FAQ

Sign in to the Nolvra Hub, find SURGE under My Products, and hit Install. The Hub downloads the plugin, puts it in the right folder, and keeps it up to date. Rescan your plugins in your DAW and SURGE appears like any other plugin.

Yes. Nolvra plugins run on both macOS and Windows, the Nolvra Hub runs on both, and one license covers both. Your two activations can be any mix of Mac and Windows machines.

Yes. Activation needs a connection once; after that the plugin keeps a signed local record and keeps working offline for up to 14 days before it checks in again. Reconnect at any point in that window and the clock resets.

Yes. Anything you make with Nolvra plugins and sounds is 100% yours, royalty-free, including commercial releases. The only thing you can't do is repackage the raw sounds or presets as another sample pack. Details are in the Terms.

Yes. It is free with a Nolvra account rather than free for a trial period: no time limit, no feature locks, no card and no licence key. Create a Hub account, sign in, install it. Nothing on its face is dimmed and there is no upgrade button on it.

No, and there is nowhere to type one. The plugin has no activation field at all: it looks for the signed session the Nolvra Hub writes when you sign in, and it opens when that session is valid. This is the same door RELIC uses.

Your voice passes through untouched. With the gate shut the output is sample for sample identical to the input, measured across 882,000 samples with a largest difference of zero, so a session that loses its sign-in loses the effect and not the take.

PRISM MINI is the talkbox on its own: ten controls on its face, four factory presets, and a key you set by hand, a root and major or minor. PRISM PRO ships twelve factory presets and the parts that decide notes and finish them. Its pitch order has one more rung, MIDI, then the chord it hears on its sidechain, then key and scale, then the follower running free, against PRISM MINI's MIDI, then key, then free. It carries up to three harmony voices, a third below, a third above and an octave below, each quantised through the same gate as the lead. With SCALE set to AUTO it works the key out on its own, where PRISM MINI's key is always the one you set. And its wet path ends in a compressor and a chorus, which PRISM MINI does not have. PRISM PRO is the paid plugin's name from version 1.2.2 on; it was called PRISM up to that release.

DRY is the fourth knob in the MOUTH band, after FORMANT, GLIDE and SPEECH, and it arrived in 1.0.5. It sets how much of your own voice reaches the output underneath the talkbox.

It finds resonances, and it separates a resonance from a note by time rather than by loudness. That distinction is the whole reason it can work: a sung note is also a narrow peak standing well above its own neighbourhood, and across the 27 mixes, stems and loops the scan was calibrated on, the notes stood 13 to 37 dB above their surroundings, which is more than a deliberately planted ring managed. So the scan watches a five second window and reports a frequency only once it has been present through enough of it. What it misses follows from the same rule. A problem that only exists in one section is a section problem, not a resonance, so scan that section rather than the whole song. A ring shorter than the window will not earn a mark. And a scan that reports nothing is giving you an answer: turning Sensitivity to 100% to make it say something is asking it to be wrong. The search is bounded by controls you set. Sensitivity runs 0 to 100% and opens at 50%, the frequency window opens at 60 Hz to 16 kHz, and Max Count is how many marks you want back, one to eight, opening at four. Two sightings within a semitone of each other are treated as the same thing.

Two ways, and they answer different questions. One undo takes the whole thing back, however many bands APPLY wrote. It is one step, not eight. DELETE is the other one, and it is narrower on purpose: it removes the bands the scan made and leaves the bands you made alone. A band you dragged more than a semitone away from its mark counts as yours from then on, so DELETE will not touch it either. It is also worth knowing what APPLY writes before you decide it is wrong. The band is a bell at the mark's frequency with its static gain flat; the cut lives in the dynamic range instead, sized by how far the peak stood above its own neighbourhood and clamped between 1 and 12 dB. So the band does nothing while the frequency behaves and pulls down only while it rings.

The ribbon holds thirty minutes of the session as a fixed ring of columns, so a mark is a moment as well as a frequency: you can see whether a ring was there for the whole take or for one line. Three spans decide how much of it is on screen, LAST 60 S, LAST 240 S and WHOLE SESSION, and it opens on 240 seconds. CLEAR is a row on the ribbon's own menu, and it carries the count of what it is about, so it reads CLEAR followed by how many marks are standing and is greyed out at zero rather than being a button that does nothing. What it clears is the picture: the marks, the columns behind them and the lit reading. What it does not touch is the SCAN switch. SCAN stays on and starts its window again, so the first marks after a CLEAR arrive the way they did after you first pressed it.

No, and the plugin says so rather than letting you assume otherwise. This release implements zero latency, and that is the mode the audio runs in. PROCESS still lists NATURAL PHASE and LINEAR PHASE, and choosing one of them leaves the audio on zero latency while the parameter bar prints MODE FELL BACK, spelled out as ZERO LATENCY (FELL BACK) where there is room for the sentence. The reason the modes are listed at all is the same reason 453 parameters are published and locked from the first release: adding a parameter later would shift every automation lane written before it. Latency is on the bar in milliseconds either way, including when it is zero, because zero is a measurement somebody made and a blank space is not. If a linear phase pass is part of how you work, that is the line to read twice before buying.

Neither, and both are the platform rather than a missing build. Audio Units are an Apple format, so no plugin has an AU on Windows. RIDGE's Windows package carries the VST3 alone, into `C:\Program Files\Common Files\VST3`. On macOS you get three: the AU, the VST3 and a standalone app. The Hub puts the plugins in your home Library and the app in your home Applications folder, so no admin password is needed for any of them. One licence covers both platforms, and your two activations can be any mix of Windows and Mac machines.

Either of two ways, and one is enough on its own. A licence key for that product, which reads `RIDGE-XXXX-XXXX-XXXX` or `IMAGER-XXXX-XXXX-XXXX` and goes into the plugin's activation screen. Or an active All-Access membership on the account you are signed into in the Nolvra Hub, in which case the plugin recognises the signed-in session and you never type anything. A key owner does not need the Hub installed. A member never needs a key. Both routes activate on up to two machines, and the 7-day trial opens them the way a membership does, with no card and no separate activation step.

They are the two places where a fold-down to mono has cost you a known fraction of the power, and both are arithmetic rather than taste. PROMISE 3.01 is 10·log10(2), the point where half the power is gone. SEVERE 6.02 is 10·log10(4), where three quarters are gone. The reading between them is your mix's actual fold-down loss, measured continuously at the plugin's own output, and the state is printed as one word: WITHIN PROMISE below the first line, OVER PROMISE between them, SEVERE above the second. There is a fourth word, QUIET, and it is not a fault. Below a stereo RMS of -60 dBFS there is no reading worth showing, so the plugin says QUIET instead of printing a number about silence.

Only report it, and that is a decision rather than a gap. An earlier version carried an automatic guard that pulled the width down when a band went out of phase. It was measured out of the product because it moved the level most where it helped least: 3.84 dB of movement to buy 0.75 dB of fold-down on one mix, against 0.55 dB of movement where it bought 3.45 dB. A warning was judged better than a silent correction, and that decision is final. The cost of it is honest and worth stating: without the guard, this plugin can break its own 3.01 dB line on real material. It tells you while it happens, and what you do about it is yours.

Because widening the low end is where mono fold-down damage usually comes from, and that ceiling is the only real protection the plugin carries. The mid and high bands run 0.00 to 2.00. The low band runs 0.00 to 1.00, where 1.00 is the width the signal arrived with, so the low band can be narrowed all the way to mono and cannot be widened at all. The ceiling sits on the parameter rather than inside the engine. That means it is decided before any audio arrives, and your host, your presets and your automation lane all see the same limit: a wide automation curve cannot hit it silently.

No, and that is structural rather than a promise. The monitor section sits after the point where the needle taps the signal, so nothing in it reaches the reading. DELTA plays the output minus the same signal at unity width, so you hear exactly what the knobs did and nothing else; with the bypass engaged it is zero by arithmetic. MONO folds the output down with the same sum the needle measures. SOLO puts one band's mid or one band's side on its own: six taps, LOW M, LOW S, MID M, MID S, HIGH M and HIGH S, plus OFF. It is one choice and not six switches, so one solo at a time is a fact about the parameter rather than a rule the interface has to keep.

Yes to both. The settings menu carries a THEME section with a DARK row and a LIGHT row, and the geometry is identical in both: the same plate, the same boxes, the same positions, with only the tones swapped. The choice is saved in the plugin's own settings file rather than in the session, alongside the editor width and whether you have seen the quickstart. So it follows you to the next instance and the next project on that machine, and opening an old session does not drag an old theme back with it.

Seven factory presets, each named for the job rather than for the filter: SOLID BOTTOM, CENTRE HOLD, BUS LIFT, MASTER SAFE, SYNTH SPREAD, DRUM ANCHOR and FOLD RESCUE. INIT is the plain state to build from. Three of them set the low band at or near mono, which is the point the ceiling above is making. FOLD RESCUE is the one to reach for when the gauge is already saying OVER PROMISE or SEVERE and the bounce is due. The two crossovers are yours in every preset: the low one runs 40 to 400 Hz and the high one 800 Hz to 8 kHz.

Lower than most people put them, and the low one is the decision that matters. LOW CROSSOVER runs 40 Hz to 400 Hz and opens at 120 Hz. Its control is deliberately skewed so that single hertz are reachable near the bottom of the range, because that is where the mono decision is actually made. On a kick and sub record the line usually wants to sit under the sub's fundamental. On a full band mix it often ends up higher than expected, because a bass guitar's body and the low end of a room mic are both in play. HIGH CROSSOVER runs 800 Hz to 8 kHz and opens at 2500 Hz. It is the line between the body of the mix and its air, and moving it decides whether the cymbals and the top of a pad belong to the band you are opening. The factory shelf is the fastest way to see the range people actually use: its low lines sit between 110 Hz and 220 Hz and its high lines between 1.6 kHz and 4 kHz, and every one of those is a starting position rather than a setting. The crossover in a preset is as much a part of it as the width.

No, and the reason is arithmetic rather than a missing feature. Width here multiplies the side signal, which is what is already different between your two channels. A mono source has none: with the left and right channels identical, the side signal is zero, and any width leaves it zero. There is no delay stage, no Haas layer and no synthesis anywhere in the engine, so a stereoizer is not one of the things this plugin does. That is also why it cannot smear a transient or hollow out a centre the way a delay based widener can. What you can do is put it on the bus that carries the material which actually makes a vocal sound wide: the doubles, the harmonies, the stereo delay, the reverb return. And if the problem is a lead being crowded rather than a lead being narrow, the useful move is usually the other direction, pulling the band the voice lives in toward the centre while opening the air above it.

No latency: the plugin reports zero samples, active and bypassed alike, and the shipping gate measures it rather than trusting it. At width 1.00 on all three bands, the correlation, the level and the image are exactly where they were. That is worth stating because the plugin is doing real work at that setting. A Linkwitz-Riley crossover pair sums to an all-pass rather than to a flat wire, so the mid signal is split into the same three bands and summed again even though all three of its band gains are 1.00. If only the side signal went through the network, it would come back phase-rotated against an untouched mid and the image would shift before you had touched anything. Running both through the identical network keeps them locked: at unity the output is the input through one all-pass, the same one on both channels. The three way split needs one more all-pass to stay complementary, which is why the low band is fed through an all-pass at the upper crossover. Without it the low band arrives at the wrong phase and the sum dips around the lower crossover.