Skip to content

forgegen

forgegen

Audio and video to funscript — in seconds, not hours.

forgegen is a haptic content generation engine. Drop in a music track or video file. forgegen analyses the rhythm, phrase structure, and energy, then outputs a .funscript file ready to drive any haptic device.

Quick Start Using the UI CLI Reference

Who is this for?

Content scripters

You produce funscripts for an audience — posting to EroScripts, selling on Patreon, scripting for video libraries. A 10-minute video takes 2–8 hours to script by hand. forgegen gives you a quality starting point in under a minute, leaving your time for the refinement that makes a script yours.

Music and EDM creators

You want haptic tracks synced to music — club nights, VR experiences, haptic-enabled music videos. forgegen handles the complete audio-to-haptics pipeline. No video required. No scripting tools. Drop a song, generate, play.

The estim community

Stereo audio-driven estim is a niche workflow with no good tools. forgegen's estim audio path treats stereo stim audio as a first-class input — two channels, two curves, clean export. (Coming in v0.2)

FunScriptForge users who want a starting point

Manual scripting from scratch is slow. Starting from a generated draft is fast. forgegen feeds FunScriptForge a clean, structured .funscript that you then refine, tone-shape, and export per device. The generation handles the tedious work; you focus on what makes it feel good.

Pipeline builders and studios

You produce content at scale — multiple tracks or videos per week. Because forgegen ships a full CLI and importable Python library, it fits into automated watch-folder pipelines, CI/CD workflows, and batch processing scripts. Generate an entire back-catalog overnight.

Why does it exist?

The existing toolchain is fragmented:

Tool Problem
OpenFunscripter Manual editing only — doesn't generate from media
PythonDancer Basic beat-locking — mechanical output, no phrase shaping
FunGen AI-assisted but VR-only — doesn't work on flat video or audio
FunscriptFlow Generic pipeline — no user interface, requires coding
funscript.io Micro-tools — Handy-specific, no offline use

No tool generates quality funscripts from arbitrary audio or video, handles phrase-level shaping, runs locally without a cloud API, and hands off cleanly to an editing tool.

forgegen is that tool.

How it works

Media file → Beat & energy analysis → Phrase classification → Curve shaping → .funscript
  1. Analyse — forgegen detects the beat grid, BPM, phrase boundaries, and energy envelope. For music, this uses librosa's beat tracker with optional harmonic-percussive stem separation.
  2. Classify — each phrase is labelled automatically: break, tease, slow, steady, fast, or edging, based on the phrase's average energy, tempo, and energy trend.
  3. Shape — the motion curve is sculpted per mode. A tease phrase has narrow amplitude. An edging phrase builds from 50% to full range. A break nearly stops the device. The curve feels written, not generated.
  4. Export — a validated, sorted .funscript JSON is ready to load in FunScriptForge, SyncPlayer, or any compatible player.

Where it fits

forgegen is the generation layer. FunScriptForge is the editing layer. Together they cover the full pipeline.

forgegen  →  .funscript  →  FunScriptForge  →  any device

forgegen never edits funscripts. FunScriptForge never generates from scratch. They are independent tools with a clean handoff at the .funscript boundary.

See forgegen + FunScriptForge for the full workflow.

Getting started

  • Quick Start — generate your first funscript in under a minute
  • Using the UI — full walkthrough of the Generate and Details tabs
  • CLI & Automation — run forgegen from the command line or integrate into a pipeline
  • Styles — understand the four style presets
  • Phrase Modes — understand how forgegen shapes each section of your track