DSD FME

[mtindor]

OK gentlemen, I tried everything that you said and I’m coming up short still has anybody that can take a screenshot of your SDR plus plus to make sure I everything set right
Sorry to be a bother, but I appreciate everyone helping me

If SDR++ is anything like SDR#, then you need to make sure you turn OFF "filter audio" or you are going to have difficulties decoding any sort of digital. That's all that I can recommend. Of course, you also need to make sure the audio out of SDR++ is not too hot or too low.

 

[SignalPatcher]

Hi. I think I've read the whole thread, but I haven't found such an error.

╚═════╝ ╚═════╝ ╚═════╝    ╚═╝ ╚═╝ ╚═╝╚══════╝
Build Version: AW 2025-214-g632041a (CYGWIN)
MBElib Version: 1.3.2
Enabling NCurses Terminal.
OSS Input /dev/dsp.
OSS Output /dev/dsp.
Audio In/Out Device: /dev/dsp
Error opening terminal: xterm-256color.

I am using Windows 11 and compiled it this way:

wget -O itpp-latest.tar.bz2 http://sourceforge.net/projects/itpp/files/latest/download?source=files
tar xjf itpp*
#if you can't cd into this folder, double check folder name first
cd itpp-4.3.1
mkdir build
cd build
cmake ..
make
make install
cd ..
cd ..

git clone GitHub - lwvmobile/mbelib: Full Rate and Half Rate Radio Vocoder
cd mbelib
mkdir build
cd build
cmake ..
make
make install
cd ..
cd ..

git clone GitHub - lwvmobile/dsd-fme: Digital Speech Decoder - Florida Man Edition
cd dsd-fme
git checkout audio_work
mkdir build
cd build
cmake ..
make -j `nproc`
make install

When I run it through the Cygwin terminal, it works.

Am I missing any components that would allow me to use the decoder without the Cygwin terminal? Should I compile it differently if I want to use Windows?

 

[gary123]

Stupid question time has anyone been able to build the windows binaries using Visual Studio. I have 2022 installed and am working on building the current git.

 

[Forts]

Take a look at DSD-Neo (It's a fork from DSD-FME). He's releasing current pre-compiled Windows builds. I've only played around with it a little bit so far, but it's working well for me.

 

[gary123]

That is on my list of DSD variants to play with. I was planning on putting the current FME build into my files and off the bucket list.

 

[MELERIX]

a new compiled build was released: Release 2025 · lwvmobile/dsd-fme

 

[lwvmobile]

Here is the link for that, the link mentioned above is pointing at a release tag that I ended up changing to reflect the current state of the code.

Releases · lwvmobile/dsd-fme

Digital Speech Decoder - Florida Man Edition. Contribute to lwvmobile/dsd-fme development by creating an account on GitHub.

github.com

 

[scanfan03]

Here is the link for that, the link mentioned above is pointing at a release tag that I ended up changing to reflect the current state of the code.

Releases · lwvmobile/dsd-fme

Digital Speech Decoder - Florida Man Edition. Contribute to lwvmobile/dsd-fme development by creating an account on GitHub.

github.com

So with this new version we don't run dsd-fme-aw-lz or any of the other variants anymore? Only dsd-fme?

 

[lwvmobile]

So with this new version we don't run dsd-fme-aw-lz or any of the other variants anymore? Only dsd-fme?

Its just the base dsd-fme version. I didn't build with the few variants that still remain, mostly just had forgotten about those. Other than the LZ variant, not even sure which ones are still out there that haven't been fully implemented or handled in code at this point. If you really need any of those, you can always still build it yourself using the instructions, if needed. I may eventually get back around and build a few of those and put them in a separate little zip file or something.

 

[scanfan03]

Its just the base dsd-fme version. I didn't build with the few variants that still remain, mostly just had forgotten about those. Other than the LZ variant, not even sure which ones are still out there that haven't been fully implemented or handled in code at this point. If you really need any of those, you can always still build it yourself using the instructions, if needed. I may eventually get back around and build a few of those and put them in a separate little zip file or something.

Ok all good. I was just wondering because (probably a few versions ago) you had said to run the aw-lz version for best results with NXDN trunking.

Thank you!

 

[gary123]

Here is the link for that, the link mentioned above is pointing at a release tag that I ended up changing to reflect the current state of the code.

Releases · lwvmobile/dsd-fme

Digital Speech Decoder - Florida Man Edition. Contribute to lwvmobile/dsd-fme development by creating an account on GitHub.

github.com

Thank you I saw that link and added the update to my FME folder.

 

[lwvmobile]

Ok all good. I was just wondering because (probably a few versions ago) you had said to run the aw-lz version for best results with NXDN trunking.

For pure trunking, I don't think LZ will make a difference, or if it did in the past, its been integrated into code. The significant difference of the LZ code was to load scrambler keys by frequency tuned to, instead of using a key id value or target value, I guess its common to have multiple conventional NXDN systems with a single key same id and same tg but all use a different key value and only had the frequency to seperate them.

 

[scanfan03]

I need to confirm this, but it seems like the latest version won't pass through un-encrypted NXDN if you have a spreadsheet set by key id. It looks like it actually tries to decrypt comms in the clear. Do I need to take the -4 away (I think that was the switch that forced encryption right?)? On the older aero version it just worked normally even with the -4 switch activated.

 

[lwvmobile]

I need to confirm this, but it seems like the latest version won't pass through un-encrypted NXDN if you have a spreadsheet set by key id. It looks like it actually tries to decrypt comms in the clear. Do I need to take the -4 away (I think that was the switch that forced encryption right?)? On the older aero version it just worked normally even with the -4 switch activated.

Yeah, you'll have to get rid of the -4 option now, can't remember the behavior that far back in aero, but now, its always making the attempt on scrambler.

 

[woodpecker]

Does the CRC ERR in this capture mean the MI(32) is not correct? Every capture seems to show CRC ERR

Slot 1 DMR PI H- ALG ID: 21; KEY ID: 08; MI(32): 7F500A04; DMRA RC4; (CRC ERR)

 

[lwvmobile]

Does the CRC ERR in this capture mean the MI(32) is not correct? Every capture seems to show CRC ERR

If every header / data block shows this, then its probably RAS enabled.

 

[SignalBrus]

I'm having issues to get DSD-FME working on configurations based on DragonOS Noble R7. DragonOS is installed on a physical PC, hence no virtualized environment or so. SDR++ is setup with TCP sink on port 7355, tuned to a DMR-channel with excellent quality. DMR-FME is started with command string 'dsd-fme -fs -i tcp:127.0.0.1:7355 -N -Z 2>log.txt' In SDR++ TCP sink changes from Listening to Connected and in DSD-FME it looks as it correctly finds the TCP sink connection. Then absolutely nothing happens, there are bursts and traffic on channel but nothing is decoded in DSD-FME, not even a sign of DMR decode or CC presented. I also have a GRC flow which is known to work on other systems to feed DSD-FME over TCP, but that is not either working on DragonOS Noble R7. DSD-FME comes pre-installed and I have also tried to upgrade to latest version of DSD-FME but the problem is persisting. Anyone how have any ideas on what is causing the issue?

DragonOS build for Raspberry Pi comes without DSD-FME but there it is just to install the software and it works flawlessly. Same thing with a standard Ubuntu system, just install SDR++ and DSD-FME and it works great.

 

[lwvmobile]

I'm having issues to get DSD-FME working on configurations based on DragonOS Noble R7. DragonOS is installed on a physical PC, hence no virtualized environment or so. SDR++ is setup with TCP sink on port 7355, tuned to a DMR-channel with excellent quality. DMR-FME is started with command string 'dsd-fme -fs -i tcp:127.0.0.1:7355 -N -Z 2>log.txt' In SDR++ TCP sink changes from Listening to Connected and in DSD-FME it looks as it correctly finds the TCP sink connection. Then absolutely nothing happens, there are bursts and traffic on channel but nothing is decoded in DSD-FME, not even a sign of DMR decode or CC presented. I also have a GRC flow which is known to work on other systems to feed DSD-FME over TCP, but that is not either working on DragonOS Noble R7. DSD-FME comes pre-installed and I have also tried to upgrade to latest version of DSD-FME but the problem is persisting. Anyone how have any ideas on what is causing the issue?

DragonOS build for Raspberry Pi comes without DSD-FME but there it is just to install the software and it works flawlessly. Same thing with a standard Ubuntu system, just install SDR++ and DSD-FME and it works great.

Screenshots, logs, anything?

 

[MELERIX]

If every header / data block shows this, then its probably RAS enabled.

is there a way to make DSD-FME to show info in the screen when a DMR signal is using RAS?

 

[lwvmobile]

is there a way to make DSD-FME to show info in the screen when a DMR signal is using RAS?

You can use the -F option to bypass the CRC error, but I wouldn't recommend it unless its a trunking system and all the CSBKs are RAS enabled. The CRC is already bypassed for the PI header and relevant VLC/TLC anyways.