GhostPRODUCT GUIDESLegal notices

TRANSPOSE / BRANCH / ATTRACT

Ghost Pitch Shifter

Start with one shifted voice. Grow a moving pitch family.

Transpose precisely, adjust the apparent spectral envelope independently, then introduce a source-derived population whose pitches can spread, orbit, attract and rebound.

Shift your first phrase Explore the interface

Scale Gravity organizes generated descendants. It does not correct the original input or move the primary Pitch setting onto a scale.

Actual Pitch Field showing voice ancestry and purple waveform context in a single display.
Actual plugin display · full annotated interface below

START HERE / ABOUT 90 SECONDS

Make a clean octave before building a cloud.

Use a clear instrument or vocal phrase. Insert Pitch Shifter and choose Init.

START FROM INIT

Pitch 0 st · Fine 0 ct · Formant 0% · Voices 0% · Spread 0% · Shape 0% · Gravity 0% · Gravity Target Harmonic · Key C · Scale Chromatic · Polarity 0% · Orbit 0% · Elastic 0% · Tension 0% · Fracture 0% · Character Clean · Mix 100% · Output 0 dB

0 of 4 steps checked — reading progress only.

WHAT TO LISTEN FOR

One control makes a conventional transposition; population controls extend that same processor into an ensemble rather than a manually programmed harmonizer.

Use small Fine and Spread for restrained detune. For scale relationships, select Gravity Target Scale and the actual song Key/Scale before raising Gravity.

LEARNING ILLUSTRATION — NOT CONNECTED TO YOUR PLUGIN

Grow a population, not a manual voice list.

Qualitative guide; no voices or audio are calculated here. Spread and other controls still determine whether descendants separate audibly.

A qualitative explanation, not a calculated response or audio preview. Use the plugin to hear the result.

PARAMETER REFERENCE

A small set of big gestures.

Pitch/Fine establish the root. Voices grows the family, Spread its extent, Shape its distribution. The remaining physics matter when that family exists.

Pitch-24–24 st · Default 0 st

Primary transposition in semitones. Pointer detents make exact intervals convenient; continuous entry, automation and modulation can travel between them.

Try this: Set +12 st with Voices 0% to learn the simple case.

Fine-100–100 ct · Default 0 ct

Cents offset relative to Pitch; 100 cents equals one semitone. It remains independently bounded from the main transposition.

Try this: Try +8 cents for a small detune instead of moving Pitch a whole semitone.

Formant-100–100% · Default 0%

Changes apparent spectral-envelope size/brightness independently of pitch: negative darker/larger, positive brighter/smaller. Applied to generated voice geometry, useful beyond vocals.

Try this: At a fixed octave, compare±20%.

Voices0–100% · Default 0%

Smoothly grows a bounded population around the primary voice.0% is one voice, not silence; new descendants fade in/out rather than appearing abruptly.

Try this: Try 35% before spreading the family.

Spread-100–100% · Default 0%

Extent of pitch and stereo relationships. Positive expands symmetrically; negative creates asymmetric/chasing structures. Zero gathers around the primary pitch.

Try this: At Voices 40%, compare +30% and -30%.

Shape-100–100% · Default 0%

Distribution within the available pitch space. Positive gathers toward the primary with separated outer voices; negative evacuates the center toward boundaries.

Try this: Keep population/extent fixed while comparing±30%.

Gravity-100–100% · Default 0%

Positive bends generated descendants toward harmonic or scale attractors; zero ignores them; negative repels. It never scale-corrects the input or primary Pitch.

Try this: With a spread population, compare 0% and +50%.

Gravity TargetHarmonic / Scale · Default Harmonic

Harmonic uses intervals around the primary voice. Scale uses absolute Key/Scale destinations for generated descendants.

Try this: Choose Scale only when you know the piece's key/collection.

KeyC / C# / D / D# / E / F / F# / G / G# / A / A# / B · Default C

Root for Scale gravity, C through B. Retained but inactive for Harmonic target; changing it retargets existing attractors smoothly.

Try this: Use C with Major for a simple first scale test.

ScaleChromatic / Major / Natural Minor / Harmonic Minor / Melodic Minor (ascending) / Dorian / Mixolydian / Major Pentatonic / Minor Pentatonic / Minor Blues · Default Chromatic

Shared ten-collection library: Chromatic, Major, Natural/Harmonic/ascending Melodic Minor, Dorian, Mixolydian, Major/Minor Pentatonic and Minor Blues. It defines attractions, not a piano-roll sequence.

Try this: Compare Major and Minor Pentatonic with Scale target and positive Gravity.

Polarity-100–100% · Default 0%

Interaction between voices: positive attraction, negative repulsion, zero independence. This is not audio sign inversion and differs from Gravity's musical destinations.

Try this: Try±25% with a moving family, leaving Gravity fixed.

Orbit-100–100% · Default 0%

Circulates voices through the geometry in either direction. Zero stops the current without resetting positions. The engine owns the timescale; there is no separate Rate knob.

Try this: Try +20%, then reverse through zero to -20%.

Elastic0–100% · Default 0%

Lag, overshoot and rebound as pitches seek their geometry. Zero is precise tracking; higher values let a short disturbance keep settling afterward.

Try this: Raise toward 50% and briefly move Spread.

Tension-100–100% · Default 0%

Coupling through parent/child ancestry. Positive tethers descendants; negative loosens and can oppose their movement. Zero is the normal ancestry relationship.

Try this: Compare±30% with Elastic and Orbit active.

Fracture0–100% · Default 0%

Descendants diverge in pitch, timing and formant while retaining ancestry. Reducing it reconstructs the family rather than choosing unrelated random detuning.

Try this: Try 25% before extreme branching.

CharacterClean / Elastic / Glass / Organic / Fracture / Void · Default Clean

Changes pitch-engine and crossing laws, preserving the running family. Differences need an active population/movement to be fully apparent.

Try this: Compare Clean and Elastic after raising Voices/Spread.

Mix0–100% · Default 100%

Blends aligned original and generated paths. Zero is dry before Output; intermediate blends retain original context underneath transposition.

Try this: Use 100% for the first octave test, then try 40% for parallel harmony.

Output-12–12 dB · Default 0 dB

Final gain after blending. A denser family can change perceived level; this is not population strength or a hidden limiter.

Try this: Trim after settling on Voices and Mix.

CHARACTER REFERENCE

Choose a character.

Clean

Precise transformation; voices cross cleanly.

Elastic

Emphasizes stretching, rebound and momentum exchange.

Glass

Separated bright voices with crystalline convergence/reinforcement.

Organic

Softer transient behavior and subtly differentiated formants.

Fracture

More independent branches and irregular temporal interactions.

Void

Weaker descendants can fade while stronger branches persist, opening moving gaps.

KNOW YOUR INSTRUMENT

Find your way around.

Blue nodes and ancestry represent actual pitch-family state. Purple I/O waveform context occupies the same display but shows recent time/amplitude, not pitch or a separate FFT.

Actual Pitch Field showing voice ancestry and purple waveform context in a single display.
Actual native Clean editor during the octave-up example. Open full-size capture

Select a numbered control

Choose a marker, or use the complete keyboard-accessible control reference above.

Node horizontal position means stereo location and vertical position means pitch offset in semitones. Scale target adds subtle destination lanes; it is not a piano-roll editor.

Horizontal/vertical drag edits Spread/Gravity; Shift edits Shape/Tension. Pulling a generated node requests changes to this shared geometry, not a hidden individual voice parameter.

After release the real family settles according to Elastic, Gravity, Polarity and Tension. Tab/arrows/Enter/Home/Escape provide keyboard editing, reset and cancellation.

HEAR THE IDEA

Put it into practice.

One clean octave up

Actual Clean Pitch Shifter at +12 semitones, Voices 0% and Formant 0%, fully processed. The source is latency-aligned to the measured processing path and level matching is disclosed below.

Source

Source WAV

Processed · actual Ghost DSP

Processed WAV
Example settings and provenance

{"pitch": 12, "voices": 0, "formant": 0, "mix": 100}. Actual native processor, stereo48k/127/double; source latency-aligned. Whole-file wet RMS gain 1.396241; common trim 1.0. Identical final20ms file fade; no listening approval inferred.

Reproduction record

A restrained octave layer

Use bass, guitar or a vocal phrase.

Pitch -12 st · Voices 0% · Formant 0% · Mix 35%

  1. Hear the fully wet octave first.
  2. Reduce Mix to place it under the source.
  3. Adjust Formant slightly only if the apparent size needs changing, then trim Output.

Listen for: An octave layer, not source pitch correction.

A family attracted to a scale

Use a clear pitched phrase in C major.

Voices 45% · Spread 40% · Gravity Target Scale · Key C · Scale Major · Gravity 60% · Orbit 15% · Elastic 30%

  1. Confirm Scale target, C and Major.
  2. Compare Gravity zero and positive while descendants are spread.
  3. Add restrained Orbit/Elastic and listen to how the family moves around destinations.

Listen for: Generated descendants favor scale tones while the original and primary Pitch remain where they were set.

GO DEEPER

Beyond the first move.

Two kinds of attraction

Gravity relates voices to musical destinations; Polarity relates voices to one another. Elastic supplies inertia, while Tension propagates movement through ancestry. A short parameter disturbance can leave bounded motion after the control returns.

Simple settings remain simple

Voices 0%, neutral formant/geometry and a chosen Pitch make a conventional single-voice shifter. Character differences can be modest there because there are no descendant collisions to change. Full sessions retain the family/history; portable sounds do not carry a previous performance.

In your session

Fourteen continuous controls support internal modulation and MIDI CC learn. Gravity Target, Key, Scale and Character are discrete host automation only. Effective values stay within approved ranges.

No external audio sidechain, MIDI-note generator, per-voice mixer, internal sequencer or separate Sync/Rate is provided. Mono/stereo behavior follows the available output layout.

Processing latency is reported and dry/Rack paths align. Buffered material drains after input; no unrelated sound is generated from silence. Normal bypass drains history; hard bypass/Panic ends it safely.

The preset browser offers All, Factory, User and Favorites, category/search filters and relevant Character filtering. A row's heart changes your personal favorites without loading or saving the sound. Save As creates a distinct User sound; explicit Save updates only the selected User sound. Factory content is protected. Name, category and tags support organization; deleting a User category keeps its presets as Uncategorized. Manage supports rename, duplicate, confirmed delete and individual/bulk/category import/export. Personal favorites stay on this installation rather than travelling in exported sounds.

Standalone and the matching Ghost Rack module share the same User sounds, categories and favorites. A module preset changes portable sound settings, not Rack placement, Rack wet/dry, bypass or modulation assignments. DAW recall embeds the sound settings and does not depend on a preset file remaining on disk. Rack chain presets are a different library and affect the whole chain.

Drag a knob normally, hold Shift for fine adjustment or Option for extra-fine. Double-click restores its default. Enter/context menus provide exact values, reset and value copy/paste; eligible controls also offer MIDI Learn and modulation. Tab and arrow keys provide keyboard access. A/B compares configurations; the adjacent copy button copies the active configuration to the other. Undo/Redo groups a continuous gesture into one edit. Settings provides UI scale, Reduced Motion, source configuration, portable settings and Panic. UI size/motion preferences do not change the sound.

WHEN SOMETHING ISN’T CLEAR

Find your way back.

Why do Voices/Gravity seem inactive?

At Voices 0% there is only the primary voice. Grow Voices and Spread before expecting descendant geometry or scale attraction.

Why isn't my input corrected to the scale?

Scale Gravity intentionally applies to generated descendants, not the original or your primary Pitch setting.

Why don't Key/Scale change Harmonic mode?

They are retained but used only with Gravity Target Scale.

Why does movement continue after a gesture?

Elastic and the family's relationships retain real bounded momentum. This is not a stuck control or decorative animation.