pip install sdrterm
*NOTE: Unless you intend use code on the bleeding-edge, or to modify the source, skip ahead to here.
After cloning this repo perform the following:
cd <path/to/clone>
python -m venv .venv
(if using windows)
.\.venv\Scripts\activate
(otherwise)
source .venv/bin/activate
(either)
pip install .
(or to include optional graphing dependencies)
pip install .[gui]
...<do stuff here>...
deactivate
Alternatively--if you do not wish to use a virtual environment--the dependencies may be installed via the system's
package manager; or, with caution, using pip directly.
Plotting support is optional, and dependencies can be installed by specifying by the
[gui] option during install, or if they happen to already be installed in the environment.
- For raw files, input is system-default endianness,
- For wave files, the endianness defined by the RIFF/X header
- For tcp connections, data is expected to be big-endian
- Endianness can be inverted using the
-Xflag - When necessary to specify the input datatype (
-eflag), the selections map exactly to the "integer" and "float" types listed here
- Standard mode outputs doubles (float64) with system-default endianness and alignment
- Multiple VFO mode (
--simoflag) output is always big-endian doubles
- Be aware that piping binary (i.e. non-text) data between processes in Powershell is only natively-supported in Powershell v7.4+ (https://stackoverflow.com/a/68696757/8372013), which you may have to install separately from the default version included with your copy of Windows. N.B. this does not affect modes that use sockets for IO.
python -m sdrterm -i file.wav --omega-out=5k --decimation=64 --center-frequency="15k" --plot=spec --correct-iq -o out.bin
- Input source: wave file
- Input data type: determined by RIFF/X header metadata
- Sample rate: determined by RIFF/X header metadata (
$fs$ $S \over s$ ) - Output lp: 5 kHz
- Decimation factor:
$64 \implies$ $fs \over 64$ $S \over s$ - Offset from tuned frequency: +15 kHz
- Plot(s): Spectrum Analyzer
- IQ correction: enabled
- Output destination: out.bin
python -m sdrterm -i <host>:<port> -eh -r1024k -w5k -d64 -c"-30k" --plot=water | ...
N.B. sampling rate and datatype must be specified for raw input
- Input source: TCP socket at <host> on <port>
- Input data type: 16-bit, signed integer
- Sample rate:
$1024k$ $S \over s$ - Output lp: 5 kHz
- Decimation factor:
$64 \implies$ $1024k \over 64$ $= 16k$ $S \over s$ - Offset from tuned frequency: -30 kHz
- Plot(s): Waterfall
- Output destination: stdout
Select multiple frequencies to process from data input via a socket and output them via separate sockets
See exmaple_simo.sh for a complete example
More examples can be found in the shell scripts in the root of this repo.
Usage: sdrterm.py [OPTIONS]
╭─ Options ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ --fs -r NUMBER Sampling frequency in k/M/Samples per sec │
│ --center-frequency -c NUMBER Offset from tuned frequency in k/M/Hz [default: 0] │
│ --input -i TEXT Input device [default: (stdin)] │
│ --output -o TEXT Output device [default: (stdout)] │
│ --plot TEXT 1D-Comma-separated value of plot type(s) [default: None] │
│ --demodulation -m [fm|am|re|im] Demodulation type [default: fm] │
│ --tuned-frequency -t NUMBER Tuned frequency in k/M/Hz [default: 0] │
│ --vfos TEXT 1D-Comma-separated value of offsets from tuned frequency to process in addition to tuned frequency in k/M/Hz [default: None] │
│ --decimation -d INTEGER RANGE [x>=2] Decimation factor [default: 2] │
│ --encoding -e [b|B|h|H|i|I|f|d] Binary encoding (ignored if wav file) [default: None] │
│ --omega-out -w NUMBER Output cutoff frequency in k/M/Hz [default: 12500] │
│ --correct-iq --no-correct-iq Toggle iq correction [default: no-correct-iq] │
│ --simo --no-simo Enable using sockets to output data processed from multiple channels specified by the vfos option. N.B. unlike normal mode, which uses the │
│ system-default endianness for output, the sockets output network-default, big-endian doubles. [Implies: --vfos <csv>] │
│ [default: no-simo] │
│ --verbose -v INTEGER Toggle verbose output. Repetition increases verbosity (e.g. -vv, or -v -v) [default: 0] │
│ --smooth-output INTEGER Provide length of polynomial for smoothing output with Savitzky–Golay filter. A larger polynomial implies more aggressive filtering. │
│ [default: (0 => no filtering)] │
│ --vfo-host TEXT Address on which to listen for vfo client connections [default: localhost] │
│ --swap-input-endianness -X Swap input endianness [default: (False => system-default, or as defined in RIFF header)] │
│ --normalize-input --no-normalize-input Normalize input data. [default: no-normalize-input] │
│ --help Show this message and exit. │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
python -m rtltcp <host> <port>
Usage: rtltcp.py [OPTIONS] HOST PORT
╭─ Arguments ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ * host TEXT Address of remote rtl_tcp server [default: None] [required] │
│ * port INTEGER Port of remote rtl_tcp server [default: None] [required] │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Options ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ --server-host TEXT Port of local distribution server [default: localhost] │
│ --verbose -v INTEGER Toggle verbose output. Repetition increases verbosity (e.g. -vv, or -v -v) [default: 0] │
│ --help Show this message and exit. │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
Due to the experimental nature of this interface, do not hesitate to report suspected bugs.
N.B. if you'd like to be sure of your wave file's sampling rate beforehand,
you can use a tool like mediainfo available via most POSIX-like systems' package managers.
Also, basic information about the input file, and other settings are output in JSON format at beginning
of sdrterm's operation. e.g.,
{
"bandwidth": 12500,
"centerFreq": -25000,
"tunedFreq": 0,
"omegaOut": 5000,
"smooth": 0,
"vfos": "15000,-15000",
"host": "localhost",
"encoding": "uint8",
"fs": 1024000,
"decimatedFs": 16000
}