LAP TSL-A34
A two-channel JFET limiter / compressor built as a component-level circuit model: the whole audio board - input followers, gain stage, the JFET-controlled stage, the precision rectifier, the attack / release network, the RATIO divider and the balanced output driver - is solved as a circuit in real time.
How it works - and how it differs from a plugin compressor. This is a feedback JFET limiter, and the controls mean what they meant on the unit. There is no threshold control: the threshold is fixed by the board’s trimmers (about −6.5 dBFS at 10:1, falling to about −15.5 dBFS at 3:1). GAIN is the drive into the limiter (0 … +20 dB in front of the gain cell), not make-up gain; OUTPUT adds 0 … +6 dB after the board. RATIO moves threshold and slope together - towards 3:1 the unit starts earlier and more softly, towards 10:1 later and harder - and the knee is hard. ATTACK and RELEASE read the unit’s scale; the actual gain change is slower. The first transient after silence passes at full GAIN until the envelope has charged, and at low frequencies the envelope ripple modulates the gain within the cycle. LIMIT lights from the threshold, GAIN lights when the input stage is overdriven. BYPASS is a relay bypass: the board keeps running.
The whole board, solved live
Not a compressor algorithm: the circuit is solved node by node at sample rate, as seven small stages in signal order - op-amps with real bandwidth, slew rate and offsets, and every resistor, capacitor, pot, diode and FET with its own tolerance; each channel is trimmed like a board on the bench (the two channels differ slightly, as two real boards do). Turning a knob moves the potentiometer inside the simulated board; the knee, the slope and the timing fall out of the components.
JFET gain cell
One junction FET as a shunt in the feedback of the controlled stage takes up to 17 dB out of the loop. A hard knee, a steep slope above the threshold and the gentle second-harmonic colour of a FET doing real work - that is the sound of the circuit.
Feedback detector
A precision full-wave rectifier listens to the output and drives a diode-steered attack / release network: ATTACK 0.1–10 ms, RELEASE 0.05–2 s, as printed on the scales. RATIO 3:1 … 10:1 sets the control-voltage offset, so it moves the threshold (about −16 … −6 dBFS) as much as the slope - and it can be turned while playing without a jump in level.
GAIN · OUTPUT
GAIN 0 … +20 dB sits before the controlled stage and drives the programme into the threshold; OUTPUT 0 … +6 dB is the make-up gain after the board. Calibrated at 0 dBFS = 4.4 V peak.
Two channels, LINK
Two independent boards with their own controls. The ON / OFF switch is LINK: channel 2 follows channel 1 for stereo; switch it off for dual mono.
Output transformer · SATURATION
After the board sits a model of a real audio output transformer - the core with its magnetisation and hysteresis solved sample by sample - for weight in the low end and low-order harmonics that grow with the bass and the level. SATURATION drives it harder or softer per channel, from -12 to +18 dB, without changing the level. Off by default - switch TRAFO ON in the settings menu.
LIMIT and GAIN LEDs
Red LIMIT follows the gain reduction - on at 1 dB, full brightness at 8 dB and more; green GAIN is the overload indicator of the input amplifier - when it flickers, turn GAIN down.
Settings menu
Right-click the panel: window SCALE 80 / 100 %, OVERSAMPLING 1x / 2x / 4x, TRAFO OFF / ON and SATURATION (one row per channel when LINK is off). Everything can be changed while playing - oversampling is rebuilt in the background, without a drop-out.
Presets
A preset bar above the panel: categories, Save / Save as / Delete and ◀ ▶ to step through the settings. A preset holds the knobs and the menu settings (OVERSAMPLING, TRAFO, SATURATION); the selected preset is saved with the project. One plain file per preset - copy them between machines and systems. Loading a preset never interrupts the audio.
Panel · oversampling
A rack panel rendered in 3D with ray-traced light at three times the default resolution - sharp at every size and on high-DPI screens. Hover to read a value, double-click to reset. Oversampling 1× / 2× / 4× (default 2×), counted from 44.1 / 48 kHz - the same sound and load in a 96 or 192 kHz session; at 1× the integration warping is compensated so the response follows the circuit to 20 kHz.
New in v5.3: a lighter detector - the rectifier, envelope and RATIO / control-voltage
stages are now a compact model built from the same components and solved in closed form. Same sound
(the difference to v5.2 stays more than 50 dB below the signal on music), less CPU - on a drum bus
about 25-35 % less on average and the load spike on every hit is gone. Presets now include the
right-click menu settings (OVERSAMPLING, TRAFO, SATURATION), the selected preset is saved with the
project, and a changed preset is marked with *. Default oversampling 2x.
Since v4.0.7: a new panel rendered in 3D with ray-traced light, sharp at every size; a model of a
real audio output transformer (off by default) with SATURATION -12 … +18 dB; the right-click settings
menu; oversampling switches while playing without a drop-out. The external sidechain input and the
Noise option were removed. Projects and presets made with v4.0.x load (v4.0.x presets with SATURATION
0 dB). Upgrading from v0.3.x: delete the old lap-TSL-A34.clap / lap-TSL-A34.vst3 first - the
plugin files are named LAP TSL-A34.