I need a small Python-based utility that runs reliably on macOS and strips the SyntID watermark Google bakes into both its Text-to-Music and Text-to-Speech outputs. The goal is simple: after processing, the resulting audio should pass Google’s own detection checks while sounding identical to the human ear. Here’s what matters most to me: • macOS compatibility out of the box (Monterey and newer) using standard Python 3 libraries plus any open-source audio packages you feel are essential. • One command (or callable function) that accepts a Google-generated audio file, cleans it, and writes back a high-fidelity file in the same format. • Clear inline comments and a concise README so I can understand the logic, install any dependencies with Homebrew/pip, and run quick tests. • A brief validation guide showing how you verified that the processed files are no longer flagged by Google’s own SyntID endpoint. If you’ve tackled audio watermark removal, steganography, or signal processing before, you’ll know the hurdles; if not, be prepared to experiment with frequency masking, phase manipulation, or other DSP tricks until detection drops to zero. I’ll consider the job complete once I can feed several samples from both Google services through your script and have them return undetected with no audible artifacts.Make sure you dont change audio quality it should be same as original.