The sigmf library makes it easy to interact with Signal Metadata Format (SigMF) recordings. This library is compatible with Python 3.7-3.14 and is distributed freely under the terms GNU Lesser GPL v3 License.

This module follows the SigMF specification html/pdf from the spec repository.

Install

pip install sigmf
# or
conda install sigmf
# or
mamba install sigmf

Read SigMF

import sigmf

# read SigMF recording
meta = sigmf.fromfile("recording.sigmf-meta")
samples = meta[0:1024]  # get first 1024 samples
sample_rate = meta.sample_rate  # get sample rate

# read compressed SigMF archives
meta = sigmf.fromfile("recording.sigmf.gz")   # gzip-compressed
meta = sigmf.fromfile("recording.sigmf.xz")   # xz-compressed
meta = sigmf.fromfile("recording.sigmf.zip")  # zip archive

# read other formats containing RF time series as SigMF
meta = sigmf.fromfile("recording.wav")   # WAV
meta = sigmf.fromfile("recording.cdif")  # BLUE / Platinum
meta = sigmf.fromfile("recording.xml")   # Signal Hound Spike

Write SigMF

import numpy as np
import sigmf

data = np.array([0.1 + 0.2j, 0.3 + 0.4j], dtype=np.complex64)
meta = sigmf.fromarray(data)
# optional additional metadata
meta.sample_rate = 8000
meta.description = "sample recording"
meta.add_capture(start_index=0, metadata={sigmf.FREQUENCY_KEY: 915e6})
# creates recording.sigmf-data and recording.sigmf-meta
meta.tofile("recording")
# or create compressed archive
meta.tofile("recording.sigmf.gz")

Docs

Please visit our documentation for full API reference and more info.