Update README.rst

This commit is contained in:
Michal Krenek (Mikos) 2017-02-19 11:37:30 +01:00
parent e7b998f69d
commit dc27e90653

View File

@ -1,7 +1,8 @@
QSpectrumAnalyzer
=================
Spectrum analyzer for RTL-SDR (GUI for rtl_power based on PyQtGraph)
Spectrum analyzer for multiple SDR platforms (PyQtGraph based GUI for soapy_power,
rx_power, rtl_power, hackrf_sweep and other backends)
Screenshots
-----------
@ -16,14 +17,40 @@ Requirements
- Python >= 3.3
- PyQt >= 4.5
- PyQtGraph (http://www.pyqtgraph.org)
- rtl-sdr (https://github.com/keenerd/rtl-sdr)
- Optional: rtl_power_fftw (https://github.com/AD-Vega/rtl-power-fftw)
- soapy_power / rx_tools / rtl-sdr / rtl_power_fftw / hackrf
Backends
--------
Universal SDR backends
**********************
- soapy_power (https://github.com/xmikos/soapy_power)
``soapy_power`` is default recommended universal backend in QSpectrumAnalyzer.
It is based on `SoapySDR <https://github.com/pothosware/SoapySDR>`_ and supports
nearly all SDR platforms (RTL-SDR, HackRF, Airspy, SDRplay, LimeSDR, bladeRF,
USRP and some other SDR devices).
- rx_power (https://github.com/rxseger/rx_tools)
``rx_power`` (part of ``rx_tools``) is also based on SoapySDR and therefore
supports nearly all SDR platforms, but it is much slower than soapy_power, doesn't support
near real-time continuous measurement (minimum interval is 1 second - same as ``rtl_power``)
and is little buggy.
RTL-SDR backends
****************
- rtl_power (https://github.com/keenerd/rtl-sdr)
You should use `Keenerds fork of rtl-sdr <https://github.com/keenerd/rtl-sdr>`_
(latest Git revision), because ``rtl_power`` in original rtl-sdr (from osmocom.org)
(latest Git revision), because ``rtl_power`` in original rtl-sdr package (from osmocom.org)
is broken (especially when used with cropping).
Another alternative is
- rtl_power_fftw (https://github.com/AD-Vega/rtl-power-fftw)
Another alternative for RTL-SDR is
`rtl_power_fftw <https://github.com/AD-Vega/rtl-power-fftw>`_ which has various
benefits over ``rtl_power``. E.g. better FFT performance (thanks to
use of ``fftw`` library) and possibility to use much shorter acquisition time
@ -31,17 +58,26 @@ for more real-time continuous measurement (minimum interval in original
``rtl_power`` is 1 second, but in ``rtl_power_fftw`` you are only limited
by number of frequency hops).
HackRF backends
***************
- hackrf_sweep (https://github.com/mossmann/hackrf)
``hackrf_sweep`` backend enables wideband spectrum monitoring by rapidly retuning the radio
without requiring individual tuning requests from the host computer. This allows unprecedented
sweep rate of 8 GHz per second.
Usage
-----
Start QSpectrumAnalyzer by running ``qspectrumanalyzer``.
You can choose if you want to use ``rtl_power`` or ``rtl_power_fftw`` backend in
*File* -> *Settings* (default is ``rtl_power``). Path to ``rtl_power``
(or ``rtl_power_fftw``) executable can be also manually specified there. You can also
set waterfall plot history size in there. Default is 100 lines, be aware that
really large sweeps (with a lot of bins) would require a lot of system memory, so
don't make this number too big.
You can choose which backend you want to use in *File* -> *Settings*
(default is ``soapy_power``). Sample rate and path to backend executable
can be also manually specified there. You can also set waterfall plot
history size. Default is 100 lines, be aware that really large sweeps
(with a lot of bins) would require a lot of system memory,
so don't make this number too big.
Controls should be intuitive, but if you want consistent results, you should
turn off automatic gain control (set it to some fixed number) and also set
@ -55,19 +91,31 @@ table can be changed in mini-histogram widget (on *Levels* tab).
Installation
------------
**Arch Linux:**
Arch Linux:
***********
Stable version:
::
git clone https://aur.archlinux.org/qspectrumanalyzer.git
cd qspectrumanalyzer
makepkg -sri
Git master branch:
::
git clone https://aur.archlinux.org/qspectrumanalyzer-git.git
cd qspectrumanalyzer-git
makepkg -sri
Or simply use `pacaur <https://aur.archlinux.org/packages/pacaur>`_ (or any other AUR helper):
::
pacaur -S qspectrumanalyzer
pacaur -S qspectrumanalyzer-git
**Debian / Ubuntu:**
Debian / Ubuntu:
****************
::
sudo apt-get install python3-pip python3-pyqt4 python3-numpy
@ -84,9 +132,20 @@ You can install it locally only for your current user by running this (without `
Executables will be then placed in ``~/.local/bin`` directory, you can add it to your
PATH in ``~/.bashrc``.
If you want to install QSpectrumAnalyzer directly from Git master branch, you can use this procedure:
::
git clone https://github.com/xmikos/qspectrumanalyzer.git
cd qspectrumanalyzer
pip3 install --user .
Todo:
-----
- finish soapy_power backend (new universal default backend)
- show scan progress
- allow setting LNB LO frequency
- save & load FFT history (allow big waterfall plot saved to file)
- automatic peak detection / highlighting
- display average noise level
- frequency markers / bookmarks with notes (even importing / exporting .csv file with