[hpsdr] Fw: Sample rate & Audio output

[john clark]

----- Forwarded Message -----
From: Jeremy McDermond <mcdermj at xenotropic.com>
To: john clark <n0ure at yahoo.com> 
Sent: Wednesday, April 3, 2013 10:47 PM
Subject: Re: [hpsdr] Sample rate & Audio output
 
On Apr 3, 2013, at 9:17 PM, john clark <n0ure at yahoo.com> wrote:

> Thanks Jeremy for the info.
> 
> I thought that we always  had to sample 2x the bandwidth?

Yes.  The Nyquist criteria means that the sample rate has to be twice the bandwidth.  Remember, though, what we're talking about when we're referring to what comes from the hardware, though.  That's 192,000 I *and* Q samples per second.  That means that you're actually receiving 384,000 samples per second, it's just that you're getting two per period of time.  So, it still satisfies Nyquist.

> When I'm running digital modes to DM780 I turn Off display in PowerSDR and sane 20% CPU.
> I set for 48000 as DM780 can only use 3900.

Yeah, then you won't need large sample rates if you're merely trying to recover audio to spit into a digital program.  The increased sample rates are for a couple of things:

1)  Increased visualization bandwidth on the waterfall and panadapter
2)  Being able to run multiple software receivers further away from one another.

> VAC does not do float.

Yeah.  Core Audio (the Mac's sound system) is quite different than Windows.  It is fairly easy to change formats efficiently with it.  Both in my DSP implementation and in DttSP, the core of PowerSDR, all the DSP is done in single precision float.  So, it has to be converted, it's just a matter of whether its done internally or externally to the program.

> I'm having problems with TX PSK-31 via DM780. I hear pops. I assume I have problem with buffers, int rate, int span, or...? ALC is -2 and power is  max 7 watts.

If you're hearing pops, it's probably some sort of latency issue.  There are some documents on reducing latency with PowerSDR.  It tends to be very sensitive to it.  Note that the sample rate setting is only *receive* sample rate.  Transmit sample rate to OpenHPSDR hardware is *always* 48k.  Also check CPU when you're transmitting if that's potentially an issue for you.  The transmit DSP code isn't used when you're not transmitting, and the receive code is *always* used since OpenHPSDR is full duplex.  Therefore transmit takes more CPU than receive.

> VAC is reporting 44100 BW in and out?
> Does Mic on ANAN-10 also go to PowerSDR first, then back to Hermes?

Again, modulation is done in software on the computer, not in hardware.  The audio has to be filtered, modulated, and potentially have the sample rate changed to 48k to go to the hardware.  Mic samples are interleaved with the receive data stream from the hardware.

> I hope to hear from PowerSDR experts. 

Many times if someone starts the conversation you'll get others chiming in.  That's part of why I try to respond to these sorts of questions that have been sitting around for a while even though I'm not a PowerSDR user.

You also might want to mention that you're trying to interface PowerSDR to DM780 and that you're getting pops.  It's likely that someone has encountered the problem and will have helpful hints.

> j

--
Jeremy McDermond (NH6Z)
Xenotropic Systems
mcdermj at xenotropic.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20130404/5e43a33c/attachment-0004.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4361 bytes
Desc: not available
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20130404/5e43a33c/attachment-0004.bin>