HoRNet ZeroLevel
User Manual

Welcome

More consistent levels, without changing the tone

Think of ZeroLevel as a level assistant for your mix. It listens while your track plays, then adjusts track volumes to make them sit more consistently together. You choose the tracks and the style; ZeroLevel takes care of the levels.

One important thing: ZeroLevel applies one global gain. It does not EQ, compress or gate your audio, so each track keeps its character.

Learn the whole track

Start playback at the beginning and stop at the end. Learn starts its analysis automatically when you stop.

Match tracks together

Instances with the same Group ID coordinate automatically as soon as every member has finished Learn.

Fine-tune whenever you like

Choose a group preset or a source-type offset. You do not need to Learn again.

Your project keeps its work

Valid Learn results are saved with the host project, so you normally do not need to Learn again after reopening it.

Quick Start

Four steps to your first match

1
Insert ZeroLevel on the tracks you want to match.
Leave Group ID at 1 or assign the same number to tracks you want to coordinate.
2
Check Track Type.
Auto is a good starting point. If the detected source is wrong, choose a category yourself.
3
Play the whole track, then stop transport.
Instances capture audio during Play and analyze the accepted material after Stop.
4
Wait for the group to become ready.
Gain is applied automatically when every instance in the group has finished Learn. There is no Apply button.

A helpful tip: for a mix, start playback at the beginning and let the tracks in the group play. A short audition may not represent the whole song.

The Main Window

The controls are labelled in English. This example shows the real interface with illustrative names and values, not a particular user's session.

HoRNet ZeroLevel main window: track type and Learn on the left, Output fader in the middle, group and offset controls on the right
The main window. The left side shows the track, Learn status and threshold; Output is in the centre; group, preset, source offset and reset controls are on the right.

This instance

Shows the track name and its selected or detected source type.

Learn & levels

Shows Learn status, playback and captured time, plus the analyzed level and gain.

Output

This is the global gain applied to the track. Changes are smoothed to avoid abrupt level jumps.

Group controls

Choose a group and preset, adjust a source type, or reset Learn.

Analyzed LUFS / gain shows the level of the band used for analysis and the applied gain. It is not a full-mix LUFS reading. Together, the two numbers give you a useful picture of how the level has moved.

The Output fader is also exposed to your host. If you want ZeroLevel to handle automatic matching, avoid writing automation to Output Gain: a new group plan may update it.

How Learn Works

Play from the start, stop at the end

ZeroLevel starts capturing audio while transport is playing. When you stop, it analyzes what it heard during that playback. Pauses and silence do not count as captured material; loops and repeated sections during Play are included.

Analysis begins after Stop, so you may see Learning — on Stop, Queued or Analyzing track. With long tracks or many instances, analysis after Stop may take a few minutes. Hiding the progress screen does not stop Learn.

Learn progress overlay showing learned instance count, independent groups, and Capturing, Queued and Applied states
The progress screen. Its counter shows how many instances have finished Learn—not song progress or an estimate of time remaining. The highlighted group belongs to the open instance.

Learn Threshold: Choose What to Capture

Learn Threshold sets how strong the peak of an incoming audio block must be before that block is included in Learn. The default is −50 dB; the range is −70 to −10 dB in 0.5 dB steps.

If a very quiet source is not being captured, try lowering the threshold. If you want to leave low-level background noise out, try raising it—but any sound above the threshold can still be included. This is not a noise gate and does not change the output level.

The threshold affects future capture only. To start the analysis over, use Reset This Instance or Reset Group Learn.

Track Type and Sources

Auto or choose a type yourself

In Auto, ZeroLevel tries the track name first and, if no useful name is available, falls back to audio classification. Name recognition depends on what your host provides. If the type shown in parentheses is not right, choose one from the menu. A manual choice—including Other—always takes priority.

Changing Track Type does not require a new Learn. A ready group recalculates automatically.

Groups: Let Them Match Automatically

There are eight Group IDs. Every instance belongs to its selected number; the default is 1. Give the same ID to tracks you want to coordinate. Once all of them have finished Learn, ZeroLevel applies the match automatically. A group includes instances in the same host process; a plug-in running in a separate process remains independent.

A silent or unfinished track remains part of its group and can hold up matching. Move it to another Group ID or remove the instance to leave it out. Other groups keep working independently.

Instances view with track names, Track Type, Group ID, Learn status, time and analyzed level
Instances: quickly check each track's group and whether its Learn is Applied, Learned or still in progress. You can also change another instance's Track Type here without opening its editor, when both instances are in the same host process.
8 groups view showing readiness, preset and state for each ZeroLevel group
8 groups: the counter shows how many instances are ready out of the total in that group. The state tells you whether Learn is still pending or the gain has been applied automatically.

Group Presets

A preset is shared by all instances with the same Group ID. You can change it after Learn and the ready group updates automatically.

PresetWhat it does
BalancedThe default balance, with no extra emphasis.
Vocal UpfrontBrings lead and backing vocals forward.
Rhythm FocusEmphasizes kick, snare, claps, bass and Drums.
Guitar HeavyEmphasizes guitars.
Synth FocusEmphasizes synths and keyboards.
Acoustic / Keys FocusEmphasizes guitars and keyboards.
Backing LiftLifts backing vocals relative to Balanced.
InstrumentalReduces lead and backing vocals.
Low-End FocusEmphasizes kick and bass, not snare.

What about buses? Treat them as a separate stage. First let the track groups finish Learn and matching, then insert ZeroLevel on the bus and give it another Group ID (it also works when the bus is the group's only instance). ZeroLevel does not know your project routing and will not reset a bus when upstream tracks change.

Group Type Offset: Your Source-Level Tweak

Group controls with Group ID, preset and a plus-three-decibel Drums offset
Group offset. Here the control is shown at zero for Group ID 4.

The Group type offset menu selects which source type you are adjusting. Its knob and numeric field set a correction from −12 to +12 dB, shared by every track of that type in the selected group.

For example, an offset for Drums adjusts only instances identified specifically as Drums in that group. It does not automatically include Kick, Snare or Claps. Auto tracks use the category ZeroLevel recognized.

The offset is added after automatic matching. Changing it does not restart Learn and repeated edits do not stack. Double-click the knob to reset only the selected source type to zero.

Start Learn Again, Safely

Confirmation before resetting Learn for only the Lead Vocal instance
Check the scope before confirming. A local reset does not clear the other tracks' analysis.

The threshold does not clean up audio. It only decides which blocks take part in Learn. The plug-in does not mute quiet sections or remove background noise.

Frequently Asked Questions

I pressed Play, but the gain has not changed. Why?
That is expected: ZeroLevel analyzes the whole captured programme when you stop transport. Your group may also be waiting for another member. Check the progress overlay and the 8 groups tab.
Do I need to press Apply?
No. Matching is automatic. Preset, Track Type and offset changes also apply without a new Learn.
A group is still waiting. What should I check?
Open Instances & groups and look for an instance that is unfinished, silent or has failed Learn. Each instance stays in its Group ID until you move it or remove it.
The source type is wrong.
Choose Track Type manually. Select Auto later if you want ZeroLevel to identify it again. There is no need to Learn again.
Will the gain always reach the target?
Not always. Relative matching is limited to ±12 dB and total gain to ±60 dB. The plugin observes those limits when tracks are far apart.
Do I need to Learn again when I reopen my project?
Usually not: valid Learn results are stored with the host project. Reset Learn if the source material has changed or you want to analyze it again.

Controls at a Glance

ControlWhat it is for
Track TypeLeave Auto on or choose the source for this track yourself.
Learn ThresholdExcludes blocks whose peak is below the selected level; it does not control output volume.
Group IDChooses which set of instances is matched automatically.
Group PresetAdds a shared stylistic emphasis for the group.
Group type offsetFine-tunes one source type by up to ±12 dB in the group.
OutputShows and controls global gain; automatic matching can update it.
BypassTemporarily bypasses the processing.

The plug-in handles up to two audio channels and declares no added latency. Available plug-in formats depend on the package you received.