What Is This?
This site allows users to transform images into the frequency domain using an algorithm known as Fast Fourier Transform (FFT) - This allows for reversible encoding with potential applications in signal analysis, data visualization, or data obfuscation techniques such as Image Steganography
Depending on the encoding mode used, this tool can preserve or irreversibly alter pixels and bit-level data. Some modes optimize compactness, while others maximize fidelity - at the expense of file size. Below I outline each encoding mode, providing example decoded images for each.
Be mindful - the FFT algorithm works best with image dimensions equal to a power of two; 256, 512, 1024, etc. Your input image will be padded during encoding and then cropped during decoding. Original dimensions are stored using metadata. The test image used below was 512x512, 48KB. [view] RGB 16-bit output files totaled 1.44MB. The same image at 513x513 would be padded up to 1024x1024, increasing the output files to 5.33MB!
Which Mode Should I Use?
| Mode | Description | Outputs | Encoded Size | Decoded Size |
|---|---|---|---|---|
| B&W 24-bit | Convert RGB > Grayscale and store high-precision frequency data. Minor bit loss. | 2 | 1.39MB | 56KB [view] |
| RGB 16-bit | Preserves visual RGB image with minor bit loss; imperceptible to the eye. | 3 | 1.44MB | 288KB [view] |
| RGB 24-bit | Fully lossless RGB encoding with maximum color fidelity. | 6 | 4.20MB | 68KB [view] |
| RGB BIN | 64-bit FFT output stored as binary data. Lossless RGB. | 1 | 12MB | 68KB [view] |
Disclaimer: This tool is intended for educational and creative use only. While this tool has potential for use in concealment techniques, such as steganography, the author does not condone or support the use of this tool for illegal, malicious, or unauthorized activities. Users are solely responsible for how they use the outputs generated by this tool.