I’m part of a research group at Cytion, and we primarily work with cell lines like MDA-MB-231 in biomedical studies. We’re exploring unconventional approaches for analyzing data — one idea we’re toying with is converting cellular activity patterns into audio signals to better visualize (or rather, hear) anomalies or behavioral differences.
Does anyone here have experience using Audacity (or scripting with its API) to analyze or process non-traditional audio sources for scientific or pattern-recognition purposes? We’re open to creative uses of the platform.
Would love to hear if anyone has tried something similar — or if it’s even feasible.
Audacity is open source but I don’t know about the “API”. It’s a “big project” so probably not something you want to dig-into just for the possibility of discovering something.
A lot of people doing scientific work use MATLAB (or a MATLAB clone).
You can import any file as “raw data”. You can experiment with different sample rates since there is (presumably) no actual sample rate.
Besides looking at the waveform, and possibly playing it as audio, you look at the frequency spectrum with Analyze → Plot Spectrum.
Of course the waveform and spectrum will change with the assigned sample rate.
…If you don’t know this, audio data is simply a sequence of sample values, each representing the “wave” amplitude at one instant in time. So it’s not too hard to analyze except that it’s ship-load of data and you can’t tell much by simply looking at the numbers. (Digital Audio Fundamentals)