TachyWooting

Python bindings and acquisition utilities for Wooting analog keyboards.

For deeper implementation details, see documentation.md. Read the Docs/Sphinx sources live in docs/ and use NumPy-style docstrings. Console scripts are documented in docs/scripts.md.

Project Documentation

• README.md: project overview and quick start

• documentation.md: technical details and architecture notes

• development.md: maintainer workflow and SDK update process

• PLUGIN_MANAGEMENT.md: plugin installation and troubleshooting

• raw_sdk.md: direct lib/ffi SDK reference for advanced use

• Analog Key Acquisition: Read key positions (0.0–1.0) with microsecond-level timing

• Threshold-Based Triggering: Automatically capture key press trajectories around actuation threshold

• HDF5 Logging: Hierarchical per-trial logging with automatic shard merging

• Multi-Key Support: Efficiently read multiple keys simultaneously using full-buffer API

• Cross-Platform: Linux, macOS, and Windows support

• Automatic Setup: Self-contained installation with system configuration

• CLI Tools: Command-line utilities for plugin management and testing

• Read analog key pressure as floats in the 0.0 to 1.0 range.

• Convert analog pressure to integer values in the 0 to 255 range.

• Acquire one or more keys around a threshold crossing.

• Log trials to hierarchical HDF5 files.

• Build against the bundled Wooting Analog SDK headers and native libraries.

• Inspect HDF5 logs with a small visualization CLI.

Requirements

• Python 3.9 or newer (through 3.14).
• A supported Wooting analog keyboard.
• A local compiler toolchain for the CFFI interface build (see below).
• Platform-specific permissions for USB/native library access.

Compiler Setup

The CFFI extension is compiled on your machine, not shipped as a prebuilt binary, so you need a C compiler available before the interface can be built.

• macOS: install the Xcode Command Line Tools: xcode-select --install.

• Linux: install gcc via your package manager, e.g. sudo apt install build-essential (Debian/Ubuntu) or sudo dnf groupinstall "Development Tools" (Fedora).

• Windows: install the "Desktop development with C++" workload from the Visual Studio Build Tools (no full Visual Studio install needed), or from an elevated PowerShell prompt (Run as administrator), run:

  winget install -e --id Microsoft.VisualStudio.BuildTools --override "--passive --wait --add Microsoft.VisualStudio.Workload.VCTools --add Microsoft.VisualStudio.Component.VC.Tools.x86.x64 --includeRecommended"
  

  (reference for the flags above; exit code 3010 means success, reboot required — not an error.)

  pip install usually finds the compiler on its own. If it can't find cl.exe, open an "x64 Native Tools Command Prompt for VS" (or run vcvars64.bat) and retry.

Quick Start

pip install .

What setup is needed

pip install does not run system setup — pip installs from wheels, which have no reliable post-install hook. Setup is split into two parts:

1. CFFI compilation happens automatically the first time you create a WOOTING_ACQUISITION (or run wooting-demo). It needs only a C compiler — no admin rights.

2. SDK plugins + input permissions require a one-time privileged step:

   wooting-build-interface   # installs SDK plugins + permissions (needs sudo/admin)
   

If the keyboard is not detected, the error message tells you exactly to run this command — you do not have to remember it. To undo it later: wooting-delete-interface.

Development Installation

python -m pip install -e ".[dev]"
wooting-build-interface

Quick Start

from tachywooting import WOOTING_ACQUISITION

acq = WOOTING_ACQUISITION(threshold=0.8)
acq.initialize_keyboard(verbose=True)

try:
    acq.setup_logging(name="tracking", path="logs", int_analog=2)
    trial = acq.acquire_analog_values(target_keys=["A"])
finally:
    acq.uninitialize_keyboard()

CLI Demo

wooting-demo --key A --threshold 50

Visual feedback (TachyPy)

On-screen pressure feedback — the interactive fixation cross and wait_light_press_visual() — lives in TachyPy, not in this hardware package. Install it with pip install 'tachypy[wooting]', then:

from tachypy import WOOTING_ACQUISITION  # keyboard + visual feedback

HDF5 Logging

setup_logging() writes one temporary shard per trial and merges shards when uninitialize_keyboard() is called.

Final files use this layout:

/trials/0001/keys/0004/values

Each values dataset stores columns in this order:

position, time_from_onset, time_abs

Visualize Logs

python -m tachywooting.visualize logs/tracking.hdf5 --list
python -m tachywooting.visualize logs/tracking.hdf5 --trial 1 --key 4

Public API

• WOOTING_ACQUISITION: acquisition, threshold detection, readiness checks, and logging.
• convert_char_to_keycode: convert between key labels and HID keycodes.
• convert_keycode_to_char: convert between HID keycodes and key labels.
• load_trial: load a single trial from an HDF5 log file.
• load_session: load all trials from an HDF5 log file.
• trial_to_dataframe: convert a trial dict to a pandas DataFrame.
• build_interface: rebuild the CFFI interface.
• delete_interface: remove generated CFFI artifacts.
• lib and ffi: raw CFFI handles for advanced SDK access.

Troubleshooting

If importing works but acquisition fails with a missing native interface error, run:

wooting-build-interface

If no devices are detected, confirm the keyboard is connected, Wootility recognizes it, and platform permissions have been applied.

Hardware Requirements

This package was developed and tested with the Wooting UwU keypad (wooting.io/uwu), and its use is strongly recommended for optimal results.

The UwU is a 3-key Hall effect keypad using Lekker L45 V2 linear switches — contactless magnetic sensors with a smooth linear force curve (30–45 cN, no tactile bump). Keys can be configured to actuate at any depth from 0.1mm to 4.0mm.