DSD+ 1.51, Unitrunker, R820T, provoice, gain values??

[pkneeyahx]

Hi, I'm having a heck of a time trying to get things just right decoding provoice with DSD+.

I get it, but its choppy and somewhat garbled. Here is a sample of my audio.

https://soundcloud.com/slickshaft/dsdplus-edacs-provoice-decoding

I use two R820T dongles with unitrunker.

I am thinking that it's the unitrunker gain values combined with line-in level. I've been trying for a while adjusting these settings which makes a touch of improvement. but I can never seem to get that magic combination.

The system i'm monitoring is EDACS96. Receiver is set to only listen to the digital voice, output to a vac

i have the virtual audio cable with wide params.. 8000 - 119000 or some crazy thing. bigger pipe is better i heard.

I used sdr# to find out the ppm values of my rtls after they're warmed up. I use those same values in unitrunker. 56 on the signal and 88 on the voice in unitrunker.

my sort of okay gain in unitrunker is 435.
using auto-gain has little effect.
also checking thermal drift doesn't do much.

signal strength is fair. not good or poor, just fair.

my dsd+ startup string is 'dsdplus -i1 -o3 -fp"

1=vac cable
3=speakers

Just looking for some suggestions. Trying to find the magic combination of levels and figures and its frustrating. Thanks!

 

[SCPD]

If your control channel decode rate is fair, your voice decoding won't be any better.

Choppy audio may mean excessive CPU load. Does the CPU exceed 50% on a dual core or 25% on a quad core processor? If one of the cores is saturated with work, it may drop data.

A gain of 435 is a little on the high side. Are you some distance from the station?

Have you set the squelch so that the RTL sticks goes quiet when tuned between stations?

I'd work on getting the control channel decode rate as close to 100% as possible. Look for sources of EMI. You may need a USB extension cable or metal shield to keep unwanted noise out of the RTL stick.

Once you've found something that works well on the signal role receiver, duplicate that for the voice role receiver.

 

[pkneeyahx]

Thanks for your wisdom unitrunker.. i am running it now on a better workstation now. makes a huge difference. Intel Core i5 @ 3.10 GHz.. 16GB RAM. 64bit Windows 7.

I still need to work on the gain a bit. i'm using the Auto Gain and it's okay..

is there a gain rule of thumb? i believe i may be 30-40km away from the tower site.

 

[pkneeyahx]

thanks for moving the thread to a better spot. was unsure if it was dsd, unitrunker related or what.

 

[SCPD]

350 works well for me in urban areas.

In suburban or rural area you may need to turn up the gain ... 450 or so (495 is maximum) may be warranted.

Sounds like you're out in the woods. That would justify the higher gain.

 

[pkneeyahx]

I guess you could say i'm out in the woods. I have found a happy medium at a gain of 405. it's listenable now as opposed to before. Thanks a lot for your help!

another sample of what it sounds like now.
https://soundcloud.com/slickshaft/more-dsdplus-edacs-provoice-decoding