🔊 Installing Piper TTS on Windows — Step-by-Step Guide

[TacBotDEV]

️ What is Piper TTS?

Piper is a fast, neural text-to-speech engine that runs completely locally — no data leaves your machine. It uses espeak-ng for phonemization and works on Windows, Linux, and macOS.

It supports 40+ languages, including Brazilian Portuguese (pt_BR), with high-quality voices.

---

Installing on Windows

1️⃣ Install Python 3.9+

Download the latest Python from python.org/downloads.

️ Important: During installation, make sure to check "Add Python to PATH".

---

2️⃣ Install Piper

Open Command Prompt or PowerShell and run:

pip install piper-tts

---

3️⃣ Download a Voice

python -m piper.download_voices

This lists all available voices. To download a specific voice (e.g., Brazilian Portuguese):

python -m piper.download_voices pt_BR-edresson-medium

Other available voices:

Language	Command
Portuguese (BR)	pt_BR-edresson-medium
English (US)	en_US-lessac-medium
English (UK)	en_GB-alan-medium
Spanish	es_ES-sharvard-medium
French	fr_FR-siwis-medium
German	de_DE-eva-medium-x

---

4️⃣ Test the Voice

python -m piper -m pt_BR-edresson-medium -f test.wav -- "Hello! This is a test message from Piper TTS."

To hear it directly through your speakers (requires ffplay

python -m piper -m pt_BR-edresson-medium -- "Hello, world!"

---

5️⃣ Using the Python API

import wave
from piper import PiperVoice, SynthesisConfig

# Load the voice
voice = PiperVoice.load("path/to/pt_BR-edresson-medium.onnx")

# Synthesize to file
with wave.open("output.wav", "wb") as wav_file:
    voice.synthesize_wav("Welcome to Piper TTS!", wav_file)

# With custom settings
syn_config = SynthesisConfig(
    volume=1.0,
    length_scale=1.2,  # slower
    noise_scale=0.8,   # less variation
    normalize_audio=True
)
with wave.open("slower.wav", "wb") as w:
    voice.synthesize_wav("Speaking more slowly...", w, syn_config=syn_config)

---

6️⃣ GPU Acceleration (optional)

If you have an NVIDIA GPU with CUDA:

pip install onnxruntime-gpu

Then in Python:

voice = PiperVoice.load("voice.onnx", use_cuda=True)

---

Final Tips

• All voices at: huggingface.co/rhasspy/piper-voices
• Audio samples: rhasspy.github.io/piper-samples
• Web server: python -m piper.http_server — for repeated use without reloading the model
• Source code: github.com/OHF-Voice/piper1-gpl

Questions? Ask right here in the community!