DSD FME

[dotignore]

If I press PTT long time you can see result command ($ dsd-fme -i tcp:127.0.0.1:7355 -N)
I use Ubuntu 22.04

 ██████╗  ██████╗██████╗     ███████╗███╗   ███╗███████╗  ESC to Menu
 ██╔══██╗██╔════╝██╔══██╗    ██╔════╝████╗ ████║██╔════╝  'q' to Quit
 ██║  ██║╚█████╗ ██║  ██║    █████╗  ██╔████╔██║█████╗
 ██║  ██║ ╚═══██╗██║  ██║    ██╔══╝  ██║╚██╔╝██║██╔══╝    MBElib 1.3.4
 ██████╔╝██████╔╝██████╔╝    ██║     ██║ ╚═╝ ██║███████╗  AW
 ╚═════╝ ╚═════╝ ╚═════╝     ╚═╝     ╚═╝     ╚═╝╚══════╝  2024-204-g90206ea
--Input Output----------------------------------------------------------------
| TCP Signal Input: 127.0.0.1:7355; 48 kHz; 1 Ch;
| Pulse Digital Output: 8 kHz; 2 Ch; G: 30% G: 30% (+|-) Auto   HPF
------------------------------------------------------------------------------
--Audio Decode----------------------------------------------------------------
| Demod/Rate:  [C4FM][4800]
| Decoding:    [AUTO]
| In Level:    [00%]
| Voice Error: [0][0] Slot 1 (1) ON
| Voice Error: [0][0] Slot 2 (2) ON
------------------------------------------------------------------------------
--Call Info-------------------------------------------------------------------
------------------------------------------------------------------------------
--Call History----------------------------------------------------------------
------------------------------------------------------------------------------

 

[dotignore]

Yeah, I forgot to have the color code decode or display on MS sourced audio. I'll put it on the to-do list.

this problem exists only for SLOT=1

it is not exist problem of decoding when two slots SLOT=1 SLOT=2 (repiter radio)

 

[lwvmobile]

If I press PTT long time you can see result command ($ dsd-fme -i tcp:127.0.0.1:7355 -N)

Try using: dsd-fme -i tcp:127.0.0.1:7355 -N 2> log.ans
You need to route stderr to the log using 2> to a log file or to /dev/null leaving only the ncurses terminal portion in stdout.

 

[dotignore]

Try using: dsd-fme -i tcp:127.0.0.1:7355 -N 2> log.ans
You need to route stderr to the log using 2> to a log file or to /dev/null leaving only the ncurses terminal portion in stdout.

it works good thanks

dsd-fme -i tcp:127.0.0.1:7355 -N 2> log.ans

 ██████╗  ██████╗██████╗     ███████╗███╗   ███╗███████╗  ESC to Menu
 ██╔══██╗██╔════╝██╔══██╗    ██╔════╝████╗ ████║██╔════╝  'q' to Quit
 ██║  ██║╚█████╗ ██║  ██║    █████╗  ██╔████╔██║█████╗
 ██║  ██║ ╚═══██╗██║  ██║    ██╔══╝  ██║╚██╔╝██║██╔══╝    MBElib 1.3.4
 ██████╔╝██████╔╝██████╔╝    ██║     ██║ ╚═╝ ██║███████╗  AW
 ╚═════╝ ╚═════╝ ╚═════╝     ╚═╝     ╚═╝     ╚═╝╚══════╝  2024-204-g90206ea
--Input Output----------------------------------------------------------------
| TCP Signal Input: 127.0.0.1:7355; 48 kHz; 1 Ch;
| Pulse Digital Output: 8 kHz; 2 Ch; G: 21% G: 30% (+|-) Auto   HPF
------------------------------------------------------------------------------
--Audio Decode----------------------------------------------------------------
| Demod/Rate:  [C4FM][4800]
| Decoding:    [AUTO]
| In Level:    [00%]
| Voice Error: [0][0] Slot 1 (1) ON
| Voice Error: [0][0] Slot 2 (2) ON
------------------------------------------------------------------------------
--Call Info-------------------------------------------------------------------
| DMR MS - DCC: 01;
| SLOT 1 - TGT: [       0] SRC: [       0]                       |
| V XTRA |
| D XTRA |
| SLOT 2 - TGT: [       0] SRC: [       0]                       |
| V XTRA |
| D XTRA |
|        |
|        |
------------------------------------------------------------------------------
--Call History----------------------------------------------------------------
| 2024-04-22 20:11:55 TGT [  255001] SRC [       1] DCC [01]
------------------------------------------------------------------------------

 

[RobDLG]

Thought I'd just take a moment to share a project I've begun working on lately. If anybody wanted to give it a try out and see how it works, I'd love to hear any feedback, critique, constructive criticism, ideas, etc.

Thank you, lwvmobile, for Digital Speech Decoder: Florida Man Edition!

I finally sat down with the source code and Fedora 39 Linux. This effort yielded results rather quickly.

Here is DSD FME tracking an EDACS Standard system using an RTL-SDR Blog V4 on Fedora 39:



The software build process was much more straightforward than I anticipated. An overview:

1. Build the ambe_tones branch of mbelib
2. Build itpp-4.3.1 (from Sourceforge, as this package was removed from Fedora long ago)
3. Build the DSD FME audio_work branch. The remaining build dependencies are all in the Fedora 39 repositories.

The screenshot was taken after invoking the following script.

#!/bin/bash

export LD_LIBRARY_PATH=/usr/local/lib64:/usr/local/lib
export PATH=$PATH:/usr/local/bin
export TZ="America/Denver"

dsd-fme -fh -i rtl:0:852.175M:0:0:24:100:3 -T -C las_cruces_edacs.csv -G 391.csv -N 2>/dev/null

Great stuff!

 

[DaveNF2G]

Looks good. Waiting for Windows...

 

[lwvmobile]

I finally sat down with the source code and Fedora 39 Linux. This effort yielded results rather quickly.

Alright, thanks for the report on that, did have a issue on that a year ago hidden under the title of PipeWire Support but I think you may have handled the path dealings a bit more elegantly, I had just decided to either copy or simlink the needed files from /usr/local/lib to /usr/lib. At some point, I may need to look in CMakeLists.txt and see if some tweaks can be make sure it can find the files on various platforms without having to worry about setting paths all the time. Now that I think of it, I think its something to do in mbelib's cmakelists.txt file, or .in file, or something. It'll go on the 'approach cautiously' to-do list, as in, fix it in one place, hope its not broken in four others. I should probably also revisit the install_notes.md file and update it a bit, make sure its clear and concise, and has good information in it.

Other than that, good to see the EDACS decoder getting some love lately, from me and others. Funny how EDACS has been 'dead' for so long, and yet, it still out there in places. BTW, if you need to, you can use the / and * keys to adjust the analog gain, and use the NBV and H keys (CAPS) respectively to toggle on and off various filters, I was working on tweaking those, but decided to just let the user decide which ones they want to use. I've found that on the STM system in Montreal, you really need something to help the analog audio a bit, it can be whisper thin without some shaping, so there are some options now to get either nice and beefy or clean and crisp with combinations of high pass, low pass, and passband shaping.

Looks good. Waiting for Windows...

You do realize there is a cygwin version compiled for windows all zipped up and ready to go, right?

Release DSD-FME V2.1b-final (20230915) + AW (20240405) · lwvmobile/dsd-fme

This is the precompiled Windows version of the base release of v2.1b-final branch from 20230915, and the updated audio_work branch as of 20240405. Extract, and copy and paste the aw patched .exe fi...

github.com

 

[dotignore]

 ██████╗  ██████╗██████╗     ███████╗███╗   ███╗███████╗  ESC to Menu
 ██╔══██╗██╔════╝██╔══██╗    ██╔════╝████╗ ████║██╔════╝  'q' to Quit
 ██║  ██║╚█████╗ ██║  ██║    █████╗  ██╔████╔██║█████╗
 ██║  ██║ ╚═══██╗██║  ██║    ██╔══╝  ██║╚██╔╝██║██╔══╝    MBElib 1.3.4
 ██████╔╝██████╔╝██████╔╝    ██║     ██║ ╚═╝ ██║███████╗  AW
 ╚═════╝ ╚═════╝ ╚═════╝     ╚═╝     ╚═╝     ╚═╝╚══════╝  2024-204-g90206ea
.
.
.
------------------------------------------------------------------------------
--Call History----------------------------------------------------------------
------------------------------------------------------------------------------
for example: Hytera ALG KEY Motorola etc

tell me please
where I can edit the parameters? that interest me so that they drop into history (Call History)

 

[lwvmobile]

where I can edit the parameters? that interest me so that they drop into history (Call History)

Any user changes for that can be made in src/dsd_ncurses.c , barring in mind that you'll be solely responsible for any code changes and knowing how to implement them in there. The whole project is open source, so, you know, you can literally change anything you want to suit your fancy, just keeping in mind that I'm not responsible or required to help you DIY your own changes.

Just look for the area that has the call_matrix and assign the alg id into one of the array positions, and have it in turn print them as well towards the bottom. I'm not going to draw out a detailed code to tell you how to do that, if I did that, I could just literally make the change myself in a fraction of the the time it would take to explain it for you.

Again, its not a user configurable element, its hard coded, so changes will need to be made in the source code to handle it, and I don't see the value personally of logging the call history on ncurses to reflect what alg id a caller used, the log file output is much more in depth when you want to look at all that stuff. The ncurses terminal I try to keep it to just the most important things, and even then, I spill over the 80 column-ish limit I put on it.

 

[DaveNF2G]

You do realize there is a cygwin version compiled for windows all zipped up and ready to go, right?

Release DSD-FME V2.1b-final (20230915) + AW (20240405) · lwvmobile/dsd-fme

This is the precompiled Windows version of the base release of v2.1b-final branch from 20230915, and the updated audio_work branch as of 20240405. Extract, and copy and paste the aw patched .exe fi...

github.com

I meant the enhancement being discussed most recently. Has that been compiled into cygwin also?

 

[lwvmobile]

I meant the enhancement being discussed most recently. Has that been compiled into cygwin also?

The Cygwin release version is up to date as of 2024-04-05, and since then, its just been a ton of background handling of EDACS, and some M17 improvements/features, and if you don't use either on of those, then yeah, its virtually up to date with the current source code.

Only other thing I see being added to the next release outside of the aforementioned things, is adding the log / decode of Color Code from the EMB_PDU on DMR MS. I think there may be a bug when it attempts to collect and decode that on MS and possible on BS during voice (not data bursts), so was going to look into that.

 

[DaveNF2G]

I do have a couple of EDACS sites still active here. No ProVoice, though.

 

[iw2dtt]

good morning, sorry, to configure DSD FME with SDR dongle with Windows system is there something pre-compiled or how should I do it, I'm waiting for instructions, thanks

 

[ki4hyf]

good morning, sorry, to configure DSD FME with SDR dongle with Windows system is there something pre-compiled or how should I do it, I'm waiting for instructions, thanks

Release DSD-FME V2.1b-final (20230915) + AW (20240405) · lwvmobile/dsd-fme

This is the precompiled Windows version of the base release of v2.1b-final branch from 20230915, and the updated audio_work branch as of 20240405. Extract, and copy and paste the aw patched .exe fi...

github.com

 

[TheButcher]

The windows version performs the same as the Linux version. Both good.

 

[RobDLG]

I had just decided to either copy or simlink the needed files from /usr/local/lib to /usr/lib. At some point, I may need to look in CMakeLists.txt and see if some tweaks can be make sure it can find the files on various platforms without having to worry about setting paths all the time.

For the initial Fedora 39 build, I decided to just work with the default install directories and set LD_LIBRARY_PATH accordingly. I subsequently encountered some additional considerations:

1. The itpp package in Sourceforge apparently does not comply with the Fedora Packaging Guidelines. There exist Fedora test builds of this code which are patched to correct the install directories. (Unfortunately, there are only test builds, as the effort to restore itpp to Fedora appears to have been abandoned.)
2. Fedora has a CMake build system that includes helpful macros to automatically set compliant install paths, among other things.

I re-built all the binaries using the itpp patches and Fedora CMake macros, and everything still worked.

Then, of course, Fedora 40 was released on Tuesday, April 23. I approached this release with some trepidation, as I've experienced far too much breakage in Fedora since the "Core 1" release in 2003, so I was relieved to see, once again, things still worked.

Thus, I am able to report that Digital Speech Decoder: Florida Man Edition works on Fedora 40!

Thanks again.

 

[iscottybotty]

Hello I'm trying to decrypt an AES-256 frame but it doesn't work, would you know what the problem is?

Slot 1 DMR PI C- ALG ID: 0x25 KEY ID: 0x06 MI: 0x2F0FC7AC1B772901

 AMBE F3245800CD4480 err = [0] [0]
 AMBE 9F7740673C3F80 err = [0] [0]
 AMBE 58BC7E8E97E880 err = [0] [0]
 AMBE 59F85F120BAE80 err = [0] [0]
 AMBE 766D0A182C5A00 err = [0] [0]
 AMBE 01A9A8B337C400 err = [0] [0]
 AMBE AE2729CABE7700 err = [0] [0]
 AMBE B66BCB8CB68480 err = [0] [0]
 AMBE 8265FD65D19300 err = [0] [0]
 AMBE A409E1DABDE000 err = [0] [0]
 AMBE DDD69C887C7700 err = [0] [0]
 AMBE 8CDEF4254B5080 err = [0] [0]
 AMBE 6B4BE21237E880 err = [0] [0]
 AMBE 5CDC4E0DA19800 err = [0] [0]
 AMBE DB92573D799980 err = [0] [0]
 AMBE 1098DD82F81780 err = [0] [0]
 AMBE 57D6B6A7B98B80 err = [0] [0]
 AMBE CF2C02F085BB00 err = [0] [0]

AES256
plaintext to encrypt:
2F0FC7AC1B77290104AF51AF97E13FF6
key:
0000000000000000000000000000000000000000000000000000000000000001
ciphertext:
FEE54EDF09363570DE3F489ABC7CC080

AMBE F3245800CD4480 err = [0] [0]

F3 XOR FE = 0D but should be F8

Question: If the MI is the plaintext, is the Cipher Text F8 01 A9 9F 8C E0? Or if not, where is the cipher text on a log?

 

[iscottybotty]

Using the Windows compiled version, I click on the following to open:

1a.
1b.
3 - tail.

Why is the tail screen a black screen showing the command only, why isn’t it loading into the tail window? Do I have to clear something in its memory? Was working fine before, now the tail window has gone messy.

Anyone.

 

[lwvmobile]

Why is the tail screen a black screen showing the command only, why isn’t it loading into the tail window? Do I have to clear something in its memory? Was working fine before, now the tail window has gone messy.

Okay, just put log.ans for tail, like tail -f log.ans and also make sure that that is the file that the startup bat file for dsd-fme is directing its output into.

 

[TeddJohnson]

I tried checking the Example_Usage.md but maybe I missed it:

Can I get DSD-FME to give me regular audio output if it's not decoded (specifically looking to pickup analog) on Windows 11?

I use SDR++ to pipe audio over, and sometimes I jump to a ham repeater but can't hear anything.
SDR++ has an upcoming dual-sink option but I don't think there's a windows build on that branch yet, so this was my solution.