UniTrunker and DSD+ Decoding Question

[juanbanzai]

Hello.

I've poured over similar threads where other posts have asked for help with DSD+ decoding the digital signals coming from UniTrunker. I've got a somewhat different problem that I'm hoping someone can shed some light on.

I'm using the UniTrunker/Virtual Cable/DSD+ setup to listen to the P25 signals in my county. I use SDR# in order to determine the frequency correction necessary for UniTrunker to lock onto the control (Signal) and Voice channels. Due to the drift on my two RTL2832 dongles, I leave both running for about an hour before I attempt to cut UniTrunker loose. Based on the numerous instructions I've found on the web, in addition to several posts I've found on RadioReference, I've got the settings for both roles in UniTrunker set pretty much as what's been recommended.

UniTrunker is locking onto all of the channels within my county system just fine - the Signal role is grabbing all of the data coming from the control channel... and the Voice role is parked at an average frequency within our P25 channel group and has no difficulty whatsoever switching to whatever talkback frequency is being used. Likewise, the digital data is being fed to DSD+ without any issues either (or so it seems). I'm seeing a lot of data being decoded in the main window, and the waveform window in DSD+ looks like all of the other screen shots I've seen of the system working correctly - nice and clean waveforms. Occassionaly, I get DSD+ popping up a message like "Unrecoverable NID"... but I'm thinking THAT'S happening because I'm currently using the trial version of Virtual Audio Cable at the moment, and that "trial" voice that comes up every 10 seconds in VAC is messing with the data feed. I didn't want to invest any more money than I had to for this experiment until I was relatively sure it was going to work on my laptop.

So with all this "perfection" going on, I'm wondering why I'm not able to hear a clean and intelligible audio feed coming from DSD+. For the most part, the voices I'm hearing are robotic sounding and pretty much garbled. Every once in a while I can get a pretty good stream of audio coming out, but mostly it's just digital noise. It's kind of hard to describe this so searching for a solution to the problem on RadioReference or the web has me at a loss. I can't say that UniTrunker and DSD+ aren't decoding because clearly they are. It almost sounds like something isn't locking onto a signal correctly or there's not enough audio going to DSD+ in order for it to decode. But the data window fills up with raw data coming in from UniTrunker. It's just not coming out as clean audio. Or sometimes large amounts of data will fill the window... but NO audio comes out.

I hope I've described this correctly. Thanks very much for any help the group can provide.

Aaron

 

[SCPD]

VB Cable is an alternative that does not have nag audio.

For second RTL stick with the voice role VCO, try this (also works on the signal role RTL stick):

1. adjust the squelch so you hear noise when there's a signal and quiet where there isn't.
2. turn on Drift Correct.
3. for single VCO operation - use the 0.960 msps sample rate.

The correction value you've set should be close enough for the drift correct feature to take over. It will continuously adjust the correction value - so long as there's an active signal in the first VCO.

 

[juanbanzai]

Thanks for the reply. Tried following your tips, but I don't think I'm getting the results I should. By default, the squelch setting on the voice role was set to 0. Not sure if I'm doing this right, but I changed the Audio Output setting for the voice role to "speakers" so I could hear what was coming from UniTrunker. Heard noise right away. I tried changing the squelch value from 50 to 20,000 and wasn't able to squelch the noise coming from the speakers. I stopped and started the dongle each time I made a change but with no results.

Also, where would I make the sample rate change to .960 msps? Is that done in UniTrunker or Virtual Audio Cable?

I also downloaded and installed VB Cable. Much better so far. No nag.

Thanks again for the help.

Aaron

 

[bama9999]

You change the sample rate on the Unitrunker receiver... click the "Info" tab and you should see it there.

 

[SCPD]

Hi Aaron;

I tried changing the squelch value from 50 to 20,000 and wasn't able to squelch the noise coming from the speakers.

Check the Deemphasis box. Turning off deemphasis also turns off squelch.

 

[juanbanzai]

Thanks so much for pointing out the function of Deemphasis. I tried clicking that several times throughout my setup process and never noticed a difference.

I was able to set the squelch on the Voice receiver so that nothing but signal comes through now. I honestly have to say that the decoding process to the point where I'm hearing a clear voice is better... but it's only perhaps 25% better. The voices are slightly more intelligible... but it's still more garbled digital noise than plain speech.

I would assume that there's no need to check Deemphasis on the Signal receiver, correct?

Also, I've poured over my settings in the Info tab for the Voice receiver and I'm honestly not seeing anywhere I can make a sampling rate change. Attached in a screen-shot of the Voice receiver. I hope I'm not misinterpreting the settings for the obvious. I've also been playing with this for several days so I also hope that something isn't set correctly that should be.

Many thanks again for the continued assistance.

Aaron

 

[SCPD]

Aaron - you're running a slightly older version which limits you to one VCO per RTL stick.

There's a Drift Correct bug in an older version (I do not recall which). Since we can't see RSSI or Window value in the version you have - best to leave it off.

 

[juanbanzai]

Thanks for pointing out the older version. I went to the UniTrunker site this morning and found that a newer version was posted on 8/5. Downloaded and installed. Went through all the settings and am pretty much back to where things were with the older version. I'm seeing the VCO information now with the sampling rate. Both receivers are set for .96 msps. In DSD+, I'm also seeing slightly cleaner data being decoded. After a solid half hour of monitoring, I'm no longer seeing any of the occasional decoding errors I was seeing before - such as the "unrecoverable NID" bit that would pop up from time to time. However, the audio coming from DSD+ is still mostly garbled digital noise. About 50% of the time I can make out actual voices, but it's still mostly unintelligible. Using SDR#, I've made several checks to ensure that the RTL dongles are spot on frequency. In fact, I left the laptop on overnight so both dongles are good and warm and quite accurate. My receiving and decoding of the Control and Talkback channels is right on.

On the Voice receiver, I can 1 VCO selected while in the Signal receiver there are 2. Should those match? I've also seen some posts on RadioReference where the Baseband AGC should be checked with a Gain set between 300 and 450. I've got mine set at 350 in the Voice receiver. Changing any of the Gain settings seems to have no effect.

Thanks again for the continued help.

Aaron

 

[juanbanzai]

Hello...
Sorry to bump the thread, but is there any more insight into trying to resolve the decoding issue I'm having with UniTrunker and DSD+. Once again... it's not really a "decoding" problem since both programs are having no difficulties in locking onto the control and talkback frequencies for the P25 system in my county. The issue is that DSD+ is not putting out clean decoded audio. It's clearly decoding the incoming digital stream, but nothing but garbled audio is coming out. Only 50% of the transmission are understandable... the rest is garbage.

One thing I haven't considered is whether or not there are any settings in my laptop that could be hurting the final conversion process.

Thanks very much for any assistance that can be provided.

Aaron

 

[SCPD]

Run task manager, select the Performance tab. Make sure none of the CPU cores is close to 100%.

 

[DSDPlus]

You need to record the raw audio in DSD+ and post a link to it here.

 

[juanbanzai]

Will do. I'll attend to that shortly. And I did indeed check the core functions of the laptop earlier thinking that perhaps some process was maxed out. Turns out that when UniTrunker and DSD+ are running, only 10-15% of the CPU is being used. Even if McAfee were to kick in while the programs are running, the useage only jumps to 35%. It's a very beefy laptop... designed for intense video and audio editing.

Aaron

 

[DSDPlus]

McAfee??

https://www.youtube.com/watch?v=bKgf5PaBzyg

 

[juanbanzai]

Here's a link to some recent P25 activity as heard on my UniTrunker/DSD+ system.
http://www.buffalohistoryworks.com/p25/p25_system.mp3
Throughout all of this digital noise, DSD+ has been decoding everything being sent over from UniTrunker with no errors being shown. Also, nothing is coming up in DSD+ as being encrypted.

And sheepishly I add that McAfee was shut down and out of the system completely when I was making the test. I hate the software to death... but it HAS caught some nasty things from getting on my machine. But that's another story not for this thread.

Thanks again for the help.

Aaron

 

[n5pwp]

Aaron
I was having a similar issue with my audio. I had upgraded the dsdplus program to 1.074. I had used the program dsdtune before to improve my audio decoding when I first started with UniTrunker and dsdplus. It did a very good job. When I tried again I started getting an error message and not completing. I found out there was an issue with the 1.074 dsdplus and the older dsdtune program. I found there was a newer version of dstune (v0.1.1) that came out recently. I used it and now my audio decodes are very usable.
Good luck.

 

[KC5KAW]

I was just playing around with this and had a similar problem. The biggest improvement for me came when I checked my settings in VB Cable. The sample rate was far too low. I upped the input and output sample rates to 48,000 Hz. I also made the output mono. I'm still tweaking, but close to 90% of my audio is intelligible now. To get to this, go to your sound controls in Windows and double click the VB Cable lines. The sample rate is in a drop down under the Advanced tab.

I realize the thread is a little stale, but I hope this helps somebody.