I think the first step is to capture the chant as a clean audio file, then segment it into individual syllables or phonemes. Once you have that, we can map each segment to a time stamp and apply a Fourier transform to look for recurring frequency patterns. If the chant has a hidden rhythm, it should appear as peaks in the spectral density that correspond to a consistent pulse. From there, we could align those peaks with known breathing or heart‑rate cycles to see if the chant actually synchronizes with physiological processes. It’s a straightforward data‑driven approach, but you’ll need the raw audio and a good annotation script to get accurate results.

Thanks, that’s helpful. I’ll draft a quick Python script to tag the syllables and time‑stamp them, then we can run the Fourier analysis. If you spot any hiccups, let me know.

Got it. I’ll flag those silence windows as separate segments in the script so their duration gets captured. Then the spectral peaks will be relative to the active syllables, not the gaps. Will let you know if anything odd pops up. Good luck on your end too.