SherlockEQ leaves beta. This is the first release without a leading zero, and
it is deliberately not a feature release: the app here is the same app that
shipped as 0.9.8, unchanged. What changed is the label. After a run of
releases that added the last of the 1.0 work and then went quiet on fixes,
the version number now says what the app has been for a while — finished
enough to stand on its own.
If you're already running 0.9.8, updating to 1.0 changes nothing about your
sound, your profiles, or your settings. If you're new here, the rest of these
notes are an overview of what SherlockEQ actually does, rather than a list of
what moved since last week.
What SherlockEQ is
A systemwide, per-ear equalizer for macOS. It captures the system audio mix
through Apple's Core Audio tap API, processes it, and plays the result to your
chosen output device — so every app is affected at once, with no per-app
setup, no virtual audio driver, and no kernel extension. Nothing you play is
ever recorded or sent anywhere.
Sound shaping
-
Graphic EQ. A ten-band equalizer on the audiometric grid
(31 Hz to 16 kHz), with bands labelled by what they do — rumble, bass,
warmth, voice body, clarity, sibilance, air — over a live spectrum.
-
Parametric EQ. A full canvas when you want it: frequency,
gain, Q, and filter type per band, stacked bands, and drag-and-edit
directly on the response curve.
-
Per-ear chains. Independent left and right filtering, not
just a balance offset — for asymmetric hearing, or a single ear that needs
something the other doesn't.
-
Master gain and balance after EQ, with headroom and
clipping made visible rather than left to guesswork.
-
Presets organized by goal rather than by music genre, and
a preset applies without disturbing your hearing adjustment or notch.
Hearing-aware tools
-
Audiogram to EQ. Enter per-ear thresholds — from a clinical
audiogram or from the built-in Listening Check — and SherlockEQ derives a
per-ear adjustment that lifts only where you've indicated reduced
sensitivity. It lives in its own layer, so presets and EQ edits never
overwrite it. New adjustments ease in over the first three weeks instead of
arriving at full strength on day one, and an optional Adaptive style gives
quiet passages more help and loud ones less.
-
Listening Check. An optional five-minute guided estimate of
your thresholds, ear by ear, using quiet pulsed tones — headphones and a
quiet room required. It's an at-home estimate to tune by ear, not a hearing
test and not a diagnosis.
-
Tinnitus tone finder and notch. A continuous-phase sine
sweep, tunable to one hertz, to explore where a tonal tinnitus sits — with
octave comparison and an optional average-a-few-matches aid, because pitch
matching is imprecise. Turn the result into a per-ear notch with Subtle,
Balanced, or Strong presets, or set depth and width by hand.
-
Adaptive Comfort. Three processors that act only when their
target appears, then back off: bring voices forward, soften harsh moments,
and reduce sharp "s" sounds. Each shows in plain language what it's doing
right now.
-
Headphone correction. Flatten a specific headphone's
measured response using AutoEQ-style curves, imported or downloaded, as a
separate layer ahead of your own EQ.
Levels and safe listening
-
Live spectrum analyzer behind the EQ canvas, with
switchable lenses — output, pre-EQ input, your manual curve, the resulting
response, and a safety overlay.
-
Metering in both flavours: per-channel peak/level meters
and an analog VU pair anchored to your own comfortable listening level.
-
Listening-dose estimate using the NIOSH model — 85 dBA over
8 hours, 3 dB exchange rate — anchored to a playback calibration you
provide and tracking your system volume as you change it. Optional
notifications as the day's accumulation approaches the warning threshold,
plus a 7-day history. It is informational guidance, not a calibrated
dosimeter, and it says so wherever a value is an estimate or simply
unknown.
Surfaces and control
-
Main window for full editing; menu-bar
popover for a glance at status plus gain, balance, profile, and
reference mode.
-
Analog Control Unit. An optional vintage-style front panel
with volume, output, balance, bass, mid, and treble knobs, VU meters, and an
expandable spectrum panel — a separate quick-adjust mode driving its own
tone profile and the macOS volume and output device.
-
Profiles. Save a complete setup — EQ, gain, balance,
correction, notch, comfort settings — and switch between them, optionally
automatically when a given output device appears.
-
Command line and Shortcuts. A
sherlockeq CLI
bundled in the app talks to the running app over local IPC, with
--json output and scriptable exit codes; nine Shortcuts and
Siri actions cover profile switching, gain, balance, simple EQ, reference
mode, and status.
-
Help built in. Around two dozen articles covering every
feature, the reasoning behind it, and its limits — reachable in context from
the screen you're on.
Requirements and privacy
-
macOS 14.6 or later, Apple silicon and Intel (universal
binary). Core Audio process taps need Sonoma 14.2 at minimum; 14.6 is the
tested floor.
-
One permission: System Audio Recording — grouped under
Screen Recording on macOS 14, its own entry on macOS 15 and later. No
microphone access, no screen or video capture. Notifications are optional
and only used for safe-listening alerts.
-
No account, no telemetry. Profiles and settings stay in
~/Library/Application Support/SherlockEQ/, exports carry no
machine-identifying details, and nothing leaves your Mac unless you export
it yourself.
