Issue with recording P25 Phase 2 traffic

[ghost]

Hello,

I have been using a Raspberry Pi 3 B+ with trunk-recorder installed and it’s doing a great job on decoding and recording P25 Phase 1 traffic, but when the network recently started carrying Phase 2 traffic, all it records were garble and unintelligible traffic. I see the logs reflect it recognizes TDMA traffic but it just would not record Phase 2 intelligibly.

Anyone has experienced this as well?

My raspberry pi is currently running on the following
Linux version 4.14.79-v7+ (dc4@dc4-XPS13-9333) (gcc version 4.9.3 (crosstool-NG crosstool-ng-1.22.0-88-g8460611)) #1159 SMP Sun Nov 4 17:50:20 GMT 2018

Here is a sample of the trunk-recorder log that I captured:
[2020-02-07 21:57:46.454713] (info) [METRO25] TG: 9032 Freq: 8.530375e+08 TG not in Talkgroup File
[2020-02-07 21:57:46.455067] (info) P25 Recorder - Adjust Channel Resampled Rate: 30117.6 System Rate: 30000 ARB Rate: 0.996094
[2020-02-07 21:57:46.455979] (info) - Starting P25 Recorder Num [0] TG: 9032 Freq: 8.530375e+08 TDMA: true Slot: 0
[2020-02-07 21:57:46.457004] (info) [METRO25] TG: 9032 Freq: 8.530375e+08 Starting Recorder on Src:
[2020-02-07 21:57:52.001970] (info) [METRO25] TG: 9032 Freq: 8.530375e+08 Ending Recorded Call - Last Update: 4s Call Elapsed: 6
[2020-02-07 21:57:52.015310] (info) - Stopping P25 Recorder Num [0] TG: 9032 Freq: 8.530375e+08 TDMA: true Slot: 0

[robotastic]

It is definitely possible that the math gets mixed up on ARM chips. The only real difference between OP25 and Trunk Recorder for Phase 2, are these files: https://github.com/robotastic/trunk-recorder/tree/master/lib/lfsr
I will try to look through and see if I spot any obvious things where it is assuming a certain order for bits.

[GTR8000]

I also think that Phase II decoding is hosed up on the ARM platform.

We tried running on a Pi 3 B+ and could not get any intelligible audio from a TDMA system. Switched over to an ACEPC AK2 using the exact same RTL dongle and config.json, and decoding was excellent.

[ghost]

GTR8000 what distro did you use on the ACEPC AK2 may I ask?
I had some luck setting up Ubuntu 18.04 using parallel desktop on my mac and got trunk-recorder working, but currently having problem with it using airspy dongle, so more experimentation on my side is needed..

[GTR8000]

Ubuntu 18.04.3 LTS, Trunk Recorder v3.1.2, with fdk-aac and sox also installed to support the OpenMHz change from ffmpeg encoding.

[edinburger]

Just started a new setup with RasPi4 (4GB) and am experiencing the same issues with TDMA recordings. Garbled audio sounds like it is encrypted. No issues with Phase-I TGs though.

[Cellphone70]

I am seeing the same issue with P25 Phase 2 talkgroups being garbled. No issue with Phase 1 Talkgroups. I'm running the latest Docker version of Trunk-Recorder on Ubuntu 18.04 LTS. Hardware is an Intel Skylake i7-6820HQ @ 2.70GHz.

[Cellphone70]

Here is an example of the audio that is recorded. All talkgroups on this system are P25 Phase 2.
https://openmhz.com/system/mcchand

[ghost]

I can confirm that when I use my Macbook Pro (much more horse power than Rp3 B+ obviously), I have no issue decoding Phase 2 traffic.

[mikewren]

Not sure if it's related, but I experienced the garbled audio issues on a P25 phase II system on talkgroups encoded as TDMA using a TR version from March. Recompiled on 6/13 and it's now working perfectly - thanks!

Edit: Fixed typo

[Cellphone70]

I was using the docker version experiencing the issue. I abandoned that, and compiled from the ditros about 2 weeks ago, and Phase 2 has been working for me since.

[jonk6]

Recompiled on 6/13 and it's not working perfectly - thanks!

@mikewren - not? Or now?

[mikewren]

Recompiled on 6/13 and it's not working perfectly - thanks!

@mikewren - not? Or now?

Sorry, typo, it's NOW working perfect - https://openmhz.com/system/monroecony