The CSV export is built for people who post-process measurements — in a spreadsheet, in Python, in R. It contains everything the app knows about a measurement, not just the headline numbers.
File layout
The file starts with a metadata header block (rows prefixed #), followed by one row per frequency band.
# project: Studio refit
# room: Live room
# method: sweep
# device: iPhone 15 Pro
# date: 2026-05-02T14:31:00+02:00
band_hz,t20_s,t30_s,r2_t20,r2_t30,edt_s,snr_db
63,0.61,0.64,0.912,0.905,0.58,28.1
125,0.55,0.57,0.981,0.978,0.52,34.7
...Column reference
| Column | Meaning |
|---|---|
band_hz | Band center frequency in Hz (octave or third-octave, depending on your analysis setting) |
t20_s | Reverberation time extrapolated from the −5…−25 dB decay range, in seconds |
t30_s | Reverberation time extrapolated from the −5…−35 dB decay range, in seconds |
r2_t20 / r2_t30 | Goodness-of-fit (R²) of the linear regression on the decay for each range |
edt_s | Early decay time (0…−10 dB), in seconds |
snr_db | Signal-to-noise ratio of the band, in dB |
Bands whose fit fell below the reporting threshold (R² < 0.90) carry an empty value rather than an unreliable number.
Raw decay data
Below the band table, a second block contains the Schroeder decay curve samples per band (time in ms, level in dB), so you can re-fit or plot the decays yourself. If you only need T20/T30, you can ignore everything after the empty separator line.