DSDPlus DSD-1.8.2

[gary123]

Iwvmobile: there have been a few minor tweaks to the code since the 1.8.2 build. I would call the current build 1.8.3 when I get it built

 

[adcockfred]

Network Chuck has got it going on, on u-tube, WSL-2 with kernel & gui

 

[gary123]

Update for those following the thread. Updated built files v1.8.3 are now available on the link provided earlier.

 

[adcockfred]

[ 15%] Building C object CMakeFiles/dsd.dir/src/dsd_dibit.c.o
[ 17%] Building C object CMakeFiles/dsd.dir/src/dsd_file.c.o
/home/freda/dsd-master/dsd-master/src/dsd_file.c: In function ‘closeMbeOutFile’:
/home/freda/dsd-master/dsd-master/src/dsd_file.c:179:7: warning: implicit declaration of function strptime’; did you mean ‘strftime’? [-Wimplicit-function-declaration]
179 | strptime (opts->mbe_out_file, "%s.imb", &timep);
| ^~~~~~~~
| strftime
[ 19%] Building C object CMakeFiles/dsd.dir/src/dsd_filters.c.o
[ 20%] Building C object CMakeFiles/dsd.dir/src/dsd_frame.c.o
[ 22%] Building C object CMakeFiles/dsd.dir/src/dsd_frame_sync.c.o
[ 23%] Building C object CMakeFiles/dsd.dir/src/dsd_main.c.o
/home/freda/dsd-master/dsd-master/src/dsd_main.c: In function ‘main’:
/home/freda/dsd-master/dsd-master/src/dsd_main.c:577:16: warning: ‘optarg’ redeclared without dllimport attribute: previous dllimport ignored [-Wattributes]
577 | extern char *optarg;
| ^~~~~~
/home/freda/dsd-master/dsd-master/src/dsd_main.c:578:14: warning: ‘optind’ redeclared without dllimport attribute: previous dllimport ignored [-Wattributes]
578 | extern int optind, opterr, optopt;
| ^~~~~~
/home/freda/dsd-master/dsd-master/src/dsd_main.c:578:22: warning: ‘opterr’ redeclared without dllimport attribute: previous dllimport ignored [-Wattributes]
578 | extern int optind, opterr, optopt;
| ^~~~~~
/home/freda/dsd-master/dsd-master/src/dsd_main.c:578:30: warning: ‘optopt’ redeclared without dllimport attribute: previous dllimport ignored [-Wattributes]
578 | extern int optind, opterr, optopt;
| ^~~~~~

 

[adcockfred]

reda@DESKTOP-D9RVSN7 ~/dsd-master/dsd-master/build
$ make install
[ 76%] Built target dsd
[ 79%] Built target gmock
[ 82%] Built target gtest
[ 92%] Built target dsdtest
[ 95%] Built target gtest_main
[100%] Built target gmock_main
Install the project...
-- Install configuration: ""
-- Installing: /usr/local/bin/dsd.exe


Yeah, well there it is dsd 1.8.3 built on Cygwin64 with just one error on package download from respository.

I broke the 32 bit version Cygwin.

So we will just see if I have to fix it, for the audio in. There was port audio loaded in both versions of Cygwin.


nap time , back to salt mines tonight

 

[adcockfred]

P.S. both itpp-4.3.1 and mbelib installed with no warnings

 

[kb9mwr]

Great, thanks for bringing this update to my attention.

I'd still love to see someone try and improve the AMBE used for D-Star. Its pretty crude sounding since it done based off reverse engineering/guess and test, vs all the other AMBE code was based off TIA technical publications.

Here are some starting points for those with the skills:

nostar/dudestar
f4exb/dsdcc - dsd rewrite - C++ library with a single decoder object
LX3JL/xlxd
szechyjs/dsd
szechyjs/mbelib
LouisErigHerve/dsd
http://git.osmocom.org/op25/tree/op25/gr-op25_repeater/lib - Op25 has encode and decode support for AMBE (D-Star, DMR and YSF) and IMBE (P25)
balint256/op25 - Pavel's IMBE Encoder/Decoder Fixed-Point implementation
http://git.osmocom.org/osmo-gmr/tree/src/codec
on1arf/voice-ann
travisgoodspeed/md380tools - The MD380 Emulator (capable of AMBE encoding and decoding)
https://github.com/g4klx/AMBETools

 

[lwvmobile]

Great, thanks for bringing this update to my attention.

I'd still love to see someone try and improve the AMBE used for D-Star. Its pretty crude sounding since it done based off reverse engineering/guess and test, vs all the other AMBE code was based off TIA technical publications.

You aren't the only one with a wish list for DSD and open source decoding software. I think the top of my list has coverting the audio system in DSD to use ALSA/Pulse Audio instead of OSS/Port Audio. My programming chops are pretty poor, but I do tinker here and there. One day, I'm going to look into ways to get that audio output going and doing something better than 8k mono output that sounds like Charlie Brown's teacher talking. Not so much on the decoding aspect, just the output audio quality and audio system it uses.

 

[kb9mwr]

You aren't the only one with a wish list for DSD and open source decoding software. I think the top of my list has coverting the audio system in DSD to use ALSA/Pulse Audio instead of OSS/Port Audio. My programming chops are pretty poor, but I do tinker here and there. One day, I'm going to look into ways to get that audio output going and doing something better than 8k mono output that sounds like Charlie Brown's teacher talking. Not so much on the decoding aspect, just the output audio quality and audio system it uses.

Yeah there I work to be done, and I am thrilled to see Louis step up to the plate. I have an AMBE dongle that I'd certainly loan to anyone with the skills and desire to make improvements.

 

[R0am3r]

FYI for everyone there is no code for decoding BP.

What is BP?

 

[kb9mwr]

What is BP?

Basic Privacy I assume, used in DMR.

 

[boatbod]

What is BP?

Basic Privacy I assume, used in DMR.

BP is just an XOR of certain bits of the xMBE codewords using one of the 255 predefined 16 bit keys. It is trivially easy to reverse if you know (a) which bits are xor'd and (b) either the expected codewords or the expected key value.

 

[thewraith2008]

It's a case when silence is key

 

[KA1RBI]

I'd still love to see someone try and improve the AMBE used for D-Star. Its pretty crude sounding since it done based off reverse engineering/guess and test, vs all the other AMBE code was based off TIA technical publications.

It seems sad that various national amateur radio organizations have not stepped up to support this effort. Also, it should be noted that TIA specs are incomplete in that they document what is known in APCO-world as the "legacy" xMBE vocoder. The specs are unapologetic when telling you that all of the "enhancements" to the vocoder (roughly speaking, the "+2" stuff) is undocumented. Paraphrasing, they also say something like "information not needed for interoperability is not included".

As far as the quality of the existing ambe that is in dsd/d-star, I've become convinced that the following cryptic comment in the original DSD provides the necessary clue:

  // decode V/UV parameters
  // load b1 from ambe_d
  //TODO: use correct table (i.e. 0x0000 0x0005 0x0050 0x0055 etc)

73

Max

 

[adcockfred]

Yeah, these people finally settled down, after the little blow er big blow. I remember in high school in 69' it was just like this storm. We could see it in the east in the daytime and were glad the 175 mph winds were over there towards the river. Which ever river sabine mississippi.
Ok back to bussines, I got those build errors fixed. In the CMakeLists.txt added add_definitions (-D_XOPEN_SOURCE=700) and in the dsd_file.c added #include <time.h>
this strptime is written in C

 

[adcockfred]

KB9
this is really cool HERVE has a video on his packages for cygwin
I have been to the grind stone, I did get dsd.exe to open.
I just cp & mv that missing dll from mbe to dsd
PPS do not try that new KEX for gui in kali, I been hard on wsl-2 also

 

[adcockfred]

This is what I was telling LWV, and that is why i coded the home folder. the mbe files are not being coppied to root- bin folder because you are in the win-10 home dir.
so it is a tree problem maybe boatbod could shed some light
has that other linux ghrue passed ? you know the one that looks like an old radio

 

[adcockfred]

PortAudio version number = 1246720
PortAudio version text = 'PortAudio V19.6.0-devel, revision 396fe4b6699ae929d3a685b3ef8a7e97396139a4'
Number of devices = 2
--------------------------------------- device #0
Name = pulse
Host API = ALSA
Max inputs = 32, Max outputs = 32
Default sample rate = 44100.00
--------------------------------------- device #1
[ Default Input, Default Output ]
Name = default
Host API = ALSA
Max inputs = 32, Max outputs = 32
Default sample rate = 44100.00


this build is ubuntu 20.04 with VcXsrv dsd 1.8.3, wsl 1, but i am running win 10 2004 with 2008 updates also linux ext and vm enabled. no kernel installed.

 

[thewraith2008]

Is it just me or can no one else follow what the OP is trying to do, or say for that matter.
After three pages, I still can't determine what your trying to do. You may see more response if you can clarify this.

What are you trying to do:

• Compile DSD in Linux so you can run it in Linux
• Compile DSD in Linux so you can run it in Windows(10)
• Compile DSD in Windows(10) so you can run it in Windows(10)

 

[adcockfred]

yeah that was the crux of the problem. run dsd on linux on win 10 and then run dsd in win 10 at the same time. then shut down the both dsd runs, then restart the linux dsd to populate the sc number. so if the population of the dependecies worked like in a straight version of linux everything would work, but of cource that was not the case. so all of the path problems, shared dll problems, what ever else problems have been solved by physically populating itpp dll, mbe dll and dsd in shared folder now that audio in has been established in wsl-1. I liken it to having a tiger by the tail. i had to fall back to ubuntu where i had some chance of success, but i do like that kali.