Blog

Resynthesis: turn any sound into an editable Vital preset

Drop a one-shot into Audial Synth, a free GPLv3 Vital fork, and get back a patch you can open, edit and modulate like any other preset.

October 9, 2026 5 min readAudial

Sampling a sound gives you that sound. Resynthesising it gives you the instrument that made it. The difference shows up the moment you want the sound a fifth higher, a little brighter, or with a slower attack: a sample needs stretching and filtering, while a patch just needs a knob.

Audial Synth does the second one. Drop a one-shot into it, wait, and a preset appears in the synth with every oscillator, envelope, LFO and effect laid out and editable.

What Audial Synth is

Audial Synth is a free, GPLv3 fork of the Vital wavetable synthesizer written by Matt Tytel, with one addition: a Resynth panel. Everything else is the synth you may already know. Spectral warping oscillators, a wavetable editor, a modulation matrix with live graphical feedback, and a full effects chain.

It is not affiliated with or endorsed by the upstream project. It is a separate product with a separate name, upstream copyright notices are kept intact, and the fork's source is public at github.com/AudialAI/sound2vital-gui as GPLv3 requires.

Builds ship for macOS as a VST3, an Audio Unit and a standalone app, so it loads in Ableton Live, Logic Pro, FL Studio and anything else that takes VST3 or AU. The current build is 8.8 MB and is not yet notarized, so on first launch right-click and choose Open rather than double-clicking. Install is the usual copy:

  1. Unzip the archive
  2. Copy AudialSynth.vst3 to ~/Library/Audio/Plug-Ins/VST3
  3. Copy AudialSynth.component to ~/Library/Audio/Plug-Ins/Components
  4. Rescan plugins in your DAW

The synth itself works with no account at all. Only the Resynth step calls Audial's hosted engine, and that needs an Audial User ID, an API key, and an active subscription.

Drop a clip, get a patch

The RESYNTH button sits in the header. Clicking it opens a panel that takes one audio file of up to 20 seconds and holds your credentials. Nothing is sent anywhere until you drop a file in.

When you do, the clip goes to the Audial API, which analyses it and returns a patch. The patch is written into your user preset folder and loaded into the synth, so it is a real preset on disk from the moment it arrives, not a temporary state you can lose.

Jobs typically take somewhere between one and four minutes depending on the clip. Alongside the preset, the service produces a render of the patch and a short report, so you can hear how the patch behaves before you open it.

What to feed it

Resynthesis is a timbre problem, so the input that works is a single clear sound, not a passage of music.

Good candidates: a single synth note held for a second or two, a plucked string, a bell, a bowed note, a pad chord stab, a sound design one-shot, a vocal "ahh". Anything with a stable character and a clear attack.

Poor candidates: a full mix, a drum loop, anything with reverb longer than the note itself, anything with two instruments playing at once. If you want a patch from a sound buried in a track, split the stems first, then cut a single note out of the stem and resynthesise that.

Practical notes. Keep the clip short, since the useful information is in the first moments and the sustain, not in a long tail. Trim silence off the front. Record or export it dry; the engine will faithfully model the reverb you leave in, and then you cannot turn it off. Mono is fine.

Editing the result

This is the part that makes it worth doing. The patch that comes back is an ordinary Vital preset. Nothing about it is locked, baked or black-boxed.

Open the oscillator pages and you can see what the engine chose: the wavetables it built, their levels, their tuning and phase relationships. The envelopes it fitted to the amplitude and the filter are on their usual pages with their usual handles. The modulation matrix shows every routing it set up.

So the edits you already know all work:

  • Slow the amp attack and a plucked source becomes a pad.
  • Pull the filter envelope depth down and an aggressive source becomes a usable bass.
  • Add an LFO to wavetable position for movement the original sound never had.
  • Detune or add a voice in unison to widen a mono source.
  • Replace the effects chain entirely; the engine's choices there are a starting point, not a verdict.
  • Save it as your own preset, and it lives in your library like anything else.

The useful mental model is that resynthesis gives you a sketch of a patch in the right ballpark, and your edits make it an instrument. A patch that is 90 percent of the way to the source sound is often more useful after you deliberately move it away from the source.

The same thing without the plugin

Resynthesis is a hosted function, so the plugin is one of several ways in.

In the browser, upload a clip at the resynthesis tool and download the .vital file, then drop it into the preset browser of Audial Synth or open it from the Load menu.

In Python, it is one call:

import audial

audial.config.set_api_key("your_api_key")
audial.config.set_user_id("your_user_id")

result = audial.sound2vital("clip.wav")
print(result["files"]["folder"])

Over REST, send the clip inline as multipart form data:

curl -X POST "https://api.audialmusic.ai/api/functions/run/sound2vital" \
  -H "X-API-Key: your_api_key" -H "X-User-ID: your_user_id" \
  -F "userId=your_user_id" \
  -F "original=@/path/to/clip.wav"

That returns an execution id immediately. Poll GET /api/db/{userId}/execution/{exeId} until state is completed or failed; the completed execution carries the .vital preset along with the renders and a report. Without an active subscription the call answers HTTP 402 with code SUBSCRIPTION_REQUIRED.

And from an AI assistant, the Audial MCP server exposes it as the sound2vital tool, so "turn ~/samples/bell_c3.wav into a synth preset" is a complete instruction. Batch resynthesising a folder of one-shots is a reasonable thing to ask for there.

Where to go next

If you already use Vital, this is the fastest way to turn a sample folder into a preset library you can actually modulate. If you do not, the synth is free and the Resynth panel is the reason to try it.

Download Audial Synth from downloads, try resynthesis in the browser at the resynthesis tool, and see the sound2vital function on the API reference.

resynthesissynthsplugins