Measured 14 November 2026 on the Resonance & Balance 1.1 code (the version on the App Store), rendered offline on a Mac, against soothe3 1.0.5, Gullfoss Live 1.12.0 and Trackspacer 2.5.10 (AU).
We ran the same vocals and mixes through our plug-ins and through the desktop tools people know for the same jobs, and measured what came out. This comparison is audio in, audio out. Here is what we found, both ways.
Measured after release, for context. OneMove products are our own designs; other products are named only to compare, and were run as licensed copies at the settings listed. On other material and settings, results will differ.
How we compared
- Material: real vocals and mixes (our bench), plus singers we never tuned on (the held-out set), 20-second excerpts in the players. The music is our own.
- Rendering: every clip was rendered by us with licensed copies of each plug-in, at the settings listed below.
- Loudness: whole-mix and vocal versions are matched to the dry clip (RMS), so you judge tone, not level. Keyed versions aren't: making room changes the level, and that is the point.
- Settings:
| job | ours | reference |
|---|---|---|
| Harsh frequencies on a vocal | Resonance, default settings, Depth 2.5–10 | soothe3, default settings, the same Depth |
| Rebalancing a whole mix | Balance, Tame 50 + Recover 50, and Tame 100 + Recover 100 | Gullfoss Live, the same Tame and Recover |
| Making room for a vocal | Balance keyed by the vocal, Voice Duck 12 dB (the default) | Trackspacer 2.5 at 15 % and 30 % |
Resonance and soothe3
Resonance 1.1 and soothe3 on a harsh vocal
Both at their default settings, Depth 5. Loudness-matched to the dry clip (RMS).
"Stayin' at Home – Vocals" by mindmapthat (ccmixter.org/files/mindmapthat/61340), licensed CC BY 3.0 (creativecommons.org/licenses/by/3.0). Excerpt, rendered through each plug-in.
| measure (7 vocals) | Resonance 1.1 | what it means |
|---|---|---|
| Where it cuts, compared with soothe3 (shape correlation, Depth 5 / 7.5 / 10) | 0.76 / 0.93 / 0.96 | 1 = the same places, in the same proportions |
| Difference in cut (rms, Depth 2.5 / 5 / 7.5 / 10) | 0.18 / 0.56 / 0.60 / 1.10 dB | how far apart the two cuts are |
| How much less it cuts at Depth 10 | 0.7 dB | on average, across the spectrum |
| Same cut at 0 dB and at −20 dB input (Depth 10) | within 0.05 dB (0.18 at most) | soothe3: 0.07 |
- soothe3
- Resonance 1.1
The values (dB)
| 350 | 700 | 1.4k | 2.5k | 3.5k | 4.5k | 6k | 8.5k | 13k | |
|---|---|---|---|---|---|---|---|---|---|
| Resonance 1.1 | +0.5 | +0.6 | -0.1 | -1.7 | -3.6 | -1.8 | -2.0 | -3.9 | +0.2 |
| soothe3 | +0.5 | +0.5 | -0.7 | -2.7 | -3.9 | -2.0 | -3.3 | -6.2 | -1.2 |
- What soothe3 offers that Resonance doesn't: more controls for shaping the detection (Sharpness, Selectivity), a longer track record, and more ways to see what it does. At Depth 10 it still cuts a little more.
- What Resonance offers that soothe3 doesn't: it runs on iPad and iPhone, with zero latency, and can be keyed by your vocal through Key Link with no side-chain routing.
Balance and Gullfoss Live
Balance 1.1 and Gullfoss Live on a mix
Both at Tame 50, Recover 50. Loudness-matched to the dry clip (RMS).
Vocal: "Lost_Pella" by queeniemusic (ccmixter.org/files/queeniemusic/65508), licensed CC BY 3.0 (creativecommons.org/licenses/by/3.0). Music: OneMove Audio. Excerpt, rendered through each plug-in.
| measure (step test, Tame 100) | Balance 1.1 | Gullfoss Live |
|---|---|---|
| Cut on the loud band | −7.2 dB | −6.1 dB |
| Reacts (63 % of the change) | 40 ms | 29 ms |
| Lets go (63 %) | 90 ms | 89 ms |
| Lift elsewhere (make-up) | +2.8 dB | +4.8 dB |
| Level change on mixes, Tame/Recover 50 / 100 | −1.8…−2.1 / −2.9…−3.5 dB | −1.2…−1.5 / −1.8…−2.2 dB |
- Where Gullfoss Live differs: it reacts faster, lifts what is hidden more, and changes the overall level less at full settings. It also offers more modes in its larger editions.
- Where Balance differs: it runs on iPad and iPhone with zero latency, and it can be keyed by a voice (Voice Duck), which Gullfoss has no input for.
Making room for a vocal: Balance and Trackspacer
Making room for the vocal: Balance 1.1 and Trackspacer
The music keyed by the vocal: Balance at its default (Voice Duck 12 dB), Trackspacer at 15 % and 30 %. The vocal is untouched. Not loudness-matched: making room changes the level.
Vocal: "Lost_Pella" by queeniemusic (ccmixter.org/files/queeniemusic/65508), licensed CC BY 3.0 (creativecommons.org/licenses/by/3.0). Music: OneMove Audio. Excerpt, rendered through each plug-in.
| measure (bench / held-out singers) | Balance 1.1, Voice Duck 12 | Trackspacer 15 % | Trackspacer 30 % |
|---|---|---|---|
| Covered voice made audible | 60 % / 67 % | 52 % / 64 % | 87 % / 91 % |
| Over-ducking (cut beyond what the voice needs) | 2.0 / 2.0 dB | 10.2 / 7.5 dB | 27.6 / 28.9 dB |
| Cut where the voice isn't covered | 2.3 / 2.1 dB | 1.4 / 1.9 dB | 5.2 / 7.3 dB |
| Ducking in the pauses | 0.1 / 0.0 dB | 0.1 / 0.0 dB | 0.2 / 0.1 dB |
| Music lost overall | 1.7 / 1.7 dB | 1.2 / 1.4 dB | 3.0 / 3.5 dB |
- Where Trackspacer differs: at 15 % it cuts less where the voice isn't covered and takes less out of the music overall; at 30 % it frees more of the voice, at the cost of deep over-ducking. It works on desktop hosts with side-chain routing, with a long track record.
- Where Balance differs: it frees more of the voice than Trackspacer at 15 %, with about a quarter of its over-ducking. It runs on iPad and iPhone, keyed through Key Link in hosts that can't route a side-chain, such as Loopy Pro.
What the tools look like
Their makers publish no terms for reusing screenshots, so we describe them instead. soothe3 has a large spectrum display with the cut drawn over it and a big Depth knob; Gullfoss Live has two main sliders, Recover and Tame, plus Bias, Brighten and Boost; Trackspacer has one main Amount control and a side-chain input. Our own panels are at the top of this page, and in full on the Resonance & Balance page.
What we haven't measured
- CPU of the desktop tools, and any of these tools on iPad or iPhone.
- Material beyond our bench and held-out singers: dense finished masters, speech, instruments on their own.
- Sound is personal: the numbers say where the tools differ, the players let you decide.
Credits
- "Stayin' at Home – Vocals" by mindmapthat, ccmixter.org/files/mindmapthat/61340, CC BY 3.0.
- "Lost_Pella" by queeniemusic, ccmixter.org/files/queeniemusic/65508, CC BY 3.0.
- The tables also use other singers from our bench; their recordings are only measured, not published here.
- Music: OneMove Audio.
soothe is a trademark of oeksound Oy. Gullfoss is a trademark of Soundtheory. Trackspacer is a trademark of Wavesfactory. They are named only to identify the tools we compared with, measured black-box as installed. OneMove Audio is not affiliated with or endorsed by any of these companies.

