Ghost Utility

Wide-range signal trim. Zero is exact unity; use fine adjustment or typed values for small changes. This is Gain, not an invented separate Input/Output pair.

Try this: Enter -3 dB and compare the actual meter change, then use Unity.

Amount of attenuation while Dim is active. It remains stored while Dim is Off.

Try this: Start at the retained -18 dB and adjust for comfortable monitoring.

Frequency-dependent all-pass rotation, displayed as phase shift at 1 kHz. Zero bypasses rotation; other frequencies do not receive the same degree shift.

Try this: Try a small move only after polarity and delay have been checked.

Right all-pass rotation using the same 1 kHz reference. It changes phase relationships, not simple level.

Try this: Disable Phase Link when an independent right-channel correction is required.

Causal left-channel delay. Milliseconds, samples and distance are views of this one saved time value. This is intentional alignment, not an acausal advance.

Try this: Use a one-sample nudge at the current sample rate, then compare correlation and mono.

Causal right delay with fractional-sample precision. Distance uses 343 m/s; sample conversion follows the current rate.

Try this: Delay the earlier signal, not the already late one.

Positions a mono source or applies a linked position offset to existing stereo channel positions. It does not silently sum stereo to mono.

Try this: Compare a small Pan move with Balance to hear positioning versus level.

Attenuates the opposite L/R side with unity at center and no endpoint boost. It changes channel level, not each channel's pan position.

Try this: Use for a literal left/right level correction.

Positions the existing left-channel content independently. Default full left preserves its original position.

Try this: Move inward to narrow that channel's location without selecting a matrix mode.

Positions existing right-channel content independently. Default full right preserves its original position.

Try this: Compare an inward right position with a Width reduction.

Exact static M/S width:0% is mono, 100% retains the original Side relationship, 200% doubles Side contribution. It is not Ghost Stereo's adaptive spatial macro.

Try this: Compare 0%, 100% and 150% and check the mono consequence.

Independent gain for the sum-derived Mid component in the M/S processing path. Zero is unity.

Try this: Lower a few dB to compare center versus sides.

Independent gain for the difference-derived Side component. Mono input has no Side to amplify.

Try this: Try -3 dB before raising Width on already unstable material.

Attenuates the opposing Mid/Side component. Center is unity, without an endpoint boost.

Try this: Compare a small move with independent Mid/Side Gain.

Cutoff for the enabled subsonic maintenance filter. The retained frequency is inactive while Subsonic is Off.

Try this: Start at 10 Hz and inspect the low-energy/headroom consequence.

Click-free monitoring silence. It does not reset the settings, measurements or stored sound.

Try this: Toggle to silence this path without changing Gain.

Applies the retained Dim Level as an explicit monitoring attenuation. It is separate from Mute and ordinary Gain.

Try this: Use Dim for a temporary quieter check.

Exact left-channel sign inversion. It is not a delay or a constant frequency-dependent phase correction.

Try this: Use only for a deliberate polarity check, then compare in mono.

Exact right-channel sign inversion, independent of left. A polarity mismatch can radically change mono summing.

Try this: Compare with Left untouched on a known related stereo pair.

A linked intentional edit sets both phase controls to the edited value in one undo transaction. Unequal recalled values remain intact until an edit.

Try this: Turn Off for independent phases; On does not silently rewrite a recalled pair.

Explicit Stereo, Mono Sum, Left/Right Audition, L→Both, R→Both, Swap, Mid/Side Audition and Mono Left/Right routing. Audition mutes the other stereo channel; copying to both does not.

Try this: Choose Mono Sum for a whole-image mono check, not Left Audition.

L/R processes ordinary stereo; Encode M/S emits Mid on left/Side on right; Decode expects that representation; M/S workspace works internally in M/S then returns L/R.

Try this: Use ordinary L/R unless you intend an explicit encode/decode chain.

Half-sum uses Mid=(L+R)/2 and Side=(L−R)/2. Orthonormal uses 1/√2. Match the convention across encode/decode for a correct round trip.

Try this: Keep Half-sum unless the surrounding M/S chain explicitly expects Orthonormal.

Selects Linear(-6 dB), Equal Power(-3 dB) or Compromise(-4.5 dB) positioning behavior. It is explicit, not an undocumented host-law guess.

Try this: Keep Equal Power initially and compare center versus edge on a mono source.

Auditions Mid within the active M/S processing path. Both solos retain both components; a mute takes precedence.

Try this: Use to inspect what remains common between channels.

Auditions the difference-derived Side component. Silence on a truly mono source is expected.

Try this: Use to inspect stereo difference information, then release Solo.

Removes Mid in the M/S path, even if Mid Solo is also active.

Try this: Use for an explicit Side-only check rather than assuming the result is normal stereo.

Removes Side in the M/S path. Its effect differs from muting one physical input channel.

Try this: Compare with the Mono Sum matrix and inspect the routing.

Inverts the sign of Mid, not the whole left output channel.

Try this: Check the workspace before using it and compare the resulting stereo relationship.

Inverts Side in the M/S path. This can exchange the image's lateral relationship; it is not a width increase.

Try this: Compare briefly on an identifiable left/right recording.

Enables precision 5 Hz DC removal. DC detection remains available while correction is Off; there is no automatic gate.

Try this: Enable only when actual DC offset needs correction.

Enables the narrow-range maintenance high-pass. Off preserves the unfiltered signal; it is not an artistic full-range filter.

Try this: Use only for unwanted very-low-frequency energy.

Chooses 12 or 24 dB/oct for the enabled maintenance filter. More slope is not automatically better.

Try this: Use 12 dB/oct first, then compare 24 only if the technical problem requires it.