE reads L and R, pushes them to a **lock-free bus** and writes nothing back, so it **never touches your audio**. An **analysis thread** drains that bus and computes only the modules the open lens needs (loudness, always), publishing **frames** for the lenses to draw and a **per-second history** — 10 minutes at 1 Hz since the last RESET — that VERDICT reads to say *where*. File analysis runs the same classes offline and comes out equal to the bit. Chain: AUDIO IN (read-only tap) → LOCK-FREE BUS → ANALYSIS THREAD → FRAMES + PER-SECOND HISTORY → 12 LENSES SHOW → VERDICT SAYS (21 RULES) Highlighted: the analysis thread. The audio is not a stage of this chain — TELESCOPE reads it and writes nothing back: 0 samples of latency, 0 tail, bit-exact output. Only the modules the open lens needs are computed; loudness is always on. #### Every control **01 · How loud** - **LOUDNESS** — How loud the programme is, to the broadcast standard — integrated, short-term and momentary LUFS, LRA and true peak, also per channel. How: **INTEGRATED** follows ITU-R BS.1770 with its double gate (absolute −70 LUFS, relative 10 LU below); **MOMENTARY** is a 400 ms window and **SHORT-TERM** a 3 s window at 10 Hz; **LRA** follows EBU Tech 3342; **TP MAX** is the true peak of BS.1770 Annex 2, and **L / R** give it per channel. Four bars share one 0 … −60 scale; while a window is still filling, the bar shows the partial value, dimmed, and it matches the official number to the bit once the window fills. The history shows the last three minutes. Controls: **RESET** · **PAUSE / RESUME** · **TARGET** (a streaming platform, drawn as a line with the distance to your integrated value). Tip: The 4× true peak has a known maximum under-read (EBU test 3341-17 reads −6.3160 dBTP against a real −6.0): if your master measures −1.0 dBTP here, treat it as a little higher. **02 · How much headroom** - **DYNAMICS** — Not how loud the mix is, but how much headroom is left in it. How: **PSR** = max true peak of the last 3 s − short-term; **PLR** = max true peak since RESET − integrated (AES TD1004). Below them, the short-term histogram since RESET in 61 bins of 1 LU, and the clip events above a threshold in dBTP (default −1.0, range −3…0) on a 10-minute timeline. An event opens on the first sample over the threshold and closes 100 ms below it: 10 bursts of 50 ms read 10 events. Controls: **RESET** · **PAUSE / RESUME** · the threshold — changing it restarts the count. Tip: A steady sine reads PSR = PLR = 0.0 dB, and that is the right answer. The 8 dB line is a reference, not a verdict: the lens never paints your number red. **03 · Where the energy sits** - **SPECTRUM** — The spectrum with an exact dB reference: each bin in dBFS referred to a full-scale sine. How: **FFT** 1 024 – 32 768 · **WINDOW FN** Hann, Blackman-Harris or Kaiser β=9 · **OVERLAP** · **CHANNEL** L · R · M · S · L+R · **BANDS** free / ⅓-octave ISO 266 (30) / Bark (24) · **SLOPE** (default 3, so pink noise reads flat; a display transform, and the number under the cursor is the tilted one) · **AVERAGE** none / exponential τ 0.1–10 s / infinite · **HOLD** 0–60 dB/s · **RANGE** · **SMOOTHING** off / 1/24 / 1/12 / 1/6 octave — a screen setting only: the readout, the peak hold and every number keep reading the raw bins. Tip: Changing the FFT size moves the noise floor, not the tones; and a band with no bin in it is marked differently from a band at zero. For bass, use a big FFT. **04 · How it moves in time** - **SPECTROGRAM** — Level over time: frequency in Y (log, 20 Hz – 20 kHz, 512 rows), time in X with now on the right, level in the colour. How: **HISTORY** 10 / 30 / 60 s · **RANGE** 60 / 90 / 120 dB · **CHANNEL** · **PALETTE**. It uses the instantaneous power, never the averaged one, and each row takes the maximum of the bins in its cell. The FFT settings are SPECTRUM's — the same knob. Changing FFT, overlap, channel, range or history clears the display instead of mixing two mappings; changing only the history does not restart the analysis. Tip: The palette is one setting for four lenses — SPECTROGRAM, STEREO SPECTROGRAM, WATERFALL and FIELD. ovni, inferno and viridis are monotonic in luminance, so brighter always means louder; spectrum orders the level by hue instead. **05 · The same, in depth** - **WATERFALL** — The same data as the spectrogram, in depth: frequency in X, level as height, time in Z with now at the front. How: **LINES** 60 / 90 / 120 · **TILT** · **HISTORY** · **RANGE**. The projection is oblique and drawn in software — a 2.5D view, with no GPU and no divide by z — so the same distance in frequency measures the same at any depth. Occlusion is drawn front to back with a horizon, proven identical to the literal painter's algorithm. If the ring holds fewer columns than the lines you ask for, it draws the ones there are. The readout gives frequency and dB of the front line only. Tip: Height is easier to compare than colour, but a short event can hide behind a later one. Read it next to SPECTROGRAM: they are complementary. **06 · Which notes** - **CQT** — The spectrum by note: constant-Q with 24 bins per octave from A0 (27.5 Hz), one window per bin — 229 bins at 48 kHz. How: One bar per bin over a note axis, with a 114-key keyboard drawn to scale underneath, then the chromagram in twelve bars with the tonic in amber and the key. **CHANNEL** · **CHROMA** (the chromagram's smoothing: 0.5 / 2 / 5 s). The key is never stated alone: it always carries its confidence and its share of time, because the confidence has a floor and some music is genuinely ambiguous. Silence reports no key. Tip: The bass is late by physics: A0 needs 1.2410 s of listening at 48 kHz, and the lens prints A0 · 1.24 s on screen instead of hiding it. **07 · Octaves at a glance** - **SPIRAL** — The same constant-Q, coiled: one turn per octave, so equal notes land at the same angle. How: Angle = pitch class, with C at the top and clockwise; radius = octave, from A0 inside to the highest bin outside; each spike's brightness and thickness follow the level. In the centre, the chroma wheel with the tonic and the key. **CHANNEL** · **CHROMA**. Tip: The guide circles mark the radius where octave n starts, not octave n itself: read as rings, they read one octave too much. **08 · How wide, and how it folds to mono** - **SCOPE** — The classic ways of looking at stereo, over a 100 / 300 / 1000 ms window: correlation, width, balance and mono loss. How: The mode: **LISSAJOUS**, an auto-scaled goniometer that says its scale (the edge is the hop's peak); **POLAR SAMPLE**, the same angle with the radius as level in dB; and **POLAR LEVEL**, built for mixing — a half-circle with mono at the top and L and R on the base, one ray per degree, the average in time filled and the peak-hold as an outline, and what is out of phase folded onto the base with its percentage. **TRIGGER** · **WINDOW** 100 / 300 / 1000 ms · in POLAR LEVEL, **PEAK DECAY** 12 / 24 / 48 dB/s and **LIN** / **dB**. The oscilloscope shows 40 ms. Tip: Folding loses nothing: each out-of-phase direction lands next to where it belongs by panning, and the percentage beside it says how much. It is pan direction by energy — not localisation. **09 · Stereo, band by band** - **BAND CORRELATION** — SCOPE's five sums, per ⅓-octave band (ISO 266, 30 bands). How: **WINDOW** 0.3 / 1 / 3 s — the effective window is published as measured (0.2987 / 1.0027 / 3.0080 s) · **ROW** (which figure sits on top). The broadband figure computed from the bins agrees with the time-domain meter: correlation within 0.00076, mono loss within 0.0033 dB. A band with no bin in it is marked as unmeasurable, not as zero, and the readout says how many bins measured each band. Tip: Measured on an 80 Hz sine in mono plus out-of-phase noise above 2 kHz: broadband reads corr +0.24, while the 80 Hz band reads +1.00 and the 4 kHz band −1.00. **10 · Phase over time** - **STEREO SPECTROGRAM** — The sonogram with the colour's meaning changed: colour is the phase (per-bin coherence), brightness is the level. How: Red out of phase, green wide, white mono. A cell with no energy is black, and silence sits at the centre of the scale, meaning undefined. **HISTORY** · **RANGE** · **WINDOW**. The coherence is smoothed over the window — a single frame over two independent sources would be confetti — but the energy is not: it keeps its time axis. Tip: Both channels are summed for the brightness, so a mono signal reads 3.01 dB above what SPECTRUM shows for the L channel alone. **11 · Energy by pan direction** - **FIELD** — Energy per pan direction × frequency, 64 columns × 96 rows, with decay and a trail, in 2.5D. How: Per bin, pan = (ΣRR − ΣLL) / (ΣRR + ΣLL); under the constant-power law it equals −cos 2θ, checked to 4.9·10⁻⁸. **DECAY** 0.5 / 1 / 2 s · **WINDOW** 0.3 / 1 / 3 s · **PALETTE**. The axis reads L … C … R, the readout gives the pan in per cent and the dB relative to the grid's maximum. Under the axis, fixed and impossible to switch off: **energy by pan direction · not localisation**. Tip: FIELD cannot tell mono from out of phase — both land in the centre. Read it with BAND CORRELATION and STEREO SPECTROGRAM: the three go together. **12 · Against a reference** - **TONAL BALANCE** — What colour your mix is, compared with a reference track you load. How: The reference is analysed whole, offline. Both ⅓-octave curves are normalised by their own integrated LUFS, so the lens compares tilt, not level — and the delta sums to zero, checked band by band to 0.006 dB. Above, the two curves on a fixed +6 … −42 dB scale; below, the delta as ±12 dB bars with a labelled ±3 dB reference band, one vertex per band. **RESET** (keeps the reference) · **LOAD** (or drag a file onto the lens) · **REMOVE**. Tip: Add 6 dB above 2 kHz and the delta will not read +6 up top: the loudness rises too, so it draws +1.70 dB above 8 kHz and −4.30 dB below 500 Hz — the same 6 dB, counted at equal volume. **13 · What it concludes** - **VERDICT** — Twelve lenses show; this one says. A deterministic rules engine — 21 rules over a per-second history — where every sentence carries the number that supports it and the id of the rule that produced it. How: First a headline with the count, no adjectives — for example *15 checks within range · 3 to look at, the first at 0:20*. Then four sections: **within range** · **how it will feel** · **where it translates** · **what to check, and where**. A finding puts the number first and ends with where to look, never with what to do; a dip across several neighbouring bands is one finding. Without a reference it compares against your material's own trend; with one, against the reference. The six per-device checks are generic definitions of a class of system, not speaker curves. **RESET** · **MODE** (live / FILE) · **FILE** → **LOAD**. The footer cannot be removed: *Measurement, not taste. Device checks are generic. Re-run the analysis after every change.* Tip: No findings is not the same as finished. When the rules find nothing it says *Nothing outside the ranges of these rules. They measure; what they can't hear is yours.* — never ready. #### Tips & tricks - Load a reference in **TONAL BALANCE** — a previous mix of yours or a commercial track — and VERDICT compares against it as well: its sentences then say which region does not look like the mix you chose as a target. Against your material's own trend it is a blunt instrument; a reference is strictly better. - In **POLAR LEVEL**, the out-of-phase fold is not an alarm. Every mix with decorrelated sources has instantaneous samples out of phase; the colour's emphasis follows the measured mono loss, and the percentage beside the base says how much. - **Re-run the analysis after every change.** VERDICT's times come from its per-second history since the last RESET, and if you open it halfway through a track the report starts there. For exact times over a whole track, use FILE mode. - Checking a bounce? Drop it on VERDICT in **FILE** mode: the offline analysis uses the same classes as the live one and comes out equal to the bit — 60 s of audio in about a quarter of a second on an M4. - Read the stereo lenses together: **SCOPE** for the whole, **BAND CORRELATION** per band, **STEREO SPECTROGRAM** over time and **FIELD** for where the energy sits. FIELD alone cannot tell mono from out of phase. - Pick a streaming **TARGET** in LOUDNESS and its line is drawn across the LUFS bars and the history, with the distance to your integrated value written out. #### Troubleshooting - **VERDICT says “no complete seconds yet”.** VERDICT works from a per-second history, and no whole second has been analysed yet. Let the track play from the start, or use FILE mode — which is exactly what the message says: *no complete seconds yet: let the track play from the start, or use FILE*. - **TONAL BALANCE says “reference not found”.** A session stores the reference's path, not its numbers. If the file was moved or deleted, the lens says *reference not found:* with its name and draws no curve, instead of showing saved numbers as if they were still the file. Load it again from where it is now. - **With reduced motion on, some lenses stop animating.** TELESCOPE honours the system's reduced-motion setting lens by lens: SCOPE leaves no trail and POLAR LEVEL has no memory, SPECTRUM stops smoothing between frames, and FIELD drops its trail. SPECTROGRAM, STEREO SPECTROGRAM and WATERFALL keep running — their time axis is the data, and freezing them would show less of it. - **How do I change the language?** **LANGUAGE**, at the foot of the lens strip, switches the whole plug-in — all thirteen lenses — and shows each language by its own name. English and Spanish are reviewed; Portuguese, French, German and Italian are translated with the industry's terms and pending review by a native speaker. A missing phrase falls back to English, never to a blank. - **macOS warns me about the installer.** It should not: the package is signed with a Developer ID and notarized by Apple, and Gatekeeper opens it with a double-click. If macOS ever warns about this .pkg, do not bypass the warning — download it again from the release and check its SHA-256 against SHA256SUMS.txt. - **My DAW lists parameters I can't see in the window.** **inGain**, **output** and **monoSafe** come from the label's shared chassis, which every OVNI plug-in carries. TELESCOPE does not show them because it does not process; at their defaults the audio is bit-exact — which is what the pass-through test verifies. #### Specs - Formats: VST3 · AU (macOS 11+, universal) - Latency: 0 samples · bit-exact pass-through - Loudness: ITU-R BS.1770 · EBU Tech 3341 tests 1–5 & 15–19 · Tech 3342 tests 1–4 - True peak: BS.1770 Annex 2 · 4× below 96 kHz · 2× from 96 to 176.4 kHz · 1× from 192 kHz - Analysis: on its own thread · file analysis equal to the bit - Channels: mono or stereo in → stereo out · stereo only - Languages: EN · ES · PT · FR · DE · IT - Licence: AGPLv3 - Windows: VST3 x64 · Windows 10+ · unsigned ZIP ## How to install The same steps for all of them: pick your system, then let your DAW find the new plugin. ### macOS macOS 11 or newer · Apple Silicon and Intel 1. Download the `.pkg` of the plugin you want. 2. Double-click it. The installer is signed with a Developer ID and notarized by Apple, so macOS opens it without warnings. 3. macOS’s own Installer asks for your admin password and puts the VST3 in `/Library/Audio/Plug-Ins/VST3` and the AU in `/Library/Audio/Plug-Ins/Components`. 4. Nothing lands in quarantine and there is no Terminal step. Open your DAW and rescan. SUPERNOVA also installs its standalone app in `/Applications`. ### Windows Windows 10 or newer · 64-bit · VST3 On Windows today: ORBIT and TELESCOPE — a VST3 in an unsigned ZIP, so Windows may warn. SUPERNOVA: macOS only. 1. Before extracting: right-click the ZIP › Properties › tick **Unblock** › Apply. 2. Right-click the ZIP › **Extract All**. 3. Copy what is inside the package’s `VST3` folder to `C:\Program Files\Common Files\VST3` — Windows asks for administrator permission: it is a system folder. 4. If SmartScreen shows “Windows protected your PC”: **More info** › **Run anyway**. 5. Open your DAW and rescan VST3 plug-ins. The same steps as the README that comes inside the ZIP. ### Your DAW Checked against each maker’s documentation on September 15, 2026. | DAW | On Mac | On Windows | To see a new plugin | Docs | |---|---|---|---|---| | Ableton Live 12 | VST3 or AU | VST3 | Settings › Plug-Ins › Plug-In Sources: turn on your VST plug-in sources (on a Mac, also **Use Audio Units**), then **Rescan**. | https://www.ableton.com/en/live-manual/12/working-with-instruments-and-effects/ | | Logic Pro | AU | — | Settings › Plug-in Manager › **Reset & Rescan Selection**; if it is not listed, **Full Audio Unit Reset** and reopen Logic. | https://support.apple.com/en-us/122179 | | REAPER | VST3 or AU | VST3 | VST3: Options › Preferences › Plug-ins › VST › **Re-scan** (it also scans at launch). | https://www.reaper.fm/userguide/ReaperUserGuide780.pdf | | FL Studio | VST3 or AU | VST3 | Options › File settings › Manage plugins › **Find installed plugins**, with Verify plugins on. | https://www.image-line.com/fl-studio-learning/fl-studio-online-manual/html/basics_externalplugins.htm | | Cubase · Nuendo | VST3 | VST3 | Studio › VST Plug-in Manager › **Scan for new and blocked plug-ins** (Shift rescans all). | https://www.steinberg.help/r/cubase-pro/15.0/en/cubase_nuendo/topics/installing_and_managing_plugins/installing_and_managing_plugins_vst_plug_in_manager_toolbar_r.html | | Studio One · Fender Studio Pro | VST3 or AU | VST3 | View › **Plug-in Manager** › **Update Plug-in List**, for plug-ins installed while it is open. | https://fenderstudiopromanual.fender.com/en/Content/The_Browser_Topics/Instruments_and_Effects_Tabs.htm | | Bitwig Studio | VST3 | VST3 | The folders it scans: Dashboard › Settings › Locations › **Plug-in Locations**. | https://www.bitwig.com/userguide/latest/the_dashboard/ | | Pro Tools | AAX only | AAX only | OVNI does not make AAX, so these plugins do not load in Pro Tools. | https://kb.avid.com/pkb/articles/compatibility/en343311 | ### Which format do I need? On a Mac, Logic Pro uses the AU; Cubase, Nuendo and Bitwig use the VST3; Live, REAPER, FL Studio and Studio One load either — the installer puts both in place. On Windows, VST3. ### Uninstalling Delete the bundle. On macOS, the `.vst3` in `/Library/Audio/Plug-Ins/VST3` and the `.component` in `/Library/Audio/Plug-Ins/Components`; on Windows, the `.vst3` folder in `C:\Program Files\Common Files\VST3`. | Plugin | What it installs (macOS) | Installer receipts | |---|---|---| | ORBIT | OVNI ORBIT.vst3 · OVNI ORBIT.component | com.ovni.plugins.orbit.vst3 · .au | | SUPERNOVA | SUPERNOVA.vst3 · SUPERNOVA.component · /Applications/SUPERNOVA.app | com.ovni.plugins.supernova.vst3 · .au · .app | | TELESCOPE | TELESCOPE.vst3 · TELESCOPE.component | com.ovni.plugins.telescope.vst3 · .au | The `OVNI Audio` folder in `/Library/Audio/Plug-Ins` holds the licence texts and every OVNI plugin shares it: delete it when the last one goes. Optional, on macOS, so the Installer forgets the receipt: `sudo pkgutil --forget com.ovni.plugins..vst3` (and `.au`; SUPERNOVA also `.app`). `pkgutil --pkgs | grep ovni` lists them. ## Which one do I need? Five things people come here for, which module does each one, and where each module stands today. - **Move a sound through space** — Out now: ORBIT places a source by angle and distance and orbits it. PULSAR and DUST are coming soon. ORBIT (Free · v0.3.0), PULSAR (Locked · coming soon), DUST (Locked · coming soon). - **Open up a space** — Nothing released does this yet: NEBULA and HALO are both coming soon. NEBULA (Locked · coming soon), HALO (Locked · coming soon). - **Play with the spectrum** — Also coming soon: HORIZON freezes an instant and gates it to the beat; AURORA sends each band to its own place. HORIZON (Locked · coming soon), AURORA (Locked · coming soon). - **See the sound** — Out now, macOS only: an image or a video that reacts to your audio, as a plugin or as an app. SUPERNOVA (Free · v0.4.0). - **Measure and read the mix** — Out now: thirteen lenses, and VERDICT writing each finding with its number and its rule. TELESCOPE (Free · v0.1.0). ## Questions about the catalog **Is it free — and until when?** Yes, with no deadline: the code is free forever (AGPLv3) and so are the installers for ORBIT, SUPERNOVA and TELESCOPE. New products released later may be paid. **What does AGPLv3 mean here?** It is the licence of the code: the code stays public, and anyone can build the plugins from it for free, forever, without asking or leaving an email. What has a date is the ready-made installer. **Which formats and systems?** VST3 and AU on macOS 11 or newer, universal (Apple Silicon and Intel). VST3 64-bit on Windows 10 or newer where there is a ZIP — today, ORBIT and TELESCOPE. There is no AAX, so they do not load in Pro Tools. **Is there a Windows version?** ORBIT and TELESCOPE: yes, a VST3 in a ZIP (unsigned — SmartScreen may warn). SUPERNOVA: macOS only. **Why does the download ask for my email?** The ready-made installers go through the list: leave your email once per browser and the download starts right away — you get new modules and updates, nothing else. The source on GitHub needs no email, ever: AGPLv3 keeps it public. ## Price, launch window and licence Two different things, and the difference matters: - **EN:** Yes, with no deadline: the code is free forever (AGPLv3) and so are the installers for ORBIT, SUPERNOVA and TELESCOPE. New products released later may be paid. - **ES:** Sí, y sin fecha: el código es libre para siempre (AGPLv3) y los instaladores de ORBIT, SUPERNOVA y TELESCOPE también. Los productos nuevos que salgan más adelante pueden ser pagos. It is the same model Ardour uses: the public repository (https://github.com/ovniaudio/ovni) stays open under AGPLv3 and anyone can build the plugins for free, forever, without asking or leaving an email. What has a date is the ready-built binary packaged in a double-click installer, signed and notarized by Apple. It is NOT a “limited-time offer”: nothing in the software stops being free. Whoever joins the list before the window closes also gets the Launch Preset Pack free when it drops. Voluntary support (never required): https://ko-fi.com/ovniaudio ## Downloads and releases - ORBIT macOS (double-click .pkg installer, VST3 + AU universal, macOS 11+): [OVNI-ORBIT-v0.3.0.pkg](https://ovniaudio.com/download/OVNI-ORBIT-v0.3.0.pkg). Signed and notarized by Apple (Developer ID): double-click, no Gatekeeper warnings. - ORBIT Windows (VST3 x64, Windows 10+): [OVNI-ORBIT-v0.3.0-Windows.zip](https://ovniaudio.com/download/OVNI-ORBIT-v0.3.0-Windows.zip). Unblock the ZIP (Properties → Unblock) and copy the .vst3 folder into C:\Program Files\Common Files\VST3. - Upgrading ORBIT: Coming from 0.2.x — the plug-ins are now called **OVNI ORBIT.vst3** and **OVNI ORBIT.component**. On macOS the installer retires the old *ORBIT* for you, but only if it is ours (*com.ovni.orbit*): someone else's *Orbit* is never touched, and nothing is deleted — it is moved to *~/Library/Application Support/OVNI Audio/replaced//*. Then run *killall -9 AudioComponentRegistrar* and rescan in your DAW. On Windows, delete *C:\Program Files\Common Files\VST3\ORBIT.vst3* by hand before copying the new folder in. Your saved projects keep working: only the file name moved. - SUPERNOVA macOS (.pkg installer: VST3 + AU plugins and the standalone app, universal, macOS 11+): [OVNI-SUPERNOVA-v0.4.0.pkg](https://ovniaudio.com/download/OVNI-SUPERNOVA-v0.4.0.pkg). Signed and notarized by Apple (Developer ID): double-click, no Gatekeeper warnings. SUPERNOVA is macOS only. - TELESCOPE macOS (double-click .pkg installer, VST3 + AU universal, macOS 11+): [OVNI-TELESCOPE-v0.1.0.pkg](https://ovniaudio.com/download/OVNI-TELESCOPE-v0.1.0.pkg). Signed and notarized by Apple (Developer ID): double-click, no Gatekeeper warnings. - TELESCOPE Windows (VST3 x64, Windows 10+): [OVNI-TELESCOPE-v0.1.0-Windows.zip](https://ovniaudio.com/download/OVNI-TELESCOPE-v0.1.0-Windows.zip). Unblock the ZIP (Properties → Unblock) and copy the .vst3 folder into C:\Program Files\Common Files\VST3. - Checksums: [SHA256SUMS.txt](https://ovniaudio.com/download/SHA256SUMS.txt) for the v0.2.0 wave; each release since carries its own: ORBIT [orbit-v0.3.0](https://github.com/ovniaudio/ovni/releases/tag/orbit-v0.3.0), SUPERNOVA [v0.4.0](https://github.com/ovniaudio/ovni/releases/tag/v0.4.0), TELESCOPE [telescope-v0.1.0](https://github.com/ovniaudio/ovni/releases/tag/telescope-v0.1.0). - Source code: [github.com/ovniaudio/ovni](https://github.com/ovniaudio/ovni) (the catalog) and [github.com/ovniaudio/orbita](https://github.com/ovniaudio/orbita) (ORBIT).