WOW / MANUAL

Wow
Manual

What it does

Wow continuously changes playback speed to create slow warble, fast flutter, and natural drift. It does not add saturation, hiss, dropouts, or an EQ curve.

Main controls

Wow Rate
Sets the slow oscillator from 0.1 to 4 Hz.
Flutter Rate
Sets the fast oscillator from 6 to 30 Hz.
Wow / Flutter
Blends the two oscillators with constant power.
Amount
Sets the total pitch movement.
Drift
Adds repeatable, smoothly changing variation to both oscillator rates.
L/R Phase Offset
Offsets left and right modulation by up to 180°.

Quality

HQ is the default. Normal, HQ, and Ultra use rate-aware windowed-sinc readers with progressively tighter spectral accuracy. Draft uses a lower-cost cubic interpolator for auditioning and comparison.

Pitch range and latency

Rate-scaled keeps delay excursion bounded, so faster modulation produces more pitch movement. At 48 kHz it reports about 8.0 ms latency.

Constant keeps the perceived pitch range more consistent across oscillator rates. At 48 kHz it reports about 34.2 ms latency.

The host is responsible for latency compensation. Changing Pitch Range can make the host restart processing.

FAQ

How do I change the interface size?
Click OIKO AUDIO in the top row of the plug-in, then choose an Interface Scale value from 50% to 200%.
Where do I install it?
Copy either or both plug-in bundles into the matching folder, restart the DAW, then rescan plug-ins.
macOS~/Library/Audio/Plug-Ins/CLAP~/Library/Audio/Plug-Ins/VST3
WindowsC:\Program Files\Common Files\CLAPC:\Program Files\Common Files\VST3
Linux~/.clap~/.vst3
Are the beta builds signed or notarized?
No. The current builds are unsigned and the macOS builds are not notarized. macOS and Windows may ask you to allow the plug-in before the host can scan it.

Why clean modulation matters

Pitch modulation changes the speed at which stored audio is read. Most requested read positions land between samples, so the plug-in has to reconstruct values that were never recorded directly. That reconstruction method has a large effect on the output.

Wow separates modulation, stereo offset, quality, and pitch-range controls.

The interpolation problem

A simple interpolator is inexpensive, but it can lose high-frequency level and create tones unrelated to the requested pitch movement. When playback speeds up, frequencies near the top of the input spectrum can be pushed beyond the output Nyquist limit. If they are not removed before resampling, they fold back into the audible band as aliasing.

Those errors are separate from the sidebands produced by pitch modulation itself. Frequency modulation naturally creates related components around the original tone; those are part of the effect. Extra interpolation spurs and folded high-frequency content are implementation artifacts.

What Wow does

Normal, HQ, and Ultra use polyphase windowed-sinc fractional-delay readers. Their low-pass cutoff follows the instantaneous playback rate: around normal speed, nearly the full input band is retained; as the read rate rises, the cutoff moves down far enough to stop content crossing the output Nyquist limit.

The filter banks are prepared away from the real-time audio thread and interpolated smoothly during processing. Normal uses 80 taps, HQ uses 96, and Ultra uses 128. The increasing tap count tightens spectral accuracy at a higher processing cost. Draft deliberately uses a simpler cubic interpolator.

Spectrogram comparing an unfiltered pitch-modulated delay, typical wow and flutter plug-ins, and Oiko Wow
Matched input, modulation depth, sample rate, level, FFT settings, and render duration. Competing products are anonymised.

Reading the comparison

The bright central curve is the modulated test tone. The additional arcs and haze around it show energy created by the resampling process. The unfiltered delay folds substantial energy through the spectrum. The middle examples reduce that energy but retain repeated spurs. Wow keeps the intended moving tone and its physically correct modulation sidebands while suppressing the unrelated structure.

The difference is easiest to hear on exposed high frequencies, extreme settings, and layered modulated signals. On darker or dense material it may be subtle, but clean resampling prevents the artifacts from accumulating.