Scribe manual
Matching EQ and translation check. Scribe learns the tonal balance of a reference and pulls your material toward it, broadly and musically. Switch mode and the same plugin plays your mix the way a phone, earbuds, a laptop or a car would.
Warren Labs · Trueness line · VST3 · AU · Standalone · macOS
What it's for
Two questions that are really one: how does my tonal balance compare to something else?
- Match answers it against a record you trust. Play the reference, let Scribe learn it, then play your mix. Scribe works out where yours is heavier or lighter, band by band, and nudges it toward the reference.
- Audition answers it against the places your mix will actually be heard.
Scribe is a tonal fit, not surgical EQ. It measures in 20 broad, overlapping bands from 30 Hz to 16 kHz, so it moves balance (too much low mid, not enough air) and leaves individual notes and resonances alone. For those, use Bevel or Pare.
Until you teach it a reference, Match does nothing. A fresh Scribe is a wire: the output is the input, bit for bit, and the screen reads NO REFERENCE: HOLD LEARN. The Match controls (Amount, Tilt, Smoothing, Mix, Freeze, Clear) stay dimmed, reading —, until a reference is learned. Learn is the one control that is live from the start.
The controls
Mode
Match, or one of four audition voicings: Phone, Earbuds, Laptop, Car.
Learn
Hold it on while the reference plays, then switch it off. That's the capture. While Learn is on, Scribe passes audio through untouched so you hear the reference as it is. It averages over a couple of seconds at a time, so give it at least 20–30 seconds of the part of the reference that matters (the chorus, not the intro).
The learned reference is saved with your session. It is not part of a saved preset.
Clear
Discards the learned reference. Match goes back to doing nothing until you learn another.
Amount
0–100 %, default 60 %. How far toward the reference Scribe pulls. Around 80 % is a full match; 100 % goes slightly past it. Most of the time, 30–60 % is the useful range: you want the reference's balance, not its exact curve. No band is ever moved more than 15 dB.
In the audition modes, Amount blends the voicing in instead: 0 % is your mix, 100 % is the full caricature.
Tilt
−6 to +6 dB per octave, pivoting at 1 kHz, default 0. Leans the target darker or brighter on top of the reference. Match only.
Smoothing
0–100, default 50. Spreads each band's correction across its neighbours, up to four bands either side. Higher is broader and gentler; lower follows the reference's shape more closely. Match only.
Freeze
Stops Scribe re-measuring your material. The correction it has at that moment becomes a fixed EQ. Without Freeze, Scribe keeps listening and adjusts as the song changes, which is usually what you want on a whole mix and not what you want on a part that changes arrangement. Match only.
Mix
0–100 %, default 100 %. Dry/wet blend in Match. In the audition modes Amount does that job, so Mix dims and reads — there.
Power
Bypass, with a click-free crossfade.
The audition voicings
These are caricatures, voiced to exaggerate what each kind of playback does. They are not measurements of any particular device.
| Mode | What it does | Listen for |
|---|---|---|
| Phone | No lows, no highs, a boxy midrange bump | Does the bass line still read? Does the vocal? |
| Earbuds | Fairly full, a presence lift, a little low-mid dip | Harshness in the upper mids |
| Laptop | No bass, a tinny upper-mid lift | Whether the kick and bass disappear completely |
| Car | Boomy lows, scooped low mids, soft top | Low-end build-up; whether the mix turns to mud |
Audition modes are for listening only. Switch back to Match, or bypass Scribe, before you bounce.
The display
- Dim spectrum: your material, live.
- Amber curve: the correction being applied right now, per band. Flat at 0 dB until a reference is learned.
- Faint line: the learned reference's shape, when there is one.
- Dashed line: the tilt, when there is a reference.
- Top right:
NO REFERENCE: HOLD LEARN,LEARNING REFERENCE…orMATCHING REFERENCE. - In audition modes the plot shows the voicing curve instead, with a wider dB scale.
A worked example: matching a mix to a reference record
- Put Scribe on your mix bus, last. Import the reference into the session on a track routed through the same bus, and level-match it roughly with your mix.
- Solo the reference. Switch Learn on, play 30 seconds of its chorus, switch Learn off. The screen now reads
MATCHING REFERENCEand shows the reference's shape. - Mute the reference and play the same section of your mix. The amber curve shows where your balance differs.
- Start at Amount 40 %. Toggle Power at matched level. If the mix is closer to the reference but still sounds like your mix, you're in the right place.
- If the correction is too busy, raise Smoothing. If you like the balance but want it a shade brighter, nudge Tilt up.
- Then check it travels: Phone, then Car. Fix anything they reveal at the source, not with Scribe.
Presets
PRESETS saves and recalls every control. It does not store the learned reference. Files live in ~/Music/Warren Labs/Scribe/Presets/.
Licence
Drag the .wllicense file from your purchase onto the plugin window, or open the About panel (ⓘ, top right) and choose Import License.
Specifications
| Formats | VST3 · AU · Standalone |
| Platform | macOS 11+, Universal (Apple Silicon + Intel) |
| Channels | Mono, Stereo |
| Latency | None |
| Match | 20 log-spaced bands, 30 Hz–16 kHz · ±15 dB maximum per band |
| Audition | Phone · Earbuds · Laptop · Car |
Warren Labs. Tools that tell the truth. · warrenlabs.com