🎨🔊 PixelWave

PixelWave is a unique image↔audio transformation toolkit that allows you to sonify grayscale and color images into audio signals — and reconstruct them back! It's built for artists, researchers, and curious developers to explore how visual data can be encoded into sound.

---

🌟 Features

• 🔄 Image to Audio (Sonification) — Convert grayscale or RGB images to .wav sound
• 🔁 Audio to Image (Desonification) — Reconstruct images from encoded .wav files
• 🎨 Supports grayscale and full-color (RGB)
• 🔊 Uses multichannel audio encoding for color images
• 🧪 Useful for data experiments, art installations, and audio-visual hacking

---

📂 Project Structure

PixelWave/ │ ├── i2a.py # Grayscale image ➝ mono audio ├── a2i.py # Mono audio ➝ grayscale image ├── colouredi2a.py # RGB image ➝ 3-channel audio ├── coloureda2i.py # 3-channel audio ➝ RGB image ├── pengbrew.png # Sample grayscale image ├── a2.png # Sample RGB image ├── qr_sonified.wav # Output from i2a ├── color_encoded.wav # Output from colouredi2a ├── reconstructed.png # Output from coloureda2i └── README.md # This file

🛠️ Technologies Used

• Python 3
• NumPy — for image/audio array manipulation
• Pillow (PIL) — for image loading, resizing, RGB splitting
• SoundFile — for reading/writing multi-channel .wav files
• SoundDevice — for optional audio playback
• OpenCV — for displaying and saving grayscale images

---

🚀 How to Use

1. 🔄 Convert Grayscale Image to Audio

python i2a.py
Input: pengbrew.png

Output: qr_sonified.wav

2. 🔁 Convert Audio Back to Grayscale Image

python a2i.py
Input: dog-bark.wav

Output: bark_image.png

3. 🎨 Convert RGB Image to 3-Channel Audio

python colouredi2a.py
Input: a2.png

Output: color_encoded.wav

4. 🖼️ Reconstruct Color Image from Audio

python coloureda2i.py
Input: color_encoded.wav

Output: reconstructed.png

📦 Installation
Install the required libraries:

pip install numpy pillow soundfile sounddevice opencv-python
📸 Example
Input Image	Encoded Audio	Reconstructed Image
	🎵 color_encoded.wav	

🧠 Applications
🎶 Audio-driven art

🧪 Data representation experiments

🎓 Teaching signal processing or encoding

🕵️ Hidden image transmission via sound

⚖️ License
This project is open-source and free to use under the MIT License.