DSD FME

[ki4hyf]

If the "-Z" switch doesn't get you what you want, I'm not sure what will. Just out of curiosity, what is missing from the voice superframe?

 

[intraHJJ2345sd32]

If the "-Z" switch doesn't get you what you want, I'm not sure what will. Just out of curiosity, what is missing from the voice superframe?

I compared the DSD+ Fastlane and I notice it spits out binary, I haven't fully grasped each separate part yet but for example I read in this one presentation it says:
Every voice transmission:
• begins with signalling burst(s) (Voice LC Header)
• is followed by so many voice superframes as needed to carry speech
• ends with terminator signalling burst(s) (Voice LC Terminator)

And I see it's protected by it says bptc which after googling I learn is block product turbo code and a type of forward error correction, so that's then abbreviated fec? And then there's checksum within the header itself? Why fec + checksum? Then also it says LC(72) + Checksum(5) but down there it says in ETSI BPTC(196,96); I got lost here, if 72+5 = 77 bits, how does 96 play into there? I figured 196 would be value with fec somehow, but 96-77 = 19, where are the missing 19 bits? Or have I missed the point completely? No idea.

What my goal was is to try and understand and imagine in my hear if I was speaking dmr what would I be saying, how would I start the conversation etc, how they sync between two tdma slots, its very fun. I never went this deep into computer science before, but I must say I like it! There's way more math than I expected.

73

 

[ki4hyf]

I compared the DSD+ Fastlane and I notice it spits out binary

Interesting. How do you get it to "spit out" binary?

 

[gary123]

It would be nice to have Fastlane also output the same frame data as FME does. Thanks to the members here I am now using FME to capture and log the frame data.

 

[ki4hyf]

I could be very wrong, but it's my impression that's not going to happen. But you're right, it would be nice. AFAIK, only DSD-FME, OP25 and the original DSD show (at least some) voice frame information.

 

[DaveNF2G]

I don't see an option in the help for decoding DMR only. I am getting a lot of gobbledygook and false decodes using the -fa switch.

This run doesn't show the DStar and other junk that was showing up in an earlier run. Bandicam got a little crazy on Windows 11.

BTW, what does YSF mean? I found a conversation about it but no definition.

 

[Whiskey3JMC]

BTW, what does YSF mean? I found a conversation about it but no definition.

Yaesu System Fusion

 

[Forts]

Try using the -ft option

Decoder options:
-fa Auto Detection
-fA Passive Analog Audio Monitor
-ft TDMA Trunking P25p1 Control and Voice, P25p2 Trunked Channels, and DMR
-fs DMR TDMA BS and MS Simplex
-f1 Decode only P25 Phase 1
-f2 Decode only P25 Phase 2 (6000 sps) **
-fd Decode only DSTAR
-fx Decode only X2-TDMA
-fy Decode only YSF
-fz Decode only M17*
-fU Decode only M17 UDP/IP Frame***
-fi Decode only NXDN48* (6.25 kHz) / IDAS*
-fn Decode only NXDN96* (12.5 kHz)
-fp Decode only ProVoice*
-fh Decode only EDACS Standard/ProVoice*
-fH Decode only EDACS Standard/ProVoice with ESK 0xA0*
-fh344 Decode only EDACS Standard/ProVoice and set AFS to 344 or similar custom 11-bit scheme*
-fH434 Decode only EDACS Standard/ProVoice and set AFS to custom 11-bit scheme with ESK 0xA0*
-fe Decode only EDACS EA/ProVoice*
-fE Decode only EDACS EA/ProVoice with ESK 0xA0*
-fm Decode only dPMR*
-l Disable DMR, dPMR, NXDN, M17 input filtering
-u <num> Unvoiced speech quality (default=3)
-xx Expect non-inverted X2-TDMA signal
-xr Expect inverted DMR signal
-xd Expect inverted ICOM dPMR signal
-xz Expect inverted M17 signal

 

[ki4hyf]

I don't see an option in the help for decoding DMR only. I am getting a lot of gobbledygook and false decodes using the -fa switch.

The DMR only switches are -fs (DMR Trunking +) and -fr (DMR Simplex). The -fr is no longer listed in the help, but it still works for me. The -fa switch is the default and is redundant on the command line.

 

[DaveNF2G]

I already get false indications for P25P2 when monitoring DMR signals. -ft should cut out the Fusion and DStar junk.
I am monitoring a conventional DMR repeater, so I had originally ruled out -fs as an option. Was I wrong?

Also, when running in the Console window, is there a quit command or do I just have to close the window on it? I've tried Esc, q and c, but it appears there are no active keys in this mode.

 

[DaveNF2G]

After a few hours of monitoring, I observe that the idle bursts come in sets of 4, and that the first block in every set has an error.

 

[Ubbe]

I don't see an option in the help for decoding DMR only. I am getting a lot of gobbledygook and false decodes using the -fa switch.

-fs are DMR decoding. You can push the Esc key on the dsd-fme window and change between decoding formats.
DMR use 12,5KHz wide channels so your 8KHz setting are too narrow. I haven't seen any SDR-RTL dongle having zero frequency error.
My V2 have a 2KHz error and my V3 5KHz. What you do are using SDR# and zoom in on a steady digital frequency and you can left click on the edge of its bandwidth markings in the spectrum view and drag it to exactly cover the signal. If your dongle have the wrong ppm set it will be unsymetrical in that bandwidth window, so go to the cogwheel settings and adjust ppm until the signal fits perfectly. Then use that ppm in DSD-FME.

/Ubbe

 

[ki4hyf]

I already get false indications for P25P2 when monitoring DMR signals. -ft should cut out the Fusion and DStar junk.
I am monitoring a conventional DMR repeater, so I had originally ruled out -fs as an option. Was I wrong?

Also, when running in the Console window, is there a quit command or do I just have to close the window on it? I've tried Esc, q and c, but it appears there are no active keys in this mode.

You should use -fs. I'm not familiar with the windows version, but Ctrl+c should get you back to a prompt.

 

[ki4hyf]

After a few hours of monitoring, I observe that the idle bursts come in sets of 4, and that the first block in every set has an error.

That sounds like it's not getting a clean signal (or the PPM is off like @Ubbe pointed out) or maybe RAS. I don't have any systems using RAS, so I'm not sure. Can you post a screenshot of the beginning of a voice transmission?

A good signal without RAS should look like:

 

[DaveNF2G]

Yeah, I've been using DSDPlus for quite a while, so I'm familiar with basic setup.
ki4hyf - thanks for the info on the -fs switch.

Here's an observation about the NCurses console display.
The Audio Decode segment seems not to repaint correctly. Also, there is no screen width that allows the terminal portion of the display to align properly.

 

[ki4hyf]

I can't really tell anything from either picture. Could you please post the whole terminal window?

 

[lwvmobile]

Here's an observation about the NCurses console display.

I've noticed the same on Windows, usually if you resize the window to taller and wider dimensions, it'll usually fix that. At one point, I did have a command in the .bat files to fix the window size, but I figured out on newer Windows, it doesn't change the window size, but instead just shifts the ncurses terminal in the window to an incorrect location.

I think its just that Windows PowerShell and Command Prompt are 100% compatible with the ncurses implementation. I've also noticed, if you build your own and run it in the default Cygwin terminal (MinTTY) it works perfectly fine.

 

[sh7411usa]

Hi @lwvmobile and thank you for the wonderful product! Observed an interesting issue using the lates windows release with the update in a side by side comparison with DSD+ FL. When monitoring a single P25 Phase 1 channel (conventional/non-trunked) which reports motorola location sentences, that DSD-FME gets the latitude right but the longitude is way off, while DSD+ gets them both right. Interesting that the latitude is correct while longitude is not decoded properly!

00:00:00 Sync: +P25p1 nac: [ ABC] src: [ 100] tg: [ 1] LDU1
MFID90 (Moto) GPS: Lat: xx.xxxxx°N Lon: 105.77612°W (xx.xxxxx, -105.77612) Current;

the correct latitude in this case should start with a -7x.xxxxx
I can pm you a wav recording if you can tell me how to make one. Thanks again for this amazing software!

 

[lwvmobile]

I can pm you a wav recording if you can tell me how to make one. Thanks again for this amazing software!

Yeah, if you want to send it, I'll take a look at it. I almost want to say this issue has already been fixed a while back, depending on which version you are running, you might have a version that is older than the fix. Judging from the print there, I'd say that is possible, the way it prints P25 data also changed.

 

[sh7411usa]

Yeah, if you want to send it, I'll take a look at it. I almost want to say this issue has already been fixed a while back, depending on which version you are running, you might have a version that is older than the fix. Judging from the print there, I'd say that is possible, the way it prints P25 data also changed.

Comparison files and audio pm-ed. DSD-FME version: AW 2025-21-g42cb14b (CYGWIN).