Save raw audio from WSPR for science analysis

Related: Load raw WSPR data for analysis

Upload your raw .wav WSPR data to the Zenodo data archive with this procedure. Please contact me with any questions.

1. Saving WSPR science raw data

(One-time setup)

In the main WSPR window, click:

Save > Save All

File > Save User Parameters

This saves one .wav file per two minute cycle.

2. Compress and upload raw WSPR data to science archive

Create three files:

  • MANIFEST: a list of filenames in the file so others don’t have to extract the file to see what’s in it
  • SHA256SUM: a checksum assuring the file was not changed since you created it
  • .tar.lz

This assumes you are using Linux / Mac / Windows Subsystem for Linux / Cygwin

cd ~/.wspr/save

find . *.wav > MANIFEST

nice tar cvaf mycall.tar.lz *.wav

nice sha256sum mycall.tar.lz > SHA256SUM

3. Upload raw data to Zenodo

Create a Zenodo account to upload WSPR data to Zenodo. When you click Publish the data is assigned a DOI and is citable.


Where are WSPR raw data saved

operating system WSPR .wav file location
Mac ~/.wspr/save
Linux ~/.wspr/save
Windows c:/wspr/save

In general you cannot use a virtual machine for WSPR because of the issues with broken/choppy audio.

How much raw data does WSPR save

Consider saving about one compressed file per day or week or so to avoid gigantic files to download. WSPR collects about 1.7 GByte/day depending on how often you transmit (no recording occurs when you transmit).

This file size is estimated by 12000 samples/sec * 16 bits/sample / 8 bits/byte * 86400 sec/day * 0.8 RX duty cycle = 1.7 GByte/day

Lzip will compress this to about 85% of original file size (the WSPR .wav files are themselves uncompressed PCM audio). The main point of compression is to make them one file instead of thousands of files per day. Don’t bother with intense compression settings because the files are full of noise, which by definition in general is poorly compressible.

Note on compression

  • replace mycall and date with your callsign and approximate date recording began for this group of files.
  • nice is used to help avoid disrupting the WSPR operation that is probably running while you’re compressing.