The ultra-fast, zero-latency gaming audio router, 10-band parametric equalizer, and ChatMix controller for Linux & PipeWire.
The genuine SteelSeries Sonar alternative engineered in Rust for Arch Linux, Hyprland, Sway, KDE Plasma, and GNOME.
Language: English | العربية (Arabic)
Run this single command in your terminal to build and install VoiceGG rootlessly to ~/.local/bin:
curl -fsSL https://raw.githubusercontent.com/1MEshh/voiceGG/main/install.sh | bashTip
Launch anytime from your application runner (Rofi, Wofi, tofi, dmenu) or run voicegg-launcher in terminal.
When I wiped Windows and made the full jump to Arch Linux with Hyprland for my daily gaming and workstation setup, there was one piece of proprietary Windows software I genuinely missed: SteelSeries Sonar (SteelSeries GG).
For years on Windows, my entire competitive gaming flow revolved around Sonar:
- Having separate virtual channels for Game, Chat, and Media so I could balance game audio against Discord on the fly.
- The physical ChatMix dial to immediately silence loud teammates during clutch rounds in Counter-Strike and Apex Legends.
- A tuned 10-band parametric equalizer with footstep isolation profiles that made subtle enemy movements unmistakable.
- Live microphone monitoring (sidetone) so I didn't yell into my headset while talking with friends.
When I searched for a Linux alternative to SteelSeries Sonar, I hit a dead end:
- ❌ Traditional patchbays (qpwgraph, Helvum): Inundated with confusing spiderwebs of hundreds of raw PipeWire audio nodes, with zero gaming presets, no per-app persistent routing, and no ChatMix logic.
- ❌ Windows SteelSeries GG in Wine/Proton: Impossible to run. It demands proprietary Windows kernel audio drivers, idles at 800 MB to 1.2 GB of RAM, and runs constant background telemetry services.
- ❌ Generic sound managers: Lacked application memory, lacked dynamic filter chains, and had zero Wayland / Hyprland bar integration.
I refused to settle. So I built VoiceGG.
Engineered from the ground up in Rust directly interfacing with the native PipeWire filter-chain engine, VoiceGG delivers everything SteelSeries Sonar provided—and more:
- Consumes only ~6 MB of RAM (over 100× lighter than Windows SteelSeries GG).
- Uses 0.0% CPU when idle with sub-6ms roundtrip processing latency.
- Features zero cloud telemetry, zero accounts, and 100% offline local privacy.
- Includes authentic Thmanyah (
ثمانية) typography and full bidirectional Arabic (RTL) localization.
— Built with passion for the Linux gaming community by @1MEshh
| Document | Description |
|---|---|
| ⚡ Quick Install | One-liner curl install script and dependencies |
| 📖 The Story Behind VoiceGG | Personal journey from Windows Sonar to Arch Linux & Rust |
| 🏗️ PipeWire DSP Architecture | Deep technical specification of virtual node graph and zero-copy filters |
| 🎯 Competitive Presets & Audio Tuning | Calibrated 10-band EQ curves for CS2, Apex, Valorant, Tarkov, and R6 |
| 🖥️ Desktop Integration (Waybar & Hyprland) | Keybindings, Waybar status modules, SwayOSD, and Rofi selectors |
| 📊 Benchmarks & Performance Audit | 6.1 MB RSS vs 940 MB SteelSeries GG, <5.3ms latency, and 0.0% CPU |
| 🌍 Localization Architecture | Authentic Arabic pro-audio terminology and Thmanyah font setup |
| 🔧 Diagnostic & Troubleshooting Playbook | Step-by-step fix for RNNoise, Xruns, latency tuning, and panic reset |
| 💻 CLI Reference Manual | Complete syntax for scripting, keybinds, ChatMix, and stream routing |
| 🇸🇦 الدليل باللغة العربية | التوثيق الكامل باللغة العربية مع دعم اتجاه اليمين-إلى-اليسار (RTL) |
Dedicated virtual channels for Master, Game, Chat, Media, Aux, and Mic. Drag and drop running applications (Spotify, Chrome, Discord, Steam) directly between channels with instant effect. Includes a real-time ChatMix balance slider.
Interactive, responsive parametric EQ graph with draggable frequency nodes. Includes built-in competitive profiles for Apex Legends, Counter-Strike 2, Valorant, and Overwatch 2, quick Bass/Vocal Clarity sliders, and Spatial Audio (Virtual 7.1 Surround) virtualization.
Studio-grade microphone processing suite running directly inside the PipeWire audio graph:
- Live Voice Monitoring: Zero-latency hardware loopback with feedback safety guard and dedicated volume slider.
- AI Noise Remover: Deep neural background noise cancellation via native RNNoise (
noise-suppression-for-voice) with honest plugin availability detection. - Noise Gate: Calibrated gating with live
● GATE OPEN/○ GATE CLOSEDindicator driven by real peak metering. - Dynamic Compressor: Broadcast dynamic range control with live real-time gain reduction monitoring.
- Arabic & RTL Localization: Full native Arabic support with authentic pro-audio terminology and custom Thmanyah (ثمانية) typography.
| Feature | VoiceGG (Linux) | SteelSeries GG / Sonar (Windows) | Generic Linux Patchbays |
|---|---|---|---|
| Operating System | Linux (Arch, Hyprland, Sway, KDE, GNOME) | Windows 10/11 only | Linux |
| Idle Memory Footprint | ~6.1 MB RSS | ~650 MB – 1.2 GB | ~80 MB – 250 MB |
| Idle CPU Utilization | 0.0% | 1.5% – 5.0% | 0.5% – 2.0% |
| Telemetry & Tracking | Zero (0%) — 100% Offline & Private | Constant cloud telemetry | None |
| Dedicated Channels | 5 (Master, Game, Chat, Media, Aux, Mic) | 5 channels | Manual wire patching |
| Drag & Drop App Routing | Yes (Automatic & Persistent) | Yes | No (Manual wire links) |
| Streamer Mode (Dual-Bus) | Yes (Personal Headphone vs Stream Mix) | Yes (Sonar Streamer) | No |
| ChatMix Balance | Yes (Hotkeys, GUI & Waybar module) | Yes (SteelSeries hardware or GUI) | No |
| 10-Band Parametric EQ | Yes (Interactive Canvas + RBJ Biquad Math) | Yes | Requires third-party plugins |
| AI Voice Isolation | Yes (NVIDIA Broadcast-grade VAD envelope) | Yes (ClearCast AI) | Basic RNNoise only |
| Live Voice Monitoring | Yes (Feedback-safe loopback + volume control) | Yes | Manual loopback creation |
| Safe Lifecycle Teardown | Yes (100% cleanup on SIGTERM / btop kill) | System driver hook | Leaves dangling ghost nodes |
| Arabic & RTL Localization | Yes (Full RTL layout + Thmanyah typography) | Partial / LTR only | No |
| Emergency Audio Reset (Panic) | Yes (1-click restore to hardware defaults) | Reinstall driver | Restart pipewire |
VoiceGG runs unobtrusively in your desktop environment. Closing the main window minimizes to the system tray, keeping your audio routed seamlessly in the background.
VoiceGG includes an official status query script and ready-to-use custom module for Waybar users:
- Install helper script:
cp packaging/waybar/waybar-voicegg.sh ~/.local/bin/
chmod +x ~/.local/bin/waybar-voicegg.sh- Add custom module to
~/.config/waybar/config.jsonc:
- Style in
~/.config/waybar/style.css:
#custom-voicegg {
padding: 0 10px;
margin: 3px 4px;
background: #11151F;
border: 1px solid #262E40;
border-radius: 6px;
color: #10B981;
font-weight: 700;
}
#custom-voicegg.offline {
color: #64748B;
border-color: #1C2230;
}Bind VoiceGG commands directly to your keyboard or macro keys in ~/.config/hypr/hyprland.conf:
# Toggle Microphone Mute
bind = SUPER, F9, exec, voicegg ctl mute mic toggle
# Adjust ChatMix balance (More Game / More Chat Comms)
bind = SUPER, F10, exec, voicegg ctl chatmix -10
bind = SUPER, F11, exec, voicegg ctl chatmix +10
# Master Volume Controls
bind = SUPER, F12, exec, voicegg ctl volume master +5
bind = SUPER SHIFT, F12, exec, voicegg ctl volume master -5
# Emergency audio reset (Instantly restore hardware routing)
bind = SUPER CTRL, ESCAPE, exec, voicegg panicVoiceGG creates virtual sink and source nodes directly in the PipeWire graph. Applications connect into their respective virtual bus, processed through low-overhead DSP filters, and summed cleanly into your physical playback device with safety limiter protection.
No. VoiceGG works with any headset, headphone, USB DAC, or audio interface (HyperX, Logitech, Sennheiser, Beyerdynamic, Corsair, Razer, Apple EarPods, etc.). ChatMix operates directly inside the PipeWire audio engine.
VoiceGG is officially developed and packaged for Arch Linux and Arch-based distributions (EndeavourOS, Manjaro, Garuda, CachyOS), and runs on any Linux distribution with a modern PipeWire audio server (Fedora, Ubuntu 24.04+, Debian 13+, openSUSE).
VoiceGG includes a rootless, safe uninstallation script:
voicegg-uninstallOr via curl:
curl -fsSL https://raw.githubusercontent.com/1MEshh/voiceGG/main/uninstall.sh | bashVoiceGG is free, open source software licensed under the GNU General Public License v3.0 (GPLv3). Pull requests, bug reports, and game preset contributions are always welcome!
- GitHub Repository: https://github.com/1MEshh/voiceGG
- Author: @1MEshh


