MonkSynth Revives Delay Lama's Cult Vocal Formant Sound for Modern DAWs

JonET's free MonkSynth arrived in beta this week carrying something producers have chased in freeware bins for two decades: the exact X/Y formant engine that made Delay Lama, the 2002 AudioNerds chanting-monk plugin, one of the most cult-beloved and most-mocked synths ever released.

Delay Lama earned that reputation by combining FOF synthesis (Fonctions d'Onde Formantique, a grain-based formant method) with an animated monk designed by a fourth collaborator, Frank Post, and behavior so lovably bizarre the plugin appeared on best-and-worst synth lists, featured in Bad Gear episodes, and quietly found its way onto more than a few successful records. MonkSynth, now at v0.2.0-beta.2, preserves that same FOF engine while adding an ADSR envelope, stereo delay with feedback, up to 10-voice unison, and five factory presets.

The talkbox-style melodic hook is the most immediately grab-and-go application. Set MonkSynth to monophonic unison, dial in portamento, and drag the X/Y pad slowly across the vowel axis while playing a simple three-note line. The formant shapes are idiosyncratic enough to cut through a dense mix unaided, and pairing them with reverb or a reverse delay amplifies the chant-melody effect without a vocoder or a sample library anywhere in the signal chain.

For choir pads and slow-burn atmosphere, sparse filter modulation against the built-in stereo delay turns a single sustained chord into something that sits between a Tibetan overtone ensemble and a malfunctioning speech synthesizer. Set a long ADSR release, reduce the delay feedback incrementally, and let the vowel shapes bleed into each other across the trail. The movement does the compositional work.

Pushed through a ring modulator or bitcrusher, MonkSynth sheds its comedy-freeware origins entirely and produces robotic leads that carry none of the original's cheerful absurdity. The built-in delay contributes enough harmonic smear on longer feedback settings to generate genuine grit, making the plugin more versatile than its cartoon-monk reputation suggests.

The fastest proof-of-concept takes under 30 seconds: load the plugin, hold any note in the middle register, and drag the X/Y pad corner to corner. Within that single gesture, the sound moves from something approximating a human vowel, into a cartoon gargle, and out the other side as a dissonant harmonic drone. Zero preset editing required.

MonkSynth runs as AU and VST3 on Windows, macOS, and Linux. The plugin ships without a built-in skin; on first launch, a setup screen prompts an import from the original Delay Lama DLL, freely available from AudioNerds. Custom themes can be loaded by right-clicking the GUI. At v0.2.0-beta.2, expect occasional visual or installer quirks, and skin imports may fail depending on host or OS security settings. JonET is actively soliciting community feedback and custom preset sharing via the project's GitHub repository.

With Mythoz having released a separate Delay Lama revival, Krazy Sandi, back in January, 2026 is shaping up as the year the internet finally gave Frank Post's chanting monk the preservation treatment it probably always deserved.