Bad DSD+ decode rate, help with tuning PC input requested

[mitaux8030]

I'm having troubles with poor/zero decode rate on DSD+ v1.074 with 9600bps digital voice, on a HP 6730B laptop.

Here's how things have gone down so far:
PC A is a desktop.
PC B is a laptop, as above.
RX Y is a BC780 with discriminator tap installed.
RX Z is an Pro2006, with discriminator tap installed.
Receiving a clean & strong 9600 provoice signal.

PC A + RX Y = good decode.
PC A + RX Z = good decode.
PC B + RX Y = no decode at all.
PC B + RX Z = no decode at all.

Then I saved raw files, and swapped them between PCs:
PC A + RX Y recorded, then transferred to PC B: good decode.
PC A + RX Z recorded, then transferred to PC B: good decode.
PC B + RX Y recorded, then transferred to PC A: no decode at all.
PC B + RX Z recorded, then transferred to PC A: no decode at all.

From the above, I conclude that both receivers are working just fine, and that PC A and PC B have the computing horsepower to successfully decode from files, and that the DSD+ software is working on either PC without problem. The problem appears to be PC B's (the laptop) ability to accurately sample the discriminator tap being input to it.

For comparison, I have included a picture of the scope being used on PC B while decoding from a 'good' and 'bad' recorded .wav file. Not much difference between the two is there? And listening by ear, there's barely any difference between the samples either.

I've tried inverting polarity, which made absolutely no change. I've tried tweaking input levels on the laptop, again no dice. AGC / microphone boost is off for the microphone input. What else can I do to improve things?
Interestingly, PC B live input seems to decode OK when tuned in to a 4800 DMR system.
For reference: Laptop is a HP 6730b, running WinXP, and for whatever reason I have been unsuccessful in getting the microphone socket to work as a line in port. It'll only work as a microphone input.

 

[racingfan360]

Laptop built-in microphones are notoriously difficult to get working with DSD+ : if you try searching this forum you'll often see it mentioned. For best results it really needs a line-in feed, and it's configuring many laptop sound cards in windows for line-in that can be a nightmare if not impossible. If you're hard set on sticking with the mic in feed, then try playing about with the sampling rate: higher sampling rate usually equals better decode results, and low sampling rate will guarantee you'll hit issues.

An external USB sound card will almost certainly solve your problem: I thoroughly recommend the "StarTech.com ICUSBAUDIO2D USB Stereo Audio Adapter External Sound Card with SPDIF Digital Audio Out" which covers up to 96khz sampling rates.

 

[slicerwizard]

Any USB sound device, even the inexpensive ones found on eBay or DealExtreme, will probably solve the OP's problem.

Like this: Virtual 5.1 Channel USB Sound Card Adapter - Black - Free Shipping - DealExtreme


BTW, those two screenshots look to be near-identical; it would be better if the OP uploaded his samples to ZippyShare, etc.

 

[mitaux8030]

Thanks for that info. I'm testing on a laptop for the first time and decided to throw it at the hardest test of 9600bps voice. If it worked for that, then it'd have to work for 4800bps DMR, which is all I really needed it to do anyway. Since it's working for that, no need to go any further - but might pick up a USB sound dongle anyway, just in case I want to play with some other modes.

 

[slicerwizard]

Say what? You threw simple two level ProVoice at it. The four level signals used by NXDN, DMR and P25 bring their own set of issues.