Getting started
Where it goes in the chain
Put Faderhand on the vocal track after corrective EQ and de-essing, and before your main vocal compressor. It handles the phrase-to-phrase and word-to-word level, which is the job that otherwise eats all of a compressor's range.
Simple and Advanced
The plugin opens in Simple, which shows the controls most vocals need: Learn target, Target, Range, Speed, Bloom, Grid and Output. Advanced, in the header, adds everything else. Switching only hides controls: every setting keeps its value and keeps working, and the choice is saved with your project.
In Simple, Range moves Range Up and Range Down together. If a preset sets them apart, the readout shows both, for example 6/0. Turning the knob then sets both to the same value; to keep them apart, use Advanced.
Help while you work
Rest the pointer on any control, meter or the history strip for a moment and a short note explains what it does and how to use it. The notes, like this manual, are in Czech on a Czech system, in Spanish on a Spanish one and in English otherwise. The ? button in the header opens this manual.
Setting the target
This is the one setting worth getting right before you touch anything else.
- Press Learn target, at the top right of the Rider group.
- Play a representative part of the vocal. The button counts up while it listens; pauses and breaths do not count, so it needs about six seconds of actual singing.
- When it is done the button shows the level it found for a moment, and Target moves there. Press the button again while it is listening to cancel.
A target far from the vocal's natural level makes every other control behave oddly, because the rider spends its whole range just getting there. It is also the usual reason the plugin seems to make everything louder: at the default −18 dBFS, a vocal recorded at −24 is lifted by the full Range Up.
The learned level is where the vocal already sits, so the rider evens it out without making it louder or quieter overall. Move Target from there if you want it to.
Presets
The Preset menu in the header holds twelve factory starting points. Standard matches the defaults. They all ride quickly and listen well ahead, because a late gain is what sounds unnatural: it arrives after the phrase has changed, and a quiet passage's boost is still there when a loud one starts. They differ in how far they move, how quickly, and how much of the vocal's own dynamics they leave alone.
| Preset | Up / Down | For |
|---|---|---|
| Gentle | 4 / 4 dB | The lightest touch that still follows the vocal |
| Standard | 6 / 6 dB | The defaults; most vocals |
| Warm Lead | 6 / 6 dB | Standard with Bloom and Grid, for warmth and presence |
| Lift Quiet Phrases | 6 / 0 dB | Only brings up what gets lost; never cuts |
| Tame Loud Phrases | 1 / 6 dB | Mostly pulls down what jumps out |
| Pop Lead | 8 / 8 dB | An even, forward lead vocal |
| Ballad | 6 / 4 dB | Keeps a dynamic delivery dynamic |
| Rap & Spoken | 6 / 6 dB | Fast, even diction |
| Podcast & Voice-over | 8 / 6 dB | Speech; the gain drifts back to 0 dB in pauses |
| Backing Vocals | 4 / 4 dB | Backing vocals that should stay behind the lead |
| Upfront | 10 / 10 dB | Tight and forward, for dense mixes |
| Live Vocal | 8 / 6 dB | Stage microphones: higher idle threshold, 8 ms lookahead |
Live Vocal trades evening for latency. With 8 ms of lookahead the gain arrives about 100 ms after the vocal, so a sudden loud phrase still comes through. That is the price of monitoring through the plugin while singing; in a mix, use any of the others.
A preset sets how the rider moves and the colour. It never changes Target, so a learned target stays where it is, whichever you do first. Presets also leave Output, Ride / Read, Bypass and peak safety alone.
A starting point
The defaults are deliberately conservative. Leave Bloom and Grid at zero until the level riding sounds right.
| Control | Start at | Why |
|---|---|---|
| Target | Learn target | Starts from where your vocal really sits |
| Range (Up / Down) | 6 / 6 dB | Enough to even a vocal, small enough that mistakes are audible |
| Speed | 40 dB/s | Fast enough that the gain stays with the vocal |
| Tolerance | 0 dB | The rider aims at the target the whole time |
| Lookahead | 100 ms | Moves before the phrase, not after it |
| Detector | 50 ms | Follows phrases without chasing the waveform |
Activation
Faderhand needs to be activated once with the licence key from your purchase e-mail. Until it is, the activation panel covers the controls and audio passes through Faderhand unchanged, so a session never breaks.
- Open Faderhand in your DAW, or open the Faderhand app.
- Copy the licence key from your e-mail and click Paste. Some DAWs do not pass typing through to plugin windows, which is why the key is pasted rather than typed.
- Click Activate. This needs the internet once; after that Faderhand works offline.
Activating once is enough for everything on that computer: the plugin in every format and the app share the same licence. One licence covers up to 3 computers.
Moving to another computer
Click the Baracca Records logo at the top right to see who the licence belongs to and this computer's ID. Deactivate this computer frees its place, so you can activate another one.
Could not reach faderhand.com - check the internet connection and any firewall. If your DAW blocks plugins from the network, activate in the Faderhand app instead; the plugin picks the licence up.
Already active on 3 computers - deactivate one of them first. If a computer was lost or reinstalled before you could, write to info@baraccarecords.com with your licence key.
Reading the display
The history strip shows the last ten seconds and answers most questions about what the rider is doing.
- Input level - the filled taupe area, in dBFS, on the upper lane. This is what the detector hears, before any riding.
- Effective target - the dashed teal line. Normally flat at whatever Target is set to. It moves only when a sidechain is driving it.
- Applied gain - the amber line on its own lane below, from −12 to +12 dB. This is the move the plugin is making.
The two lanes exist because the quantities are in different units. Input level and target are absolute levels in dBFS; applied gain is a relative move. Drawing them on one axis would misrepresent one of them.
On the right are two meters.
- Gain - the same applied gain as a bar, with zero in the middle. Amber above the line is a boost, teal below it is a cut.
- Out - the loudest sample leaving the plugin, from −24 to 0 dBFS. The dashed amber line is the Ceiling. Anything above it turns red.
Above the Out meter, LIM −1.2 means the limiter just pulled a peak down by that much. CLIP means a sample went past 0 dBFS; it stays lit until you click the meter, so a single over cannot slip past unseen.
The amber line should drift, not twitch. If it is moving on every syllable, Speed is too high or Tolerance too low. If it steps between phrases and then sits still, that is exactly right.
Rider controls
The level the rider aims the vocal at, measured as RMS. Everything else is in service of this. Set it with Learn target.
Range Up and Range Down on one knob. Switch to Advanced when a vocal needs more room one way than the other.
The most the rider may boost. Set it by how quiet your quietest usable phrase is. Too much and room tone comes up with the vocal.
The most the rider may cut. Usually needs less than Range Up, because shouted phrases are rarer than quiet ones.
How fast the gain may move. This is the control that decides whether the rider sounds like an engineer or like a compressor.
Cuts happen twice as fast as boosts. A loud phrase caught late is more obvious than a quiet one lifted late, so the rider is allowed to duck at twice the rate it lifts.
Above roughly 25 dB/s it starts levelling individual syllables, which flattens the performance and dulls consonants. If you want that, use a compressor - it is better at it.
A deadband around the target. While the vocal is within this much of the target, the gain does not move at all. Raise it if the rider fidgets; lower it if the level wanders.
Idle
Between phrases there is no vocal to ride - only breaths, headphone bleed and room tone. Without an idle rule the rider would treat those as a very quiet vocal and boost them hard. Both controls are in Advanced.
Below this level the rider stops moving and holds its gain. Set it above your breaths and below your quietest real singing.
It leaves idle only once the signal is 3 dB above this threshold, so a level hovering right at the line does not chatter in and out.
At zero the gain simply holds through a pause, which is what you usually want. Set a time and the gain instead drifts back toward unity over that many seconds while idle, so a long instrumental section does not start with whatever the last phrase needed.
That is Idle Threshold set too low: the breath is above it, so the rider treats it as a phrase and lifts it. Raise the threshold until breaths sit below it.
Detector
All three controls are in Advanced.
The plugin measures the signal before it delays it, so the gain is already moving when a loud phrase arrives. This is why the rider can be slow and still catch phrase starts.
Too much and you hear the level dip before the word - a hole in front of the consonant. Too little and the word starts loud and gets pulled down a syllable in.
It sets the plugin's reported latency, which your DAW compensates for. See Latency.
How long a window the level is averaged over. Short windows follow syllables; long windows follow phrases. If the gain moves late, try shortening this before reaching for Speed.
Filters the detector's input to the ITU-R BS.1770 loudness curve, so the rider responds to perceived loudness rather than raw energy. Leave it on for voice. Turn it off if you want the rider to track absolute RMS, for example when matching a measured level.
Sidechain
Feed the backing track into the sidechain input and the rider will ride the vocal relative to the music instead of to a fixed number - up in loud choruses, down in sparse verses. Both controls are in Advanced.
How much the music moves the target. At 0 the sidechain is ignored entirely. At 1 the target follows the music one for one: music 6 dB louder, target 6 dB higher. Around 0.3 to 0.5 is a usable starting point.
The music level at which the target is left alone. Set it to your backing track's typical level, then Target still means what it says.
With Sensitivity up and no signal actually reaching the sidechain, the plugin sees silence, concludes the music has gone away, and pulls the vocal down by the full Range Down. If the vocal vanishes when you enable the sidechain, check that the bus is really feeding it.
Colour
Two stages after the rider, both off by default. They sit after the gain, so their level follows the ridden vocal rather than the raw one.
Adds a second harmonic and nothing else - no clipping, no odd harmonics. At 100 the harmonic sits about 18 dB below the fundamental.
The amount is normalised against the signal's own envelope, so a quiet phrase gets the same relative harmonic as a loud one. It does not get dirtier as you sing harder.
A fixed presence lift from two broad bands at 2 and 3 kHz.
The bands are wide and overlap, so the lift is bigger than the two band figures suggest: at 100 the combined curve peaks around +4.3 dB near 2.7 kHz. Treat 100 as a lot, not as a nudge.
Output and peak safety
A vocal's peaks sit far above its average level - often 15 dB or more. A rider that lifts the average by 6 dB lifts the peaks by 6 dB too, which is how a vocal that was nowhere near clipping ends up over. Peak safety stops that in two ways. Ceiling and Limiter are in Advanced; in Simple, peak safety stays on at −1 dBFS.
Trim after everything, before the limiter. Use it to match bypassed and active levels so you are judging the riding, not a level change.
The loudest sample the plugin will let out.
First, the rider will not boost a phrase further than its loudest peaks allow: if a peak is already at −4 dBFS and the ceiling is −1, the boost stops at 3 dB however quiet the phrase is. Cuts are never held back.
Second, a limiter at the very end catches whatever still gets past - a peak that was already over in the recording, or one the Output trim pushed up. It looks 2 ms ahead, so it turns the peak down before it arrives rather than clipping it.
Switches both halves of peak safety. Leave it on. Below the ceiling it does nothing at all - the audio passes through untouched - so it costs nothing in sound.
If LIM shows often and deep, the limiter is doing a compressor's job. Lower Output or Range Up rather than leaving it to cope.
The limiter only touches individual peaks above the ceiling and lets go within a tenth of a second. It will not make a vocal sound denser. Keep your vocal compressor after Faderhand.
Recording the ride
The rider's moves can be written to your DAW as ordinary track automation, then played back and edited by hand. That is what Ride and Read in the header are for. The Ride knob itself is in Advanced.
Ride mode
The default. The plugin computes the gain and writes it to the Ride parameter thirty times a second. The Ride knob is dimmed here because the plugin owns it - dragging it would just be overwritten.
Recording it
- Leave the plugin in Ride.
- Set the track's automation mode to Touch or Latch (Logic), arm automation (Live), set the envelope to Write (Reaper), or set the track to Write (Pro Tools, after enabling Ride for automation as described below).
- Select the Ride parameter in the track's automation lane.
- Play the song through. Writing only happens while the transport is rolling, and never while the plugin is bypassed.
- Switch the plugin to Read and the automation lane to Read.
Read mode
The rider stops computing and the applied gain follows the Ride automation instead. The knob becomes a normal control, so you can redraw any part of the curve. Edits still pass through the same smoothing, so a hand-drawn step will not click.
Moving the ride to Clip Gain (Pro Tools)
In Pro Tools the recorded curve can become the clip's own gain line, which then travels with the clip when you move or copy it. Clip Gain works before the plugins, so Faderhand's limiter still sits after the gain and keeps guarding the peaks.
- Show the Faderhand > Ride lane, select the part you want (or the whole lane) and copy it with Cmd+C.
- Switch the track view to Clip Gain, keep the same selection and choose Edit > Paste Special > To Current Automation Type (Ctrl+Cmd+V). The values carry over in dB: +6 dB of Ride becomes +6 dB of Clip Gain.
- Switch the plugin to Read, delete the Ride automation, and double-click the Ride knob (Advanced) to set it back to 0 dB.
The last step matters. In Ride mode the plugin would ride the already corrected vocal again, and with the Ride automation still in place in Read mode the gain would be applied twice.
Logic - leaving automation on Read while recording is the usual reason nothing gets written. It must be Touch or Latch. The parameter sometimes appears in the lane list only after you close and reopen the plugin window.
Live - parameters do not appear in the automation lane by themselves. Click Configure on the device, move the Ride knob once, and it will show up.
Reaper - check you are not in trim/read mode, where written values are added to a manual offset rather than replacing it.
Pro Tools - Pro Tools writes no automation for a plugin parameter until you enable it. Click Auto in the grey bar above the plugin window, move Ride to the list on the right with Add, and click OK. The usual Ctrl+Opt+Cmd-click on a control does not work on Faderhand's panel. Also check that Plug-in is lit in the Automation window (Cmd+4), otherwise no plugin automation is written at all.
Working with a file
The Faderhand app can open a recording, play it through Faderhand while you set it up by ear, and export the result as a new file, all without a DAW: a podcast, a voice-over or a demo. The plugins in your DAW have no player; there the track simply plays through Faderhand.
- Click Open… in the player bar under the header and choose the recording. On a Mac you can also drag the file from Finder onto the window. Faderhand opens WAV, AIFF and FLAC, mono or stereo. While a file is open it replaces the audio input.
- Press play, or the space bar. Click or drag the position bar to jump, and turn on Loop to keep a passage going while you work on it.
- Set Faderhand up as you would in a DAW: press Learn target while a representative part plays, choose a preset and adjust the controls.
- Switch between Original and Faderhand to compare. It is the same switch as Bypass, so the two line up without a jump. Set Output so both sound equally loud before you judge.
- When it sounds right, click Export… and choose where to save. Faderhand suggests the original name with (Faderhand) added. A panel shows the progress, and Show in Finder (Show in Explorer on Windows) takes you to the result.
Export processes the whole file from the original at full quality, with the settings on screen, in Ride mode and never bypassed. The result has the same length, sample rate, number of channels and bit depth as the original and lines up with it sample for sample, so it can go straight back into a project. The original is never changed, and a cancelled or failed export saves nothing. Moving the controls during an export does not affect the file being written.
The × next to the file name closes the file and goes back to the audio input.
Sidechain needs a separate music signal, so Music Sensitivity has no effect on a file played or exported in the app.
Latency
Lookahead is real delay, and the plugin reports it so the DAW can compensate, together with the limiter's fixed 2 ms. At the default 100 ms that is 4896 samples at 48 kHz, or 4498 at 44.1 kHz - about 102 ms. The limiter's 2 ms is reported even when the limiter is off, so switching it never shifts timing.
The lookahead is long on purpose: it is what lets the gain move before a phrase arrives instead of after it. If you sing through Faderhand live, the Live Vocal preset drops it to 8 ms, and you can set it anywhere between.
To confirm your host is compensating: duplicate the vocal, put Faderhand on one copy with Range Up and Range Down at zero, and flip that copy's polarity. The two should cancel to silence. A comb-filtered ghost instead means the delay is not being compensated.
Some hosts only apply a latency change after the transport stops. If timing drifts after you move Lookahead mid-playback, stop and start again before concluding anything is wrong.
If something is wrong
Target is above the level the vocal was recorded at. Press Learn target - see Setting the target.
The plugin is in Simple. Switch to Advanced in the header.
A sample passed 0 dBFS. With the limiter on, nothing gets past the Ceiling, so this means the limiter is off. Switch it on, or lower Output. Click the meter to clear the flag.
Idle Threshold is below your breaths. Raise it until breaths fall under it and the gain holds through them.
Too much Lookahead. Bring it down toward 10 ms.
The detector is reacting late. Shorten Detector before increasing Speed - Speed is the blunter tool.
Speed is too high. The rider should be inaudible as a rider; you should notice only that the vocal sits still.
Nothing is reaching the sidechain bus. See Sidechain.
Idle Return is set. At zero the gain holds through pauses instead of drifting back.
Where the files are
On a Mac:
/Library/Audio/Plug-Ins/VST3/Faderhand.vst3
/Library/Audio/Plug-Ins/Components/Faderhand.component
/Library/Application Support/Avid/Audio/Plug-Ins/Faderhand.aaxplugin
/Applications/Faderhand.app
/Library/Application Support/Baracca Records/Faderhand/Faderhand Manual.html
On Windows:
C:\Program Files\Common Files\VST3\Faderhand.vst3
C:\Program Files\Baracca Records\Faderhand\Faderhand.exe
C:\ProgramData\Baracca Records\Faderhand\Faderhand Manual.html
If the plugin does not appear, have your DAW rescan. Logic scans Audio Units at launch, so it may need restarting; so does Pro Tools, which only looks for plugins when it starts.