DSDPlus Fastlane DSD+ Tuning for Optimum Performance

[N2STJ]

I am running the latest Fastlane DSD+ / FMP24 with an RTL SDR Blog dongle. My setup is working pretty good and I can understand about 75% of the decoded P25p1 signal I receive from my local PD. I am running with the DSD+ default setting but I assume some adjustment is in order to tune my setup for optimum performance. I know that DSD+ has many setting (AMBE, IMBE, Rolloff filter, etc.) that can be tweaked. I have been changing these settings with little success. I read about a tool called DSDTune that appears to no longer be supported. I tried DSDTune last night and received error messages.

Are there any supported tools available to assist with the tuning of the many DSD+ parameters?

73
N2STJ

 

[mtindor]

I am running the latest Fastlane DSD+ / FMP24 with an RTL SDR Blog dongle. My setup is working pretty good and I can understand about 75% of the decoded P25p1 signal I receive from my local PD. I am running with the DSD+ default setting but I assume some adjustment is in order to tune my setup for optimum performance. I know that DSD+ has many setting (AMBE, IMBE, Rolloff filter, etc.) that can be tweaked. I have been changing these settings with little success. I read about a tool called DSDTune that appears to no longer be supported. I tried DSDTune last night and received error messages.

Are there any supported tools available to assist with the tuning of the many DSD+ parameters?

73
N2STJ

As long as you have a decent signal (which in my mind would be a 20dB+ signal (more preferred), you shouldn't have any issues and tuning shouldn't be necessary. FMP* knows how to follow / latch onto a signal of a dongle if "off freq", but if you are in scanning mode (VC/VC combined monitoring where it switches from the control channel to the voice channel to decode the audio), then it might take time for FMP to find the center of the freq if your dongles aren't tuning to the proper freq by default. Yeah you can play with the settings you mentioned, but my advice would be NOT to play with any of them and use defaults. Once you start screwing with those types of settings (and then saving them), you are likely in for a world of hurt. Every signal is different. Some signals (quality strong signals) don't need any special "tuning", and if you go adjusting rolloff and other things, that is specific to a particular signal -- and the next signal you tune to would not necessarily decode best with custom settings that you made to copy another signal.

The key is to use a decent receent vintage dongle that specifically advertises a 0.5 ppm (or less) TCXO, and one that doesn't get so hot to the touch when in use (because ones that lack sufficient cooling really need to "heat up" before you can depend upon their accuracy.

You might have to make some adjustment if your dongle is "off frequency" everywhere you tune it.

I'd use something like SDR# to test that out. Figure out a known signal (lets say you know of a semi-local company with an active repeater on 453.2375. Tune to 453.2375 in SDRSharp and see if you are absolutely centered on the freq when you tune to 453.3275. If you aren't, you'll have to figure out how much you are "off" frequency and then use the "-P" option in your FMP24 batch file to specify a PPM frequency correction value so that FMP24 can tune right to the frequency.

That's just my opinion. And you know what opinions are like. I'm sure somebody else who is an "expert" while chime in with better recommendations. I can say that if you are doing it right (not overloading the dongle with gain, and don't require PPM adjustments, and are monitoring a signal with a decent SNR, there is no reason to have to "tune" anything).

And although FMP24 is pretty good about it, you should make sure that the bandwidth "press b when focused on the FMP24 screen" is set to one appropriate for the signal you are monitoring.

NXDN48: 4.00 bandwith
NXDN96 / DMR: 7 bandwidth
P25 (non-simulcast / none-Phase II): 9.5 bandwidth
P25 P2 or Simulcast systems: 12.5 bandwidth


Mike

 

[Ubbe]

There's a program called dsdtune.exe that needs a recorded sample of the signal, push R on the keyboard to record a 10-20 sec sample and then R again to stop, and it will adjust the parameters until the optimum setting are found for that signal. Copy those values to the startup setting in the windows icon for DSD+. Create one startup icon for each troublesome system with its unique settings. If you are not using windows at all then I have no advice for you.

/Ubbe

 

[N2STJ]

As long as you have a decent signal (which in my mind would be a 20dB+ signal (more preferred), you shouldn't have any issues and tuning shouldn't be necessary. FMP* knows how to follow / latch onto a signal of a dongle if "off freq", but if you are in scanning mode (VC/VC combined monitoring where it switches from the control channel to the voice channel to decode the audio), then it might take time for FMP to find the center of the freq if your dongles aren't tuning to the proper freq by default. Yeah you can play with the settings you mentioned, but my advice would be NOT to play with any of them and use defaults. Once you start screwing with those types of settings (and then saving them), you are likely in for a world of hurt. Every signal is different. Some signals (quality strong signals) don't need any special "tuning", and if you go adjusting rolloff and other things, that is specific to a particular signal -- and the next signal you tune to would not necessarily decode best with custom settings that you made to copy another signal.

The key is to use a decent receent vintage dongle that specifically advertises a 0.5 ppm (or less) TCXO, and one that doesn't get so hot to the touch when in use (because ones that lack sufficient cooling really need to "heat up" before you can depend upon their accuracy.

You might have to make some adjustment if your dongle is "off frequency" everywhere you tune it.

I'd use something like SDR# to test that out. Figure out a known signal (lets say you know of a semi-local company with an active repeater on 453.2375. Tune to 453.2375 in SDRSharp and see if you are absolutely centered on the freq when you tune to 453.3275. If you aren't, you'll have to figure out how much you are "off" frequency and then use the "-P" option in your FMP24 batch file to specify a PPM frequency correction value so that FMP24 can tune right to the frequency.

That's just my opinion. And you know what opinions are like. I'm sure somebody else who is an "expert" while chime in with better recommendations. I can say that if you are doing it right (not overloading the dongle with gain, and don't require PPM adjustments, and are monitoring a signal with a decent SNR, there is no reason to have to "tune" anything).

And although FMP24 is pretty good about it, you should make sure that the bandwidth "press b when focused on the FMP24 screen" is set to one appropriate for the signal you are monitoring.

NXDN48: 4.00 bandwith
NXDN96 / DMR: 7 bandwidth
P25 (non-simulcast / none-Phase II): 9.5 bandwidth
P25 P2 or Simulcast systems: 12.5 bandwidth


Mike

Thanks for the reply! It makes sense that the defaults are pretty much center of the road. My dongle advertises a PPM = 0.5 so I'll try adding this to my startup script. BW is set properly already so that is good to go. My local PD doesn't transmit all that often so I have to wait sometimes quite a while to check if my changes made any difference!

73N2STJ

 

[N2STJ]

There's a program called dsdtune.exe that needs a recorded sample of the signal, push R on the keyboard to record a 10-20 sec sample and then R again to stop, and it will adjust the parameters until the optimum setting are found for that signal. Copy those values to the startup setting in the windows icon for DSD+. Create one startup icon for each troublesome system with its unique settings. If you are not using windows at all then I have no advice for you.

/Ubbe

Thanks for the reply. I tried this tool again today with no luck. When I press R/R, I see two files created in my DSD+ directory. One is a .FM file and the other is a .IQ file. Neither of these are the proper input for DSDTune and I get the error message: "could not get decode string". I tried to input the .wave files that are automatically recorded and I get the same error message.

73
N2STJ

 

[N2STJ]

Can anyone tell me if these are error messages? If so, what do they mean?
Is there anything I can do to prevent them or is this just par for the course?

73
N2STJ

 

[slicerwizard]

Thanks for the reply. I tried this tool again today with no luck. When I press R/R, I see two files created in my DSD+ directory. One is a .FM file and the other is a .IQ file. Neither of these are the proper input for DSDTune and I get the error message: "could not get decode string". I tried to input the .wave files that are automatically recorded and I get the same error message.

73
N2STJ

You're pressing R in FMP24. You're supposed to press it in DSD+.

You also should be using the latest and greatest DSDTune. It sounds like you're working with an older version.

 

[slicerwizard]

View attachment 89555

Can anyone tell me if these are error messages? If so, what do they mean?
Is there anything I can do to prevent them or is this just par for the course?

73
N2STJ

Those are error counts. Definitely something wrong with your setup. You should upload that DSD+ recording so we can see it. Use something like sendspace.com

 

[N2STJ]

You're pressing R in FMP24. You're supposed to press it in DSD+.

You also should be using the latest and greatest DSDTune. It sounds like you're working with an older version.

Very odd. Pressing "R" in my DSD+ window does nothing which is why I pressed "R" in the FMP24 window. I am using DSDTune version 0.1.3 (02/27/17) that I found through this link: DSDTune - The RadioReference Wiki

If there is a more recent release, please point me in the right direction!

Thanks,
N2STJ

 

[N2STJ]

Those are error counts. Definitely something wrong with your setup. You should upload that DSD+ recording so we can see it. Use something like sendspace.com

Let me see what I can do... Are you looking for the auto recorded wave file?

 

[slicerwizard]

Very odd. Pressing "R" in my DSD+ window does nothing which is why I pressed "R" in the FMP24 window.

You probably have keyboard commands disabled. Use the menus to start/stop a raw audio recording of a voice call if you prefer.

I am using DSDTune version 0.1.3 (02/27/17) that I found through this link: DSDTune - The RadioReference Wiki

https://github.com/dreinhold/dsdtune/releases states that 0.1.3 shouldn't be used.

Let me see what I can do... Are you looking for the auto recorded wave file?

No, raw source audio.

 

[N2STJ]

You probably have keyboard commands disabled. Use the menus to start/stop a raw audio recording of a voice call if you prefer.

https://github.com/dreinhold/dsdtune/releases states that 0.1.3 shouldn't be used.

No, raw source audio.

I do not know how to disable/enable command line entry of command. Must be an undocumented feature! Perhaps you can tell he how to enable the command line commands?

If I click on "Record AF" on DSDPlusUI v2.0.9 nothing happens.

Yes, mistake in my previous post. I am using DSDTune v 0.1.2. It appears the problem is that I cannot record raw audio because of the problem mentioned above.

73
N2STJ

 

[BM82557]

 

[N2STJ]

Wow! The event log GUI comes up minimized and I had no idea these pull-down menus were there!

Thanks!
N2STJ

 

[N2STJ]

Ran DSDTune successfully and received the suggested settings. Will give them a try and see if it makes a difference.

I appreciate all your help!

73
N2STJ

 

[N2STJ]

Those are error counts. Definitely something wrong with your setup. You should upload that DSD+ recording so we can see it. Use something like sendspace.com

I have uploaded the raw audio file I used to run DSDTune to sendspace. Here is the link: [email]VC-DSDPlus-Raw-Input_2020-08-09@203706.wav[/email] (5.27MB) - SendSpace.com

 

[slicerwizard]

That audio decodes perfectly here. Didn't have to adjust anything.

So... what's different on your end?

 

[N2STJ]

I have uploaded a HUGE file that is much longer if you want to give it another try: VC-DSDPlus-Raw-Input_2020-08-10@212608.wav

If this file is too big let me know...

My guess is that it has to do with the strength of the signal. When the local PD dispatcher transmits I copy much clearer than when the patrol cars transmit. There also seems to be a difference between correct decoding and the signal being intelligible to me. Give the other file a shot and see if you see any errors and/or issues.

73
N2STJ

 

[mtindor]

I have uploaded a HUGE file that is much longer if you want to give it another try: VC-DSDPlus-Raw-Input_2020-08-10@212608.wav

If this file is too big let me know...

My guess is that it has to do with the strength of the signal. When the local PD dispatcher transmits I copy much clearer than when the patrol cars transmit. There also seems to be a difference between correct decoding and the signal being intelligible to me. Give the other file a shot and see if you see any errors and/or issues.

73
N2STJ

Nobody can click o that link and get that file. You've got a "mailto:" in that link, which makes people's email client open up when they click on it.

 

[Ubbe]

Also zip the file, it could make it smaller in size compared to the wav format.

/Ubbe