A one-knob vocal doubler. Turn up Taste and a single take blooms into a stack of up to four detuned, formant-shifted copies that sound like different mouths — width and thickness without the chipmunk.
A single control, 0–100%. It scales four coupled quantities at once: wet level, per-voice detune depth, stereo spread, and active voice count. The value is mapped per sample through a 30 ms linear ramp, so automation and live moves introduce no zipper noise.
Up to four doubled voices are summed onto the unaltered dry signal. Voices 1–2 fade in across the first 12% of travel; voices 3–4 across 50–92%. The wet bus is normalized by the square root of the active-voice gain, holding output level roughly constant as voices are added.
Each voice passes through an STFT cepstral formant shifter (512-point FFT, Hann window, 75% overlap). It separates the spectral envelope from the fine structure and warps only the envelope, displacing apparent vocal-tract size independently of pitch. Per-voice ratios sit within ±8% of unity.
Voice delay times are modulated by smoothed sample-and-hold noise — an xorshift source filtered by a 6 Hz one-pole low-pass — rather than a periodic LFO. Hold intervals differ per voice (70–110 ms), so the delays drift with no repeating pattern.
Each voice is pitch-shifted a few cents (−12 to +10 ¢) by a grain pitch shifter. Below roughly 15 cents the formant displacement from pitch shifting is negligible, so the stack widens without audible spectral artifacts. The dry signal stays centered at unity gain.
Every copy runs through antiderivative-antialiased saturation, a 5 kHz high shelf, and a one-pole low-pass (11–15 kHz), with different settings per voice. The variation decorrelates the stack and suppresses comb filtering caused by the near-coincident delays.
1. Run the .exe installer
2. SmartScreen may warn — click "More info" then "Run anyway"
3. VST3 installs to C:\Program Files\Common Files\VST3\HojPoj\
4. Rescan plugins in your DAW
1. Open the .dmg and read install.txt
2. Drag the .vst3 bundle to /Library/Audio/Plug-Ins/VST3/
3. Drag the .component to /Library/Audio/Plug-Ins/Components/
4. Follow the instructions in install.txt
5. Rescan plugins in your DAW
1. Extract the archive
2. Copy the .vst3 bundle to ~/.vst3/
3. Copy the .so to ~/.vst/ for VST2
4. Rescan plugins in your DAW