DAB COMMAND LINE and DAB LIBRARY
The DAB library provides entries for the functionality to handle DAB/DAB+ through some simple calls. A few callback functions provide the communication back from the library to the caller. The library interface is given in dab-api.h
A number of example programs is included in the source tree, they are meant to give an idea on how to use the library code, either as library or as "built-in" sources. They might or might not work.
Note that the library depends on a device, but does not include that device. The main program is responsible for ensuring that something is available to deliver input samples and something is available for handling the output.
Invocation of the example programs, with some parameters specified, is something like
dab-sdrplay-x -M 1 -B "BAND III" -C 12C -P "Radio 4" -G 80 -A default
In this case, the example program was built with the SDRplay as device (other possibilities are DABsticks, AIRspy devices, HACKRF and Lime SDR devices (and some of the example programs can be configured to take file input)). Furthermore, the example program will set the tuner to Band III, channel 12C, will select the service "Radio 4". The Gain of the tuner is set to 80 (on a scale from 1 .. 100), the main program will select "default" as audio output device. The Library code will be set to interpret the input as being of Mode 1.
- SDRplay RSP's (using 2.13 lib or 3.06 lib)
- AIRspy
- RTLSDR based devices
- HACKRF (only example programs 2, 3, 4)
- LimeSDR (only example programs 2, 3 and 4)
- Pluto (only example 2)
and of course fileinput of ".raw" and ".sdr" files is supported, as well as input through the rtl_tcp driver.
Since I was unhappy with passing lots of individual callback functions through the whole of the computing chain, I made (am making) a change to the API and the dabInit function in the API.
Until recently one had to specify all individual callback functions as parameter to the dabInit function.
The change is that now a struct "API_struct" is defined in the API, the fields of which are the references to the different callback functions such that the call to dabInit is basically simplified to
theRadio = dabInit (theDevice,
&interface,
nullptr, // no spectrum shown
nullptr, // no constellations
nullptr // Ctx
);
Examples 1, 2, 3, 4, 5 and 6 are adapted. Other examples - apart from the python example are deleted.
The obvious advantage is that adding a callback function for a specific purpose now does not change all of the intermediate functions in tbhe library, basically opne only modifies
a. the structure
b. the main program for filling the structure and adding a handler
c. the affected function in the library
New callback functions are
a. the tii data
typedef void (*tii_data_t)(int);
where the int is encoded as mainId << 8 + subInt
b. the time
typedef void (*theTime_t)(std::string, void *);
where the time (hours::minutes) is passed as string
The handling of motdata is changed. In the previous version, slides were - automatically - written to a file. In the current version, the parameter profile of the motdata handling function is extended, the motdata is passed on as uint8_t array, the size of the data is passed, the name of the slides - as derived from the DAB data - is passed on. See the dab-api for details
The software is provided as is, it is available under the GPL-V2, the examples might or might not work. As far as I can see they work on my Linux-x86 box, no garantees are given that the library software or the examples will work on Windows or any other system.
THE EXAMPLES ARE NOT INTENDED TO BE FULL SOLUTIONS, BUT MERELY THERE TO GIVE YOU AN IDEA HOW TO USE THE LIBRARY OR ITS SOURCES. MODIFICATIONS IN THE LIBRARY ARE NOT ALWAYS TESTED ON ALL EXAMPLES, PLEASE CORRECT THEM YOURSELF IF YOU (THINK YOU) FIND AN ERROR BEFORE ASKING ME.
The examples are basically simple variations on a single theme, example 2 is the basic one, others are derived.
- example 1 is the example where the main program is linked to
a precompiled shared library,
i.e. the DAB library should be pre-installed
- example 2 has the same functionality as example 1, the sources
of the library are "compiled-in", however.
- example 3 has the same functionality as example 2, and here
the library sources are "compiled in" as well. However, the
PCM samples are being sent out to stdout.
One might use one of the available programs to make the sound
audible
dab-example-3 .... | aplay -r 48000 -f S16_LE -t raw -c 2
- example 4 has the sample functionality as examples 2 and 3, and
here the library sources are "compiled in" as well. However,
no sound decoding takes place. The MP2 frames (in case of DAB)
and the AAC frames (in case of DAB+) are just emitted
into a file or to stdout. A flag "-f XXXX" to the command
line of this example, if specified, output is written to
the specified file, otherwise to stdout.
The output can be processed by e.g. VLC.
(Note that the AAC frames have 960 rather than 1024 samples,
not all audio programs are capable of handling these).
- example 5 is a small experimental extension to example 2,
It contains a simple "keyboard listener", that will react
on entering a stroke on the return key. It will cause the
"next" (audio) service to be selected.
In example 5 there is support for tdc packet handling,
and - thanks to stefan Juhl - support for FEC protected packets.
- example 6 is an experimental version where stdin is
used as input device (and the command line parameters are
adapted to that)
- the python example seems to work, the cmake file now expects
python3.9 (it should work with other versions as well, adapt the
CMakeLists.tct file)
- the scanner example is what the name suggests, it scans the band
and shows the content of the channels that carry (detectable)
DAB data
For all examples it holds that NO garantee is
given on their functioning, feel free to improve.
Next to these examples, a simple dab-scanner was made, an example program that just scans all channels of the given band (BAND III by default) and collects and emits data about the ensembles and services encountered. Output can be sent to a file - ASCII - that can be interpreted by Libre Office Calc or similar programs.
The dab-scanner supports rtlsdr, sdrplay (with support for the 2.13 lib and the 3.XX libraries), airspy, hackrf, and lime sdr.
As with other sdr programs, a number of libraries is to be installed. Assuming that the program is to be compiled in a Debian based system (e.g. Ubuntu)
sudo apt-get update
sudo apt-get install git cmake
sudo apt-get install build-essential g++
sudo apt-get install pkg-config
sudo apt-get install libsndfile1-dev
sudo apt-get install libfftw3-dev portaudio19-dev
sudo apt-get install libfaad-dev zlib1g-dev
sudo apt-get install libusb-1.0-0-dev mesa-common-dev
sudo apt-get install libgl1-mesa-dev
sudo apt-get install libsamplerate0-dev
Installing the library for the SDRplay requires downloading the library from the "www.SDRplay.com" site.
For pluto the "libiio-dev" has to be installed. Note that on "old" Ubuntu versions, e.g. 16.04, only an old, not yet complete, version of the library is available.
Installing the library for the RTLSDR can be done by executing the following lines
git clone git://git.osmocom.org/rtl-sdr.git
cd rtl-sdr/
mkdir build
cd build
cmake ../ -DINSTALL_UDEV_RULES=ON -DDETACH_KERNEL_DRIVER=ON
make
sudo make install
sudo ldconfig
cd ..
rm -rf build
cd ..
Installing the library for the AIRspy can be done by executing the following lines
wget https://github.com/airspy/host/archive/master.zip
unzip master.zip
cd airspyone_host-master
mkdir build
cd build
cmake ../ -DINSTALL_UDEV_RULES=ON
make
sudo make install
sudo ldconfig
cd ..
rm -rf build
cd ..
Once the libraries are installed, creating an executable for example 2 is straightforward
cd dab-cmdline
cd example-2
mkdir build
cd build
cmake .. -DXXX=ON (replace XXX by the name of the device)
make
sudo make install
cd ../../..
The resulting executable is installed in "/usr/local/bin"
The full API description is given in the file dab-api.h
Initialization of the library is by a call to "dabInit". The call returns a pointer (type void *) to structures internal to the library.
"dabInit" gets pointers to callback functions as parameter and returns
a "Handle", to be used in the other functions
"dabExit" cleans up all resource use of the library.
"dabReset" cleans up resource use and restarts the library
"dabStartProcessing" returns immediately after being called
but will have created a few threads running in the background.
As soon as an ensemble is recognized, a callback function will be called. Recognition of a service within an ensemble also leads to calling a callback function, with the service name as parameter.
"dabReset_msc" will terminate operation of the handler(s) for the
currently selected service
"is_audioService" can be used to enquire whether or not a service
with a given name is a recognized audio service,
"is_dataService" can be used to enquire whether or not a service
with a given name is a recognized data service,
"dataforAudioService" is the function with which the (relevant) data,
describing an audio service with a given name is fetched.
The function fills a structure of type "audiodata",
it contains a field "defined" telling whether or not the
data is the structure is valid or not.
Note that if the last parameter is a 0-value, the main service
is looked for, otherwise, the i-th subservice.
"dataforDataService" is the function with which the (relevant) data,
describing a data service with a given name is fetched.
The function fills a structure of type "packetdata",
it contains a field "defined" telling whether or not the
data is the structure is valid or not.
Note that if the last parameter is a 0-value, the main service
is looked for, otherwise, the i-th subservice.
"set-audioChannel", when provided with a structure of type "audiodata",
with valid data, will open and add an audio stream.
"set-dataChannel", when provided with a structure of type "packetdata",
with valid data, will open and add a data stream.
Note that there is no built-in limit on the amount of open streams,
although from a practical point of view there may be limitations,
i.e two audiostreams may compete for a single library.
What might be useful is to enquire for subservices such as MOT
when opening an audio stream.
The library (whether separate or compiled in) sends its data to the main program using callbacks. These callbacks, the specification of which is given in the file dab-api.h, are implemented here as simple C functions. WHAT MUST BE NOTED IS THAT THESE FUNCTIONS ARE EXECUTED IN THE THREAD OF THE CALLER, and while the library is built around more than a single thread, it is wise to add some locking when extending the callback functions.
========================================================================
The library can be created by - if needed - adapting the
CMakeLists.txt file in the dab-library/library directory and running
mkdir build
cd build
cmake ..
make
sudo make install
from within the dab-library directory.
IMPORTANT: YOU NEED C++11 SUPPORT FOR THIS
Note that contrary to earlier versions, the "device" is NOT part of the library, the user program has to provide some functions to the library for getting samples. The interface can be found in the file "device-handler.h".
===============================================================================
Libraries (together with the "development" or ".h" files) needed for creating the library are
libfaad
libfftw3f
libusb-1.0
zlib
============================================================================
For the python-example read the README file in the python-example directory. HOWEVER: before running the example program one has to create an ADAPTED library. The CMakeLists.txt file for creating such an adapted library is in the python directory.
=============================================================================
The programs accept the following command line parameters:
-B Band
selects the DAB band (default Band III),
-M Mode
selects the DAB Mode (default Mode 1),
-C the channel
the default is 11C, the channel I am listening to mostly,
-P the program name
a prefix suffices. For e.g. "Classic FM" it suffices to give "Classic". However, when passing on a non-unique prefix (e.g. "Radio" for "Radio Maria" and "Radio Veronica") the software will select one arbitrarily. Note that letter case is IMPORTANT is the current version. The names of the programs in the ensemble being received in the selected channel will be printed during recognition.
Important: If no program names are found, or if no match can be made between the program name and the list of program names, the program has no other choice than to halt, what it does.
-G the gain
to be applied on the device, a value in the range from 1 .. 100. The value will be translated to an acceptable value for the device. In case the gain is table driven, as in the case of a dabstick, a value of e.g. 75 is translated into the element on three quarters of the table (basic assumption is that the table elements are more or less linear). For e.g. the Airspy the values are mapped upon the range 0 .. 21 of the sensitivity slider. Note that when using the rtl_tcp interface, this does not hold. The sound setting is passed on to the server.
-W waiting time
the maximum time to wait for valid data. If no valid data is found within this period, execution of the program will stop.
Note that in example-2 the -W is replaced by TWO parameters, a -d xxx indicating the maximum amount of time to reach time synchronization (which is implicit in the other examples) and a -D xxx parameter indicating the maximum amount of time to get the signal "ensemble found".
-A the output channel (example 1 and 2 only)
again as with the program name, a prefix of the name suffices. As with the programs, the names of the sound channels identified will be printed. Note, however, that in Linux not all all names appearing on the namelist are useful, some of them will just not work, a well known issue with the combination portaudio/alsa under Linux. Important: If a name is selected for a channel that cannot be opened the program will try to open the default output device.
-O filename or "-" (example 2 only)
The PCM samples of the sound output are stored in the file . If "-" is specified as filename the output is just written to stdout. This output then can be made audible by some other program.
-H hostname (example 2, 3 and 4 only)
If rtl_tcp is selected as input device, the -H option allows selection of a hostname. Default is "127.0.0.1".
-I port (example 2, 3, and 4 only)
If rtl_tcp is selected as input device, the -I option allows selection of a port. Default is 1234.
In order to use the library, the main program has to deal with two interfaces, obviously the dab library, but also the device.
The device interface is specified in "./device-handler.cpp".
One of the issues still to be resolved is the handling of data. As an experiment a callback function was added that is called from within the tdc handler. In example-2 a simple TCP server was added, one that just writes out packaged tdc frames. The package structure is : an 8 byte header followed by the frame data. The header starts with a -1 0 -1 0 pattern, followed by a two byte length, followed by a zero, followed by a 0 for frametype 0 and 0xFF for frametype 1. Install the server by adding "-DSERVER" to the cmake command line.
A simple "reader" (client), using qt is included in the sources.
Copyright (C) 2016, 2017, 2018
Jan van Katwijk ([email protected])
Lazy Chair Computing
The dab-library software is made available under the GPL-2.0. The dab-library uses a number of GPL-ed libraries, all rigfhts gratefully acknowledged. All SDR-J software, among which dab-library is one - is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.