DSDplus can't tune with DSDtune or by hand

[rubdom]

Hi everyone,

I have DSDplus decoding Provoice. I am trying to fine tune it but for some reason I can't use DSDtune or fine tune it by hand using the steps in DSDPlus.txt.

I recorded a 30 second sample of audio and it created a file (tried several files). I then tried to issue this command.

dsdplus ? DSDPlus-Raw_Input_2015-11-20@205102.wav -o0 -O NUL -fp -dr1

It opens a new window with dsdplus and it runs but the window closes too fast for me to see what is happening. I tried to log what happened by using -v4 >tunelog but the file was empty. Not sure what is going on. Am I doing something wrong?

When I use Dsdtune I get the error: Could not get decode string. I have gone through many different troubleshooting steps with DSDtune also.

I have been messing with this for 2 days and haven't been able to get either to work.

I am getting some errors with decoding and garbled sound. I have ran through every other possible tuning solution (gain, sound, taking filters off, moved off simulcast tower a regular tower, and 1000 other things).

Any ideas would be helpful!

 

[deboe]

Do you get any messages in the command window after runing that dsdplus command?
Which version of dsdplus are you running?
As for dsdtune, it's really only for fine tuning. So you probably need to work on your input signal first. Yeah like you said gain, filters will give you better results. Once you get a pretty good decode then dsdtune might be able to get you that extra 1% better.

 

[cg]

I am not good at documenting or remembering why I use certain parameters but here is what I used in my last run with the latest DSDPlus. I rename the .wav file to something simple like "test" or "test1, 2, 3, etc"" rather than the long name generated by DSDPlus.

DSDTune -i testa.wav -fn -x dsdplus.exe -b run_dsd.bat

You need to run this from within the DSDPlus folder
Change -fn to the mode you want to decode (-fp)
This will place the settings into a batch file named "run_dsd"

chris

 

[rubdom]

Thanks for the help.

@deboe
I don't get any messaged in the window once I run my commands and the window that pops up with Dsdplus closes really fast.
I am running DSDplus version 1.074
I have tried several parameters with the same results. Some decoding is really really good with zero errors. Some transmissions are really good with about 3-4 errors. I can even have transmissions come out perfect on and some really bad (even on the same talk group and frequency). Errors are usually e:1111 or e:1121 and at the and of the transmissions it gets worse e:frdrfr10r.

Could the control channel have anything to do with this? I am getting 100 health at most times. I went it there to mess with it a bit and turned down the gain....this made the signal health go down so I turned it back to 300.

@cg

I tried to run those parameters (with provoice) in dsdtune and it gave me the same error "Could not get decode string"
DSDplus will start and run then close but still get the same error.
Here is the full output:

Checking option -dr
Running dsdplus.exe -o0 -o -fp -drXXX ? test1.wav
Could not get decode string

 

[cg]

you do know that DSDTune is a separate program and you need to run that and not DSDPlus for tuning correct??

Only two of the DSDPlus "-d" options are used with ProVoice, -dr and -dh. Default for each are 0 and that is Auto settings.

You posted that you used :
DSDPlus.exe -o0 -o -fp -drXXX ? test1.wav
That does not make sense to me. If you are running DSDTune, you need to run
DSDTune.exe -i ? test1.wav -fp run_dsd.bat
that is, program, input source, mode, result file
In that example, dsd_bat would give you a batch file that had the results of the run, something like
DSDPlus -fp -dr2 -dh5
Every parameter that is not seen uses the defaults.

chris

 

[rubdom]

you do know that DSDTune is a separate program and you need to run that and not DSDPlus for tuning correct??

Only two of the DSDPlus "-d" options are used with ProVoice, -dr and -dh. Default for each are 0 and that is Auto settings.

You posted that you used :
DSDPlus.exe -o0 -o -fp -drXXX ? test1.wav
That does not make sense to me. If you are running DSDTune, you need to run
DSDTune.exe -i ? test1.wav -fp run_dsd.bat
that is, program, input source, mode, result file
In that example, dsd_bat would give you a batch file that had the results of the run, something like
DSDPlus -fp -dr2 -dh5
Every parameter that is not seen uses the defaults.

chris

Hey Chris. Sorry, what I posted was a bit confusing. I tried running DSDtune (which didn't work see how I used it below and the error it throws) and manually tuning in DSDplus (using the commands in DSDplus first starting with the dr switch).

Here is how I used DSDtune:

DSDTune -i testa.wav -fp -x dsdplus.exe -b run_dsd.bat

Here is what was returned to me on the cmd line:

Checking option -dr
Running dsdplus.exe -o0 -o -fp -drXXX ? test1.wav
Could not get decode string

So as you see there is an error of some kind with DSDtune and I am getting "Could not get decode string". Not sure why it won't work. Also running tuning manually in DSDplus won't work either....the program just runs then closes out.

 

[cg]

Are you using SDR dongles or a tapped radio?
Perhaps you could post a raw audio file.
I have never used DSDPlus on ProVoice as there is only occasional a brief exchange on a talkgroup on an EDACS system locally.

I am using the latest DSDPlus Fast Lane release and the latest DSDTune release I saw. I don't know if DSDTune has a version number but I believe the latest was from June.

Did you try DSDPlus with your batch file altered to include -dr & -dh with values? It would take a bit but there is not that many possible combinations (40).

chris

 

[rubdom]

Are you using SDR dongles or a tapped radio?
Perhaps you could post a raw audio file.
I have never used DSDPlus on ProVoice as there is only occasional a brief exchange on a talkgroup on an EDACS system locally.

I am using the latest DSDPlus Fast Lane release and the latest DSDTune release I saw. I don't know if DSDTune has a version number but I believe the latest was from June.

Did you try DSDPlus with your batch file altered to include -dr & -dh with values? It would take a bit but there is not that many possible combinations (40).

chris

I am using SDR dongles for both the control channel and voice channel.

Here is a link to a raw file. https://drive.google.com/file/d/0B-s2SlHNBUNSdy1XVmZ4Tzd3bUE/view?usp=sharing
Notice that there are lots of errors on some transmissions and very few on others. Not sure why that is. I can get some transmissions with zero errors and the next on the same talk group will have many errors.....mystery.

I am using the latest version of DSDplus and also DSDtune. I also tried using the old version of DSDplus just to see if I had better results but it was the same.

Great idea on running DSDplus with -dr and -dh values with live audio. I will try that and see if I can tune it that way.

 

[cg]

That file ran fine with DSDPlus 1.100 and the June DSDTune file.
I got -dr1 -dh2 from the sample you posted.

chris

 

[rubdom]

That file ran fine with DSDPlus 1.100 and the June DSDTune file.
I got -dr1 -dh2 from the sample you posted.

chris

Thanks for that. I am running DSDplus 1.074. Where can I get 1.1?

 

[cg]

That is the Fast Lane version. Try this post for info:
http://forums.radioreference.com/vo...323575-how-join-dsd-fastlane.html#post2481097

chris

 

[rubdom]

That is the Fast Lane version. Try this post for info:
http://forums.radioreference.com/vo...323575-how-join-dsd-fastlane.html#post2481097

chris

Thanks!

 

[FR3500]

Hello i haven ' t used dsdtune yet; i expect to give it a try today;
there's a signal on DMR i can decode with bad quality;

In order to test a sample, do i have to record voice, or idle repeater works too? cause the freq is very quiet, it may take long to catch voice calls ...

Does dsdtune work with all DSD issues ?

thanks

 

[cg]

The function of DSDTune is to automate determining the best filter settings in DSDPlus. It is nothing that you couldn't do yourself but as there are a huge number of possible combinations, it helps to automate it.
It will however, not help with poor signal, hardware problems, bad file configuration, etc. In other words, if you have a decent signal but are seeing too many errors, it may help.

It is done with the file generated by pressing "r" during a transmission. As for testing it on the control channel, you could try. I am not sure if the 5 filters have to do with voice reproduction or are in the decoding in general.

chris

 

[slicerwizard]

I recorded a 30 second sample of audio and it created a file (tried several files). I then tried to issue this command.

dsdplus ? DSDPlus-Raw_Input_2015-11-20@205102.wav -o0 -O NUL -fp -dr1

It opens a new window with dsdplus and it runs but the window closes too fast for me to see what is happening.

Usage:
  DSDPlus [options]            Decode from audio stream
  DSDPlus [options] ? file     Decode from .wav file
  DSDPlus -h                   Show help

The help text shows the "? filename" after any other options, not before.

And if you do this stuff from a command prompt, you won't have to deal with any window closing before you can read error messages.

 

[rubdom]

Usage:
  DSDPlus [options]            Decode from audio stream
  DSDPlus [options] ? file     Decode from .wav file
  DSDPlus -h                   Show help

The help text shows the "? filename" after any other options, not before.

And if you do this stuff from a command prompt, you won't have to deal with any window closing before you can read error messages.

Tried that. Still can't get it to work. It runs through the file and closes the window. I did go through command prompt....opens a separate window anytime I run dsdplus. I even tried running through a shortcut and resetting the target to where cmd.exe lies with a /k. Nothing is workin

I was able to see what it was doing before it closes out the window. Its just running through the wav file and closing out and it isn't giving me a score like the help file says. Its like it is ignoring any -dr or -dh commands.

DSDtune continues to give me the error "could not get decode string" Not sure what is going on there as well. I am going to go for the DSDplus fastlane option to see if that is any better.

 

[racingfan360]

rubdom: Please post a link to the raw file that you are actually trying to try this test with please.

Are you using this one:
>Here is a link to a raw file. https://drive.google.com/file/d/0B-s...ew?usp=sharing
Like others above, I can decode it fine. I get a marginal improvement in decode score adding the -dr1 -dh2 functions.

> It runs through the file and closes the windo
One reason it would do this is if there is NO voice traffic to decode in the raw wav sample. So are you using a valid raw wav file?

 

[slicerwizard]

Tried that. Still can't get it to work. It runs through the file and closes the window. I did go through command prompt....opens a separate window anytime I run dsdplus. I even tried running through a shortcut and resetting the target to where cmd.exe lies with a /k. Nothing is workin

I was able to see what it was doing before it closes out the window. Its just running through the wav file and closing out and it isn't giving me a score like the help file says. Its like it is ignoring any -dr or -dh commands.

DSDtune continues to give me the error "could not get decode string" Not sure what is going on there as well. I am going to go for the DSDplus fastlane option to see if that is any better.

Hm, I just tried DSDPlus ? "filename.wav" and here's the tail end of what I got:

Sync:+ProVoice VOICE
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE e:1
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE e:1
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE e:14ErEr
Sync: no sync

35.0 seconds (1,120,000 bytes) of synthesized audio appended to 'DSDPlus.wav'

decoding score = 65,988


1.074 and 1.101 are producing identical output here.

Tried some -dr and -dh values and the decoding score went down.
Looks like my recording doesn't need tuning...

 

[rubdom]

rubdom: Please post a link to the raw file that you are actually trying to try this test with please.

Are you using this one:
>Here is a link to a raw file. https://drive.google.com/file/d/0B-s...ew?usp=sharing
Like others above, I can decode it fine. I get a marginal improvement in decode score adding the -dr1 -dh2 functions.

> It runs through the file and closes the windo
One reason it would do this is if there is NO voice traffic to decode in the raw wav sample. So are you using a valid raw wav file?

That file is one of several that I have tried to use. I did ensure there is voice traffic when recording the samples. I am using a valid raw wav file.

 

[rubdom]

Hm, I just tried DSDPlus ? "filename.wav" and here's the tail end of what I got:

Sync:+ProVoice VOICE
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE e:1
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE e:1
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE
Sync:+ProVoice VOICE e:14ErEr
Sync: no sync

35.0 seconds (1,120,000 bytes) of synthesized audio appended to 'DSDPlus.wav'

decoding score = 65,988


1.074 and 1.101 are producing identical output here.

Tried some -dr and -dh values and the decoding score went down.
Looks like my recording doesn't need tuning...

Thanks for that. I have been running live audio using various -dr and -dh switches since I can't seem to get anything else to work and I seem to get more errors when i use them. Still no way to get a decoding score but my tests are inline with your decoding score going down with any switches turned on.