Version: see the VERSION file at the repository root (single source of truth for CMake and releases).
A JUCE-based Neural Amp Modeler implementation, derived from a large rewrite of Tr3m/nam-juce.
This project builds on Steven Atkinson’s NeuralAmpModeler ecosystem: run .nam captures and cabinet IRs inside AU, VST3, and a standalone app.
- Major UI and architecture rewrite — mock-driven layout, new
NamUistack, streamlined workflow. - JUCE 8 — vendored under
Modules/JUCE/and wired through CMake. - Presets redesigned — JSON preset library under
NAM/Presets/withmanifest.json, save/rename/delete/reorder, and paths stored relative to the NAM root (no UI-only aliases in files). - Single opinionated content layout — one root folder (your
NAMdirectory) with fixed subfolders; see below. - Built-in metronome — compact always-available click control with BPM recall outside the preset library.
- Tighter, compact plugin window — optimized for a narrow portrait layout.
The app stores a NAM root path (via standalone File → Settings → Set NAM Directory, or equivalent). All captures, IRs, and presets live under that root.
Folder names are fixed: under that root you must use exactly Captures, IRs, and Presets (same spelling and capitalization). Renaming or nesting them differently will not be picked up.
NAM/
├── Captures/ # .nam models
│ ├── *.nam # optional: loose files here show under “Standalone” (first in UI)
│ └── <collection>/ # optional subfolders of .nam files
├── IRs/ # .wav impulse responses (this fork uses folder name IRs)
│ ├── *.wav # optional: loose files show under “Standalone” (first in UI)
│ └── <collection>/
└── Presets/ # user preset library
├── manifest.json # ordered list of { "id", "name" }
└── <id>.json # one preset file per id (paths relative to NAM root)
Standalone (collection row): If there are .nam or .wav files directly in Captures/ or IRs/, the browser shows a synthetic Standalone entry first, then other folders in natural sort order.
Download the latest macOS arm64 ZIP from GitHub Releases:
Download Neural Amp Modeler (macOS arm64)
Unpack the archive and open Neural Amp Modeler.app (you may need right-click → Open the first time for Gatekeeper).
The release ZIP is generated from the same Neural Amp Modeler.app that lives under binaries/macos-arm64/ in the repo. Both are refreshed by the Release arm64 build path (see Building below).
Supported and tested here: macOS arm64, Release builds via CMake. Other platforms and architectures have not been compiled or validated in this fork.
Prerequisites: Xcode command-line tools (or Xcode), CMake ≥ 3.15.
git clone <your-repo-url> nam-juce
cd nam-juce
# Preferred: build Release standalone and refresh binaries/macos-arm64/
python3 Scripts/build_macos_arm_binary.py
# Equivalent manual CMake commands
cmake -S . -B build-release-arm -DCMAKE_BUILD_TYPE=Release -DCMAKE_OSX_ARCHITECTURES=arm64
cmake --build build-release-arm --config Release --target NEURAL_AMP_MODELER_StandaloneOn Apple Silicon, this project’s CMake prefers CMAKE_OSX_ARCHITECTURES=arm64 when the host reports ARM hardware.
Artifacts (names may match your generator):
- Standalone:
build-release-arm/NEURAL_AMP_MODELER_artefacts/Release/Standalone/ - AU / VST3: under
build-release-arm/NEURAL_AMP_MODELER_artefacts/Release/
Note: For manual audio checks on Apple Silicon, prefer Release (or your usual ARM Release artefact). Debug standalone behavior may differ for real-time audio.
USE_NATIVE_ARCH=1— On x86_64 hosts this can enable extra CPU flags; on macOS arm64 the project skips the x86-only tuning. Safe to leave off for typical Apple Silicon builds.
- AU
- VST3
- Standalone
Community captures and impulse responses are widely shared on sites such as Tone3000.
There are no maintained build or install instructions for Windows or Linux in this README, and no Chocolatey or other package-manager claims. If you build on another OS or CPU, expect to adjust toolchain paths and JUCE dependencies yourself; issues on untested platforms are not guaranteed to be reproduced here.
See LICENSE.txt.