Skip to main content
Join the official 2019 Python Developers SurveyStart the survey!

D-STAR library and utilities in Python

Project description

pydv

Collection of Python code to experiment with D-STAR and the proposed vocoder extension that allows the use of the open source Codec 2 with D-STAR.

Provides Python interfaces to manage DExtra and DPlus connections (protocols used by reflectors), convert from network data to D-STAR streams (header and frames) and vice versa, as well as encode and decode voice data using mbelib (decode only) and codec2, and transcode using an AMBEd server (the version included in my xlxd fork).

Installs the following executables:

  • dv-recorder, which connects to a reflector and records traffic in .dvtool files
  • dv-player, which plays back a .dvtool file to a reflector
  • dv-encoder, which converts a .wav fle to a .dvtool file using the Codec 2 vocoder
  • dv-decoder, which converts a .dvtool file using any vocoder to .wav
  • dv-transcoder, which connects to an AMBEd server and converts a .dvtool file using the AMBE vocoder to a .dvtool file using the Codec 2 vocoder and vice versa

D-STAR vocoder extension

I propose the use of the "Flag 3" byte of the header, to mark the vocoder type in the voice frames, as follows (in accordance to section 2.1.1, page 4 of the D-STAR specification):

Bit Meaning Function
0000000x Vocoder 0: AMBE (backwards compatible)
1: Codec 2
000000x1 Mode 0: Codec 2 3200 (160 samples/20 ms into 64 bits)
1: Codec 2 2400 (160 samples/20 ms into 48 bits) plus FEC (22 bits)
00000100 to 11111111 Undefined Use for future expansion

The total space available in the frame for voice data is 72 bits. With Codec 2 3200, there is not much space left for FEC, so this option should only be used for communications over the Internet. On the other hand, using Codec 2 2400, allows us to protect the first 24 bits of the voice data with two applications of the (23, 12) Golay code.

Note: An early version of the extension used one more bit to control the presence of FEC. It was later decided to always include FEC when there is available space in the frame.

The vocoder extension is compatible with all current D-STAR hardware (repeaters, hotspots, etc.) and software (repeater controllers, reflectors, etc.), except - of course - transceivers that assume voice data to be in AMBE format and use the corresponding chip for processing.

D-STAR reflectors, like xlxd, can be used to transcode and bridge the two formats. Although implementations of the vocoder extension can use the same reflector connections as AMBE-only clients, it should be avoided, to save the user confusion and establish interoperability.

The solution implemented in the vocoder-extension branch of my xlxd fork uses another DExtra listener on a different port (30201 instead of 30001). The new port is to be used by reflectors using the "ORF" prefix (Open ReFlector). Any client connected to an ORF reflector will receive streams encoded with Codec 2. All other D-STAR protocol handlers will still send out data encoded with AMBE. Note that the protocol/port only affects data transmitted by the reflector. The stream vocoder is recognized by all protocol handlers, so a client can still transmit data using any vocoder on any port. The rationale behind this is that DExtra links may be used by repeaters or other reflectors, so it is not really possible to know what their clients support. So, nothing will change when linking a repeater to an XRF reflector, but will do when linking to an ORF one.

The open source vocoder, allows homebrewing transceivers using a Rasbperry Pi, an MMDVM modem (even one constructed with through-hole components), and an old radio. Thus, one could use a D-STAR hotspot as a transceiver, assuming a method to attach a microphone and speaker. It also allows using software clients to communicate through reflectors without the need of any AMBE hardware.

All included utilities implement the vocoder extension. Use the -p dextraopen flag, or substitude the reflector's callsign prefix with "ORF", to make dv-player and dv-recorder use the "open" DExtra port.

Building

To build, you must first build and install mbelib and codec2.

On Mac OS X, I used MacPorts to install cmake, speexDSP, and libsamplerate. mbelib compiles and installs to /usr/local without problems. To build codec2, I had to export LIBRARY_PATH=$LIBRARY_PATH:/opt/local/lib before running make, and edit the following files to remove unsupported gcc flags (from the build folder):

  • unittest/CMakeFiles/ofdm_stack.dir/flags.make, to remove -fstack-usage
  • unittest/CMakeFiles/ofdm_stack.dir/link.txt, to remove -Wl,-Map=ofdm_stack.map

Based on ircDDBGateway and xlxd. Tested with my xlxd fork.

73 de SV9OAN

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for pydv, version 2.0
Filename, size File type Python version Upload date Hashes
Filename, size pydv-2.0.tar.gz (22.3 kB) File type Source Python version None Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page