DSDPlus

[KC1UA]

Gary, did you check to make sure that the Filter Audio box in SDR# is NOT checked off?

I have this working, sort of. When I'm on a DMR channel the display scrolls and I see data displaying. From reading this thread I'm seeing indications of Slot 1 and occasionally Slot 2. I'm getting Color Codes, mostly 7. I'm getting BS, which indicates I'm getting data. and I'm getting an Idle message. Once in a wile I get voice fragments and a snippet of voice. I'm just not getting consistent voice.

I parked on my towns DPW VHF frequency and at first I got all of the above. Now I get nothing although SDR# shows activity on that frequency. I do hear the data sound and the "Trial" marker from VAC. Just no audio.

I think I'm close, but not quite there yet.

I'd appreciate any thoughts anyone might have.

I'm running DSD with no additional parameters right now. I was specifying the audio input and no recording, but I stopped doing that.

 

[MtnBiker2005]

Quick question...

Does this new program work with PSR-800 IF Out audio?

 

[cobraDIHO]

Another huge improvement is the reduced CPU load. I can decode multiple formats on my Intel Atom netbook flawlessly.

 

[garys]

Gary, did you check to make sure that the Filter Audio box in SDR# is NOT checked off?

Yep. I read every idea in this thread. The funny thing is it did decode the display the data at one point. I don't know if the problem with with VAC or maybe the repeater or the set up. I can hear the digital data stream on a scanner when they transmit, but DSD+ appears deaf to it.

I just moved SDR# to the 900 Mhz band and both DMRDecode and DSD+ are scrolling data. Maybe the town is on a format that DSDPlus won't process.

 

[jje64]

I am having similar issues with DMR decode. Placed the radio on a UHF DMR freq. First tried the DSDPlus program with -fa switch. I get scrolling data with colorcodes, etc. No audio decode.

Then tried the -fr switch and the -D1 switch. In both cases I get an error that reads "invalid switch" And It shows DSDPlus 1.0.3. But when running P25 all switches work and it reads DSDPlus 1.7.0

Any updates on what folks have found with DMR?

Thanks
jje64

 

[br0adband]

I have this working, sort of. When I'm on a DMR channel the display scrolls and I see data displaying. From reading this thread I'm seeing indications of Slot 1 and occasionally Slot 2. I'm getting Color Codes, mostly 7. I'm getting BS, which indicates I'm getting data. and I'm getting an Idle message. Once in a wile I get voice fragments and a snippet of voice. I'm just not getting consistent voice.

I parked on my towns DPW VHF frequency and at first I got all of the above. Now I get nothing although SDR# shows activity on that frequency. I do hear the data sound and the "Trial" marker from VAC. Just no audio.

I think I'm close, but not quite there yet.

I'd appreciate any thoughts anyone might have.

I'm running DSD with no additional parameters right now. I was specifying the audio input and no recording, but I stopped doing that.

If you're sick of hearing that "Trial" thing I highly recommend the VB Audio Cable (free competitor of Virtual Audio Cable). I set it up one time, set the rate to 16 bit 48000 Hz in the Windows mixer (both Playback and Recording for that "device") and I haven't had one lick of trouble since. Virtual Audio Cable, on the other hand, good lord that thing just irritated me to no end, so YMMV. I am using VB-Audio Virtual Cable (two of them, CABLE and CABLE-A) and again, haven't had a single issue so far.

As for your decoding issues, as long as DSDPlus is getting "something" then it's pretty much working. I would highly suggest using DMRDecode at the same time to monitor what's being sent across the frequency. There are several MOTOTRBO frequencies that I've been noting (the list is growing fast, it seems) here in the Las Vegas area that are nothing but data constantly, and while DSDPlus will show you DATA during the decode, DMRDecode will actually show you what's going on with respect to the data as in whether or not there's a data call between two units (which I'm guessing is the "Private" aka encrypted mode of MOTOTRBO (and DMR technology) as well as when data is being sent to specific units ("Data call from ...").

So far this thing is working fantastic and I haven't (and probably won't) get into the whole tuning thing but that's just me. Here's a shot of what I mean with respect to DSDPlus showing both data and voice at the same time and DMRDecode verifying it.

 

[br0adband]

Quick question...

Does this new program work with PSR-800 IF Out audio?

If that's how the PSR-800 does the discriminator tap then yep, it should work fine.

 

[br0adband]

I am having similar issues with DMR decode. Placed the radio on a UHF DMR freq. First tried the DSDPlus program with -fa switch. I get scrolling data with colorcodes, etc. No audio decode.

Then tried the -fr switch and the -D1 switch. In both cases I get an error that reads "invalid switch" And It shows DSDPlus 1.0.3. But when running P25 all switches work and it reads DSDPlus 1.7.0

Any updates on what folks have found with DMR?

Thanks
jje64

Here's what I would suggest for you and anyone else just getting started with DSDPlus (and I just started myself on Christmas night so, I'm no more of an expert than anyone else I guess):

Use DSDPlus for a while with no command line switches at all - just run DSDPlus.exe all by itself, let it get to work by fault and then see what happens. I think sometimes that people are trying to customize the command line switches too drastically at first and perhaps they're doing some self-defeating in the process.

Use the defaults (let it auto-detect auto-optimize auto-everything) by just running the .exe by itself and see what happens. When you start actually getting content then you can go and do some more specific decoding (for specific modes), etc.

Can't hurt, right? Even if you know what you're trying to decode is P25 or MOTOTRBO or whatever, it doesn't hurt to let DSDPlus do it first for you and then you can see what it's found, then go back and tell it to listen for specific modes and even more fine tuning.

Makes sense, at least to me, hopefully to others as well.

 

[garys]

If you're sick of hearing that "Trial" thing I highly recommend the VB Audio Cable (free competitor of Virtual Audio Cable). I set it up one time, set the rate to 16 bit 48000 Hz in the Windows mixer (both Playback and Recording for that "device") and I haven't had one lick of trouble since. Virtual Audio Cable, on the other hand, good lord that thing just irritated me to no end, so YMMV. I am using VB-Audio Virtual Cable (two of them, CABLE and CABLE-A) and again, haven't had a single issue so far.

I think I'm going to try this as VAC seems to be a bit fussy. Maybe that's part of my problem.

I've been running DRMDecode along with DSDPlus to see what's going on. Once in a while DSDPlus will scroll data, but DMRDecode won't process the signal. Oddly, I have to run Unitrunker to get the audio to work at all. I'm not sure what I screw up there, but obviously something.

Off to try VB.

 

[br0adband]

Whenever I've noted that DMRDecode isn't processing anything but DSD/DSDPlus show activity (a lot of it, actually), I will go to SDR# and adjust the frequency correct + or - a digit or two, and that typically shows me what's going on with a given frequency). While DSDPlus does have some great aspects, one thing I miss that DSD has is the per-packet decoding percentage while DSDPlus is not nearly as fast with the updates and they're displayed in the title bar of the DSDPlus window (which is freakin' awesome, actually).

Whoever wrote DSDPlus really knows how to add stuff that I (and I'm guessing many others) find appealing with that level of additional info display up there, very very cool stuff.

But with respect to DMRDecode, it seems to be a bit more sensitive to errors in the data stream than DSD or DSDPlus are - they'll show activity and DMRDecode just sits there so when I adjust the frequency correction even 1 ppm sometimes that's all it takes. I'm sure IanWraith will get around to improving it when he's able, it sure would be awesome to get a single app that could do all this.

Since we're dealing with "cheap USB TV tuners" they're going to drift, regularly, even if it's just a single ppm now and again, that's all it takes to kill DMRDecode from doing its job properly, least that's what I've discovered. Now that someone has modded these "cheap USB TV tuners" with a TXCO (temperature controlled oscillator) we should be able to get a rock solid lock on any given frequency - I'd love to have one or two of them myself actually but the $50-60 price for the modded ones is a bit much for me at the moment, and the guy doing the mods is in Japan iirc and he's swamped with orders so there goes that idea.

Basic gist: if you get a decent decode with DSD/DSDPlus (even with a few errors listed) but DMRDecode isn't showing much if any activity at all (if anything it could just be nothing but errors, most of the time it's just no activity whatsoever), definitely definitely make sure you check/adjust the ppm from time to time a single digit or two either way - you'll know it's working/not working by looking at the decoding percentage which will change +/- 10-20% just from one single ppm.

To the author of DSDPlus if you're out there and you happen across this thread and this post specifically:

I love the utility you've created, but if I had to ask for one thing (which is two or three all together I suppose) it would be this - I'd love to be able to see the Group/Call data displayed like it is at the bottom of DMRDecode but inside the window. Since DSDPlus appears to be much more forgiving of errors and there's no apparent need to use the error bar-type display that DSD uses, that extra space could be used to display the Group/Data info meaning we'd see info like this:

-DMR    slot1       BS DATA CC=2 DATA Header         Data call to 204 to 64245
-DMR         slot2  BS DATA CC=2 VOICE Header            Group call from 1 to 115

Notice that the Data/Group info would be offset based on slot/channel as well to make it more readable.

And so on. I know it's a bit more info than most would like I suppose, but when it comes to picking up/monitoring new traffic we may not have seen before it would be a boon to some of us. Right now to find new "stuff" it takes SDR# + DSD/DSDPlus (which gives the audio and other info) + DMRDecode (which is what displays the Groups and some other data info). Biggest thing for me aside from displaying it would be to be able to log that info in the same manner which I could then go back and look at easily.

I'm just tossing out ideas...

 

[Boatanchor]

Sounds like there is too much audio AGC. Try using a fixed audio gain.

I can find no reference to AGC settings in the DSD+ readme file.

I will try and post a couple of recordings of recovered audio of the same P25 transmission. One decoded with DSD and the other with DSD+.

In my case, both software packages give perfect decode but I think the high freq audio rolloff characteristics/Low pass filtering (if it even exists) needs a bit of tweaking in DSD+.

I have tried adjusting the 'rolloff' parameter, but this does not appear to affect audio frequency response.

Would be nice to get an idea of what the additional tuning switches actually do.

 

[slicerwizard]

While DSDPlus does have some great aspects, one thing I miss that DSD has is the per-packet decoding percentage while DSDPlus is not nearly as fast with the updates and they're displayed in the title bar of the DSDPlus window (which is freakin' awesome, actually).

I haven't seen either of those programs display per-packet decoding percentages.

But with respect to DMRDecode, it seems to be a bit more sensitive to errors in the data stream than DSD or DSDPlus are - they'll show activity and DMRDecode just sits there so when I adjust the frequency correction even 1 ppm sometimes that's all it takes.

Then DMRDecode isn't handling DC-shifted audio.

Since we're dealing with "cheap USB TV tuners" they're going to drift, regularly, even if it's just a single ppm now and again, that's all it takes to kill DMRDecode from doing its job properly, least that's what I've discovered. Now that someone has modded these "cheap USB TV tuners" with a TXCO (temperature controlled oscillator) we should be able to get a rock solid lock on any given frequency - I'd love to have one or two of them myself actually but the $50-60 price for the modded ones is a bit much for me at the moment, and the guy doing the mods is in Japan iirc and he's swamped with orders so there goes that idea.

Sounds like someone should write a basic NFM tuner/demodulator app that does auto fine tuning a la TRUNK88. No need for fancy TXCOs or DC shift-aware decoders.

And while we're at it, how about we eliminate the whole VAC, etc. baggage? I think that an auto-tuning NFM demodulator app that makes the audio (unfiltered 16 bit 48 kHz, please) available over something like a simple internal TCP/IP link, combined with decoding apps that can get their raw audio via TCP is the way to go.

I can find no reference to AGC settings in the DSD+ readme file.

Auto output audio gain sounds like AGC to me.

I will try and post a couple of recordings of recovered audio of the same P25 transmission. One decoded with DSD and the other with DSD+.

And DSD+ -g# ?

 

[br0adband]

As for that decoding percentage thing, sorry, my boo - I meant the input level, totally different thing. Was trying to get the point across that the input level reacts immediately with changes to the ppm correction (as it should, obviously). 1 ppm change for me sometimes reflects as a 10-20% change in the input level (inlvl) and directly affects whether DMRDecode gets anything to work with at all, at least in my experience.

In case anyone hasn't heard one of those 'dual channel' decodes (both slots being decoded at once) here's a sample of it, pretty cool stuff and when it first happened last night after I got DSDPlus and fired it up, I was kinda shocked at it, thought I had left DSD running at the same time because I was so used to just hearing the one channel audio. Now we just basically doubled our monitoring!!! (just kidding, just kidding)

https://dl.dropboxusercontent.com/u/19019338/Sample.zip 656KB wav, 21 seconds long

EDIT:
Redid the link with the wav + the srt log so you can see the activity, inside a Zip archive.

 

[SCPD]

I don't have any X2-TDMA or Phase 2 to test with locally. The program claims X2-TDMA support, but I don't see any reference to Phase 2 support. I guess the only way to really tell is to pipe some raw Phase 2 audio in and see what it does.

I couldnt' find any raw Phase 2 samples online. I'm sure they exist somewhere, but I wasn't savvy enough to google them out.

mike

X2-TDMA requires it to be forced from the command line but even then I had issues with this mode. After doing the -fx -xx switches I was able to get some voice decode but it was pretty bad. X2 is very touchy with signal levels so it may not be getting enough signal for proper decode.

DMR is 1000x better than DSD160 and works great.

Has anyone gotten the default MP3 recording to work with LAME? I'm ussing LAME 3.99.5 (32-bit) and DSD creates a WAV file but not an MP3.

UPDATE: So DSD is creating this proprietary WAV file. I can use LAME to manually convert the file to MP3 and it then plays fine..?

TRBO Sample: http://home.comcast.net/~k7cj/tumwater.mp3

 

[garys]

Success! I was able to listen to the DPW on VHF using DSDPlus. I installed a different version of lame_enc.dll among other things. I also got some 935 MotoTrbo audio as well.

I'm hearing the data stream and the voice. Can I get rid of the data stream?

 

[Markb]

Sounds like you are listening to the input side. Go to sound options in control panel and uncheck "listen" under the listen tab located under properties for your selected recording device.

Sent from my Samsung Galaxy S4 using Tapatalk

 

[garys]

Listen is unchecked. It's a simplex system, so I'm listening to all of it.

Sounds like you are listening to the input side. Go to sound options in control panel and uncheck "listen" under the listen tab located under properties for your selected recording device.

Sent from my Samsung Galaxy S4 using Tapatalk

 

[Markb]

sorry, I meant the input side of your audio going into DSD, not the input of the radio system

Sent from my Samsung Galaxy S4 using Tapatalk

 

[garys]

Oh. You may have a point. For some reason all of my SDR# audio is routed via UniTrunker. I have the Zeffie plug in for SDR#/Unitrunker and probably have a setting wrong. I just have to figure out what I screwed up and unscrew it. :lol:

Update: I should unscrew first and type second. I stopped Unitrunker and the data buzz went away, but the audio remained!

I'm running my 346XT holding on the channel and can hear the buzz that way. DSDPlus is producing clear audio that syncs exactly with the buzz on the scanner. DMRDecode displays the data stream as well.

I'm making progress.

sorry, I meant the input side of your audio going into DSD, not the input of the radio system

Sent from my Samsung Galaxy S4 using Tapatalk

 

[AZScanner]

Color me impressed with this program - I'm currently enjoying 99-100% perfect P25 decoding on the Phoenix RWC simulcast system. This program is decoding this system better than any scanner I have ever heard on this system, and better than any other version of DSD for that matter.

Here's the weird part - it seems to decode better with the audio filter ON in SDR#. Anyone else notice this or is this just a quirk of my particular setup?

-AZ