
Room tone from the take
you already have.
Load a dialogue take. Undertone 2 picks out the parts where nobody speaks and resynthesises them into room tone that runs as long as the scene needs. Load more sounds and build whole soundscapes from them.
The first 13.4 seconds are the original take, the same in A and B; the replacement lines run from there to the end. The loop repeats the last 6 seconds.
The new line is clean. That's the problem.
Put the room back. Same take, same air.
This clip did not load.
AAS CUT
BWITH UNDERTONE 2
Replacement lines recorded in a clean room, cut into the original take. In A the background drops out between the lines. In B, a bed that Undertone 2 made from the original recording runs under them. Both clips play at their published level. Headphones help. The full walkthrough
Undertone is my favourite tool for creating dialogue fills. It sounds more natural than anything else I have tried.
Dustin Harris, Supervising sound editor · . The Handmaid's Tale, The Expanse, Locke and KeyMore from colleagues
One window. Two engines, 10 slots.
Load up to 10 sounds, shape them with two engines that always run together, and set the balance between them with one knob.
Point at an area, in this list or on the window.Tap an area, here or on the window.
Top barPower, the Preset Browser and the name of the loaded preset, the Sample Editor, the preferences, the switch to the mini view, and the arrow that shows or hides the side panel.
Side panel: HistoryHistory keeps the settings of your last eight renders, named by the time or by your clip name: click one to load it back, and rename, pin or clear it. The Panner tab places each slot's grains in the room: see the Panner chapter.
SeedEvery grain follows this string. Lock it and the grains repeat; the arrows draw a new seed. On the right, the name your rendered clips get.
Engines mixerSINGLE on the left, MULTI on the right, and the knob between them sets how much of each you hear. Single has GRAINS LEVEL, IR LEVEL and a knob that blends the two; Multi has GRAINS, LEVEL and the PANNER switch. Click SINGLE or MULTI to show that engine's controls below.
Single, in detailFor the grains: RANDOM LEVEL, RANDOM PITCH, REVERSE and SINGLE BAND or multi-band. For the convolution: IR LENGTH, IR PITCH and the NOISE PROFILE. A high-pass and low-pass filter for each, and the active slot's waveform and name.
Multi, in detailFor the selected slot: LEVEL, PITCH, REVERSE, RANDOM LEVEL, RANDOM PITCH, GRAIN SIZE, ENVELOPE and PLAY FROM START. Every slot keeps its own.
10 slotsTen tabs and their rows. Drag sounds in from the Finder, Soundminer or DragOut 2, or capture the timeline selection in AudioSuite: up to 30 seconds per slot. Double-click a waveform to open the Sample Editor.
Slot list, in MultiIn MULTI, each row adds S and M to solo and mute the slot, its PROBABILITY, the odds it is picked for a new grain, its GRAINS LIMIT, how many of its grains play at once, and its REST TIME between grains.
Bottom barS, M or L for the window size, WAVEFORM ZOOM, A/B to compare with the audio on the timeline, PARAMETERS AREA, UNVOICED AUTO SELECTION, and DRY and OUT, which set the balance with your track's audio.
No room tone recorded? Use the take.
Load a dialogue take and the parts with nobody speaking are picked out before you touch anything. They are ordinary selections in the Sample Editor: move them, resize them or add your own. Only the selection feeds the engines, crossfaded so nothing clicks.
-
The take loads.
UNVOICED AUTO SELECTION runs on every load while it is on in the settings, or on demand from the bottom bar.
-
It selects where nobody speaks.
In the Sample Editor, the lighter spans are selected: nobody speaks there. The darker rest was left out: all of the speech.
The slot's capture button turns red,
→
: the slot holds a selection.
-
Hear what it picked, then the bed.
INThe takeThis clip did not load.
The two spans UNVOICED AUTO SELECTION picked, as in the capture above. SELECTION ONLY loops them, as the engines read them.The two spans UNVOICED AUTO SELECTION picks in this take: AFTER, above, shows them selected. SELECTION ONLY loops them, as the engines read them.
IN THE TAKE
OUTThe bed it made, no tweakingMonitor gain for this lane only. The bed is quiet: it is only the room.This clip did not load.
OUT THE BED IT MADE, NO TWEAKING
Rendered from Undertone 2 with the two spans above and default settings. Twice as long as the take; on a track it runs for as long as the transport does.
Our guide made its own bed from a longer selection of this take, then fixed a few high squeaks in it.
Two engines. Three ways to rebuild a sound.
Single and Multi share the same 10 slots and always run together, and one knob mixes them. Use one way, or blend all three.
Press play, then turn the knob.
At the plug-in's own level this is quiet: LISTEN +12 dB lifts it.
This clip did not load.
The waveform and the grains logged in this render did not load, so they are not drawn.
Traffic - L to R V2as saved
Two synths on one sound.
The granular synth keeps the bed alive and moving, close to the take it came from: right for ambiences with life in them, like keyboard clicks or chatter. The convolution synth holds on to the character of steady tones: machines, air conditioning, room tone. One knob blends the two.
The bed from our guide
The bed from our guide
The bed from our guide
This clip did not load.
How A and B were made
ADEFAULT
BTUNED
A: our guide's bed at default settings, from a longer selection of the previous chapter's take. B: the same take, grains kept below 1158 Hz, convolution above 505 Hz, mixed 61 / 39. In our guide's words, more true to the ambience.
10 sounds. Each with its odds.
Load up to 10 sounds and set how often each one plays. A background holds without a break while pass-bys come and go, each time a little different.
LOGGEDThe plug-in's slot list, with the grains each slot played in this render.
S solos a slot and M mutes it. A solo also mutes Single, as in the plug-in by default. Silenced slots stay drawn, dimmed.
At the plug-in's own level this is quiet: lifts it.
Background plays from Single; in Multi it is muted, as saved.
Mixer audio: Traffic - L to R V2 as saved, rendered from Undertone 2, seed locked to 665bce744bdc41648bb1cc4246de2400.
Grains that move through the room.
Each slot gets a position in the sound field, a random area around it and a distribution inside that area. Give it a path and its grains travel from the triangle to the circle as they play. Up to 50 grains, mono to 7.1.
All of it automates.
- Per slot, start and end
- Position X and Y, range X and Y, distribution X and Y
- Per slot
- Path, LFE, solo LFE
- Panner
- On and off, grains, level
153 parameters in all, each one a lane in your DAW.
This clip did not load.
LOGGEDEach ring is a grain of this render, where the plug-in placed it.
On, each grain travels its slot's path, left to right. Off, each stays where it starts.
PathOn
PathOff
This demo is a stereo track, so you hear the grains move left to right; on surround tracks the field also places them front to back. This preset's paths run along the front.
Render it again. Get the same thing.
Seed
Every grain Undertone 2 plays follows one seed. Lock it, and every render and preview plays the same grains for as long as no parameter and no sound changes. Locked, it also lets you print stems that, together, sound like the full render.
AudioSuite
Capture the timeline selection and render it at its exact length: one continuous file or clip by clip, with handles and your clip name. The last eight renders stay in the history; click one to get its settings and audio back.
On a track
In real time it follows the transport. With the seed locked, the same buffer size and the same start point, it plays the same grains each time. A/B lets you hear the track's own audio for a moment; Dry and Out set the balance, from −100 to +12 dB. Every parameter can be automated, Panner paths included; the sounds and their selections, the seed and the names cannot.
Undertone 2 Overview
Chapters11
More than 150 presets, from air tones to drones.
33 air tones for dialogue fills: interiors and exteriors, cars and trains, quiet or with movement. Rain, wind, birds, traffic and engine rooms. Drones, chords and things harder to name. Each one is a full setup you can open, learn from and change.
In the Preset Browser: search, tags, notes, auto load, arrow-key auditioning, and loading only part of a preset: its slots, audio, Single, Multi or Panner settings. Undertone Lite presets open too.
- A Bonfire DF v2
- A Fireplace DF v2
- A Shower DF v2
- Adrian v2
- Air Tone 01 Int
- Air Tone 02 Int
- Air Tone 03 Int
- Air Tone 04 int
- Air Tone 05 Int
- Air Tone 06 Int
- Air Tone 07 Int Open Window
- Air Tone 08 Int w quiet noises
- Air Tone 09 Int w noises
- Air Tone 10 Int w quiet noises
- Air Tone 11 Int
- Air Tone 12 Int w AC
- Air Tone 13 Int w quiet noises
- Air Tone 14 Int w wood noises
- Air Tone 15 Int w Movements
- Air Tone 16 Int w mov Zipf Distr
- Air Tone 17 Ext w movs on dirt
- Air Tone 18 Int w ittle noises
- Air Tone 19 Int noisy
- Air Tone 20 Ext noisy
- Air Tone 21 Ext windy
- Air Tone 22 Ext heavy
- Air Tone 23 Ext
- Air Tone 24 Int
- Air Tone 25 Int
- Air Tone 26 Int w movements
- Air Tone 27 Int w more movements
- Air Tone 28 Noisy w lav shash
- Air Tone 29 Car Int
- Air Tone 30 Car Int
- Air Tone 31 Train Int w Noises
- Air Tone 32 Int quiet - Single Collection
- Air Tone 33 Ext loud - Single Collection
- Alien Drone MP v2
- Alien Engine Room DF V2
- Alone in the Dark DF v2
- Antartica v2
- Atonal Trance V2
- Bad Radio 1 MP v2
- Bad Radio 2 MP v2
- Birds And Swallows v2
- Birds Randomizer DF v2
- Busy Spaceport v2
- Cathedrale Engloutie RM v2
- Chaotic Clocks Ticking MP v2
- Clashing Swords MP v2
- Clock Panic MP v2
- Cloud of Yore MP v2
- Collapsing RM
- Corda DF v2
- Cosmic Balls v2
- Cracking Moon v2
- Creepy Tonal Swarm MP v2
- Datomir_REMIX DF v2
- Digital Lake MR v2
- Distant Karma V2
- Doom Event v2
- Dry Leaves in Wind MP v2
- Earthquake MP v2
- Energy Crickets v2
- Energy field v2
- Eraser Factory v2
- Fifth and octave
- Finale v2
- Fire Crackling MP v2
- Forest Fire MP v2
- Fourth min7
- Gentle Waves DF v2
- Geofone Laundry Drone 1 MP v2
- Geofone Laundry Drone 2 MP v2
- Ghost House v2
- Ghostly Wind 1 MP v2
- Ghostly Wind 2 MP v2
- Glass Drone MP v2
- Guitar Trill MP v2
- High Pitched Machine DF v2
- Humming Drone MP v2
- Infested Mall v2
- Lunar v2
- Magic Bells MP v2
- Medieval Village MP v2
- Metal Frogs MP v2
- Minor 6th Maj
- More than Ligeti v2
- Mother Breath v2
- Nebula v2
- Night Crickets MP v2
- Noisy Car Interior 1 Driving v2
- Noisy Car Interior 2 Driving, closed windshield DF v2
- Noisy Car Interior 3 Driving, open windshield v2
- Orion Strings v2
- Parking Lot DF v2
- Plymouth Fury v2
- Rain And Thunder v2
- Rain Dance v2
- Rain V2
- Random Balinese Gamelan MP v2
- Random Reverse Metallic Drone v2
- Relaxing Chord Storm v2
- Renaissance 1 V2
- Renaissance 2 V2
- Root dim
- Root Maj
- Root min
- Root min 7-9
- Root min7
- Root sus 4-7-9
- Rushing Wind R to L v2
- Setting Karma V2
- Signal from Mars 1 v2
- Signal from Mars 2 v2
- Sines 1 DF v2
- Sines 2 DF v2
- Sirens Call MP v2
- Something MP V2
- Spinner in the Rain v2
- Spiritual Door v2
- Stringularity 1 DF v2
- Stringularity 2 DF v2
- Stronger Wind in Mountain Sh v2
- Tension Interference MP v2
- Tonal Drone DF v2
- Tonal Engine Room DF v2
- Tons Of Insects v2
- Traffic - City V2
- Traffic - Highway rush hour V2
- Traffic - L to R V2
- Traffic - Pass by Only
- Traffic - R to L V2
- Traffic Heavy near highway L to R
- USO_Aleen v2
- USO_Datomir v2
- USO_Drahgor v2
- USO_Eadu v2
- USO_Exegol v2
- USO_Garel v2
- USO_Ibaar v2
- USO_Insectus V2
- USO_Kyma_Ambiences v2
- USO_Nakara v2
- USO_Nixus v2
- USO_Plastic_Fire
- USO_Transmission_Apparatus v2
- USO_Unseen_Noises v2
- USO_Vardos v2
- Vatican Square MP v2
- Voices V2
- Water Pump Drone MP v2
- Wind from nowhere v2
- Wind in Mountain Shelter DF v2
- Wind in Trees DF v2
- Wood Squeaks MP v2
Names exactly as they ship.
In their words.
Undertone is now an essential in my dialogue-editing toolbox – for creating a clean fill, a fill with movement in it, for continuing big noises such as planes or car noise which is often hard to fill, for elongating screams – you name it. And it is a must-have for creating DX sessions for M and E too. Strongly recommended!
Michael MaroussasSupervising dialogue editor · Chernobyl, Black Mirror, Darkest Hour
Undertone has become an invaluable tool in my effects editorial and design arsenal. In many instances, I am able to take a handful of seemingly ordinary sounds and create unique beds with extraordinary movement, characteristics, and, at times, emotion.
Andrew TwiteSound designer · Everything Everywhere All at Once, Star Trek: Discovery
The grown up kids at Audiokids have enclosed two of my desert island DSP tools – granulation and convolution – in a unique and friendly device. Once you have started using Undertone, it is addictive – primarily for the tiny but important tasks such as creating air fills for dialogue tracks. Watch out though: it is a powerful instrument to create complex soundscapes and atmospheres in an intuitive, simple interface.
Matteo MilaniSound designer and founder, Unidentified Sound Object
U.S.O. Libraries, which supplied some of the factory samples, is Matteo Milani's company. He is also credited for sound design in the factory presets.Three sound editors and mixers from Rome.
Undertone started as one job we kept doing by hand: cut room tone out of a take, loop it, and hope the loop did not tick. It is still the plug-in we open most.
Davide, Fabrizio and Mattia
-
Davide Favargiotti
Sound supervisor and dialogue editor. David di Donatello for Best Sound, Dogman; nominated for Call Me By Your Name and Freaks Out.
-
Fabrizio Cheloni
Re-recording mixer on Italian film productions. Writing Audiokids software full time since 2022.
-
Mattia Persico
Sound editor on Italian film productions. Sound designer in video games since 2021.
Photos by Irene Strazzeri.More about us
UNDERTONE 2 CREDITS
- DESIGN
- Davide Favargiotti, Fabrizio Cheloni, Mattia Persico
- CODING
- Fabrizio Cheloni, Mattia Persico
- MACHINE LEARNING
- Matteo Ciferri, Fabrizio Cheloni
- GRAPHICS AND LOGO
- Corrado Frontoni
- SOUND DESIGN
- Mattia Persico, Davide Favargiotti, Matteo Milani, Wideraudio, Riccardo Marsana
- FACTORY SAMPLES, UNDER LICENCE
- Free To Use Sounds, U.S.O. Libraries, Wideraudio, Gabriele Fasano, Riccardo Marsana
A big thank you to all our users and beta testers!
The numbers. As the plug-in sets them.
ENGINES
- Engines
- 2, Single and Multi, always running together
- Mix
- Single <-> Multi, 0–100
- Single
- Granular and convolution on the active slot, Granular <-> IR 0–100
- Multi
- Up to 10 slots, picked by probability
SOURCE
- Slots
- 10, shared by both engines
- Sample length
- From about 150 ms up to 30 s per slot; longer files: the first 30 s
- Channels
- Mono per slot: the first channel of a multichannel file
- Files
- WAV, AIFF, MP3, FLAC, OGG; any sample rate and bit depth
- Loading
- Drag and drop from the Finder, Soundminer or DragOut 2; capture from the timeline in AudioSuite
- Sample Editor
- Multiple selections with automatic crossfades; waveform and spectrogram; gain 0–3x
- Unvoiced Auto Selection
- Machine learning; on load or on demand; macOS
GRANULAR (SINGLE)
- Random level and pitch
- Per grain
- Reverse
- 0–100 % probability per grain
- Multi-band grains
- On or off
- Filters
- High-pass and low-pass, 20 Hz–20 kHz, 12 dB/oct
CONVOLUTION (SINGLE)
- IR length
- Up to 5 s
- IR pitch
- 0.25x–4x
- Noise profile
- Flat, Dark, Darker, Darkest
- Filters
- High-pass and low-pass, 20 Hz–20 kHz, 12 dB/oct
MULTI (PER SLOT)
- Probability
- 0–1
- Grains limit
- 1–50 at once
- Rest time
- 0–10 s, or 0–5x the grain length
- Pitch
- 0.25x–4x
- Grain size
- 1–100 % of the sample
- Envelope
- Linear fade in and out, 0–50 % of the grain
- Also
- Play from start, solo, mute
PANNER
- Track formats
- Mono, stereo, LCR, quad, 5.0, 5.1, 7.0, 7.1
- Per slot
- Position, random area, distribution, LFE send
- Path
- Start and end per slot, each with its own area and distribution
RECALL AND RENDER
- Seed
- Up to 32 characters; locked, every render plays the same grains
- AudioSuite
- Continuous file or clip by clip, with handles; clip name; history of 8 renders
- Real time
- Sync: Host or Free; A/B; Dry and Out −100 to +12 dB
- Automation
- Every parameter; not sample loading
- Presets
- 156 factory; tags, notes, partial loading
- Window
- Small, Medium, Large; main window and Preset Browser accessible with assistive devices
FORMATS AND SYSTEM
- Plug-in formats
- AAX, AudioSuite, VST3, AU; 64-bit only
- macOS
- 10.13 or later; Universal Binaries, Intel and Apple Silicon natively supported
- Hosts
- Pro Tools 12 or later; any DAW that supports VST3 or Audio Unit
- Licence
- An iLok USB key, v2 or v3, and an ilok.com account
- Includes
- Undertone Lite
- Windows
- The last Windows version, 2.2.0, is on the Downloads page; it is no longer updated or supported
Undertone 2 or Undertone Lite
Undertone 2 includes Undertone Lite. Lite reads only the first slot of an Undertone 2 preset.
See Undertone Lite| Feature | Undertone Lite | Undertone 2 |
|---|---|---|
| Single engine: granular and convolution | Yes | Yes |
| Sounds loaded at once | 1 | up to 10 |
| Multi engine and the Single <-> Multi mixer | No | Yes |
| Probability, Grains Limit, Rest Time | No | Yes |
| Panner and grain paths | No | Yes |
| Multi-band grains | No | Yes |
| Unvoiced Auto Selection | No | Yes |
| Render history (AudioSuite) | No | Yes |
| Track formats | mono to 7.1 | mono to 7.1 |
| Factory presets | 76 | 156 |
| Price | €119 | €219 |
A real session is the only honest test.
The Downloads page emails you the installer link. Install Undertone 2, open it on a track or in AudioSuite and click Try on the iLok screen: the 15-day trial starts at once. It needs an iLok USB key, v2 or v3.
Start the 15-day trial- Undertone 2Full version, includes Undertone Lite €219 Buy
- Upgrade from Undertone 1For Undertone 1 owners. The licence also covers Undertone 1.x. €89 Upgrade
- Crossgrade from Undertone LiteFor Undertone Lite owners €100 Crossgrade
VAT is added at checkout on EU orders.
Undertone Lite: one sound, one engine.
Granular and convolution from a single sound, for fills and ADR. No Unvoiced Auto Selection: you select the quiet parts of a take by hand. No Multi engine, no Panner.
Education
50% off for students in sound and music courses, teachers and institutions.
Request form (PDF)Studios
Custom quotes for three or more licences, for individual users or a shared studio iLok.
Write to usDo I need an iLok?
Yes: Undertone 2 needs an iLok USB key, version 2 or 3. Computer and iLok Cloud licences are not supported for it yet.
Is there a Windows version?
Undertone 2 is for macOS. If you own it, the last Windows version, 2.2.0, is still on the Downloads page: it runs with the same licence, but it is no longer updated or supported.
Does it support Dolby Atmos or other immersive formats?
Up to 7.1 today. Wider formats are not supported yet. We have had several requests and we are considering it: tell us your use case.