File Conversion

Seismic waveform file conversion tooling for GaiaCode OmegaBox digitizer data: converting between binary miniSEED, ASCII .dat (for SeismoSignal ingestion), and SAC formats, with StationXML metadata injection into SAC headers.

Contents

Path	Description
seed2dat.py	Convert miniSEED → ASCII .dat (Time, Amplitude columns) for SeismoSignal, with optional sensitivity scaling from raw counts to physical units
dat2seed.py	Convert ASCII .dat back to binary miniSEED, reversing the sensitivity scaling to restore raw integer counts
SEED2SAC/GaiaCode Pipeline/	Headless, cron-safe pipeline: mirrors miniSEED data directly from OmegaBox units over HTTP (wget), converts to SAC with ObsPy, and injects station coordinates / component orientation from StationXML into SAC headers

Technology Stack

• Python 3 with ObsPy (miniSEED/SAC I/O, StationXML parsing) and NumPy
• PyYAML — pipeline configuration
• wget — mirroring OmegaBox /data/ trees (pipeline)

Hardware Dependencies

• GaiaCode OmegaBox digitizer units (24-bit ADC) — the headless pipeline mirrors continuous data directly from the units' built-in HTTP server (http://<unit-ip>/data/). The OmegaBox does not serve StationXML, so one metadata file per station is kept alongside the pipeline.
• seed2dat.py / dat2seed.py operate on local files and have no hardware dependency.

Setup & Installation

python3 -m venv venv
./venv/bin/pip install obspy numpy pyyaml
sudo apt install wget        # pipeline only (Debian/Ubuntu)

Usage

miniSEED → ASCII .dat (seed2dat.py)

from seed2dat import mseed_to_dat

# sensitivity=1.0 keeps raw counts; use the channel's StationXML
# sensitivity (counts per m/s²) to export physical units
mseed_to_dat("sample.mseed", "converted_waveform.dat", sensitivity=1.0)

Output: tab-separated Time(s) / Amplitude columns with a header row, relative time starting at 0.0 s.

ASCII .dat → miniSEED (dat2seed.py)

from dat2seed import dat_to_mseed

# use the SAME sensitivity used during export to restore original counts
dat_to_mseed("converted_waveform.dat", "restored_sample.mseed",
             sensitivity=1.0, station="ST01")

Sampling rate is recovered from the time column; absolute time and station metadata (lost by the .dat format) are supplied as arguments.

Headless OmegaBox → SAC pipeline

cd "SEED2SAC/GaiaCode Pipeline"
# edit config.yaml: station list, archive roots, per-unit web_url
../venv/bin/python gaiacode_pipeline.py [--config config.yaml]

One-shot per invocation — drive the schedule with cron (see the pipeline's own README for deployment details). Exit code 0 when every file converted, 1 if any file failed.

Data Formats & Protocols

• miniSEED — input/output waveform format (Steim-compressed records); read/written via ObsPy.
• SAC (Seismic Analysis Code) — output of the pipeline; one file per channel per continuous segment, named NET.STA.LOC.CHAN.D.YYYY.JJJ.HHMMSS.SAC. Headers populated with STLA/STLO/STEL (coordinates), CMPAZ (azimuth), and CMPINC = dip + 90 (SEED dip → SAC inclination convention).
• StationXML (SeisComP schema) — per-station metadata used for SAC header injection; example files with placeholder coordinates included.
• ASCII .dat — two-column Time(s)/Amplitude text for SeismoSignal ingestion.
• HTTP — OmegaBox units expose their /data/ archive tree over plain HTTP; the pipeline mirrors it with wget -m (timestamping, only new/changed files transferred).