Date   

Re: HF+ Latency

Alexandru Csete
 

I tested the Airspy HF+ vs. CloudIQ side by side using two instances of gqrx with audio coming out of both of them. I could barely hear and delay between them, so the difference could not have been more than few milliseconds.

You should repeat your tests using a sample rate close to 786 kHz on the CloudIQ. Somewhere in the pipeline there can be a fixed size buffer. Such buffer will be filled faster at higher sample rate and give lower latency. Indeed, if you look at your numbers:

1288 kHz / 768 kHz = 1.68
426 ms / 251 ms = 1.70

Alex
OZ9AEC


Re: HF+ Latency

Shirley Dulcey KE1L
 

The IC-7600 uses DSP so its latency will also be non-zero, though
lower than most SDRs. The acid test would be to compare the delay of
the SDR to reception on an analog reciever (a synthesized LO is OK,
but not DSP) which will have near-zero delay. Not zero because analog
filters cause delay; the tiny delays in amplifier stages and wire are
negligible.

Latency in an SDR comes from multiple sources. First, the A/D
conversion takes some time (but a small amount). Second, the digital
data has to get to the software somehow; if it's coming from an
external source over Ethernet, USB, or FireWire, it's getting bundled
into packets. The packet size is a tradeoff; larger packets mean less
processing overhead but a longer wait for the data. The SDR software
probably bunches up a number of samples and processes them all at once
rather than doing it byte by byte, so there is a delay while the
processing buffer fills. Finally, the digital processing involves
quite a bit of computing and takes non-zero time.

On Wed, Jan 10, 2018 at 1:54 PM, PMM <pmmeasures@gmail.com> wrote:
Got you, yes sorry had in my head sdr# rather than sdrc

Interesting.

On 10 Jan 2018 18:49, "John Fonseca" <fonsecanyc@hotmail.com> wrote:

PMM,
I'm measuring the time difference between when the audio from the SDR
(either HF+ or CloudIQ) and the audio from the Icom 7600 exits the audio
mixer which all of my radios are connected to. Nothing to do with frequency
changes.

Here are the scenarios for testing:

HF+ -> USB -> PC -> SDR v3 -> USB ->Focusrite 6i6g2 -> Behringer XR12 ->
Presonus 22VSL -> USB -> PC -> Audacity
or
CloudIQ -> GB ethernet switch -> PC -> SDR v3 -> USB _> Focusrite 6i6g2 ->
Behringer XR12 -> Presonus 22VSL -> USB -> PC -> Audacity
and
Icom 7600 -> Behringer XR12 -> Presonus 22VSL -> USB -> PC -> Audacity

I tune to a QSO with one SDR plus the 7600 and record the audio. I then
identify the same audio peak on each (L/R) track, zoom in, and measure the
difference.

I'm not using Spyserver.


John
AC2SV



Re: HF+ Latency

PMM
 

Got you, yes sorry had in my head sdr# rather than sdrc

Interesting.

On 10 Jan 2018 18:49, "John Fonseca" <fonsecanyc@...> wrote:
PMM,
I'm measuring the time difference between when the audio from the SDR (either HF+ or CloudIQ) and the audio from the Icom 7600 exits the audio mixer which all of my radios are connected to.  Nothing to do with frequency changes.

Here are the scenarios for testing:

HF+ -> USB -> PC -> SDR v3 -> USB ->Focusrite 6i6g2 -> Behringer XR12 -> Presonus 22VSL -> USB -> PC -> Audacity
or
CloudIQ -> GB ethernet switch ->  PC -> SDR v3 -> USB _> Focusrite 6i6g2 -> Behringer XR12 -> Presonus 22VSL -> USB -> PC -> Audacity
and
Icom 7600 -> Behringer XR12 -> Presonus 22VSL -> USB -> PC -> Audacity

I tune to a QSO with one SDR plus the 7600 and record the audio.  I then identify the same audio peak on each (L/R) track, zoom in, and measure the difference.

I'm not using Spyserver.  


John
AC2SV



Re: HF+ Latency

John Fonseca <fonsecanyc@...>
 

PMM,
I'm measuring the time difference between when the audio from the SDR (either HF+ or CloudIQ) and the audio from the Icom 7600 exits the audio mixer which all of my radios are connected to.  Nothing to do with frequency changes.

Here are the scenarios for testing:

HF+ -> USB -> PC -> SDR v3 -> USB ->Focusrite 6i6g2 -> Behringer XR12 -> Presonus 22VSL -> USB -> PC -> Audacity
or
CloudIQ -> GB ethernet switch ->  PC -> SDR v3 -> USB _> Focusrite 6i6g2 -> Behringer XR12 -> Presonus 22VSL -> USB -> PC -> Audacity
and
Icom 7600 -> Behringer XR12 -> Presonus 22VSL -> USB -> PC -> Audacity

I tune to a QSO with one SDR plus the 7600 and record the audio.  I then identify the same audio peak on each (L/R) track, zoom in, and measure the difference.

I'm not using Spyserver.  

John
AC2SV


Re: HF+ Latency

John Fonseca <fonsecanyc@...>
 

Hi, Simon.

Odd high or odd low?  The audio latency reading in SDR Console floats around the 50ms range.  The audio buffer is set to minimum.  For a sound card I'm using a Focusrite 6i6gen2.

When I switched from using an Elad S2 to the CloudIQ I did a similar comparison and saw a similar difference (unfortunately I didn't save those results) and when I compared to what I saw for the Perseus on https://www.n1kdo.com/sdr-delay-measured/ it seemed to be in the same ballpark.

John
AC2SV


Re: HF+ Latency

PMM
 

What are you trying to measure? 
Time it takes to change frequency / time it take audio to recommence after frequency change?  Other. Is that factoring in buffers? How much is being buffered.

Once you streaming it not that relivant unless.overily high think pings 400+ms I can understand a need for low latency on frequency shifting though as user may send quite a few.small frequency shifts in a small space.of time when scrolling the spectrum display instead of jumping directly onto a frequency.

There maybe work there to do on the spyserver given what I have seen of its working in my testing. Maybe programmer could put in a settling period before jumping to a.freuqncy so a number of rapid short step frequencies are ignored unless a.time period greater than x after last received step then jump frequency so fast small steps are ignored if choosen to via config line. E.g. if 100 * 1ms steps revived for just a few Hertz don't action previous 99 until number 100 time > x

I know I diverting off topic but that behaviour could be adding to perceived latency and a bit of magic coding might make things feel a bit more snappy.



On 10 Jan 2018 17:52, "Simon Brown" <simon@...> wrote:

John,

 

The main factor is the software, whether the software is engineering for lowest possible latency. There’s something a bit odd with the figures you achieved below.

 

Simon Brown, G4ELI

 

www.dxgalaxy.com
www.sdr-radio.com
www.sdr-satellites.com

 

 

From: main@airspy.groups.io [mailto:main@airspy.groups.io] On Behalf Of John Fonseca
Sent: 10 January 2018 17:26
To: main@airspy.groups.io
Subject: [airspy] HF+ Latency

 

I was just curious if anyone has compared latency on the HF+ to any other SDR receivers they have.

I did a non-scientific test comparing it to a Cloud-IQ by recording the output of my mixer with an Icom 7600 panned hard left and SDR panned hard right and then comparing the distance between the same peak on each track in Audacity with SDR Radio v3 as the SDR software.  The connected SDR receiver is split from the 7600's RX loop.

Of course there are many contributing factors to overall latency (PC/software/soundcard/DSP) but all was the same the same for both SDR receivers.

What I saw:
HF+ (BW 786 kHz):  426ms
CLoudIQ (BW 1.2288 MHz):  251ms

Has anyone else looked at this?  

Thanks,
John
AC2SV


Re: HF+ S11 measurements

W0LEV
 

So, to make a good S11 measurement, the source power should be kept at or less that -40 dBm for the HF+. 

My previously stated goal of -30 dBm max for the measurement is an absolute max (for me).  When measuring low-noise preamps, I usually hit them with -50 dBm.

Dave - WØLEV

On Tue, Jan 9, 2018 at 6:41 PM, Leif Asbrink <leif@...> wrote:
Hello Dave,

-30 dBm is too much signal for the Airspy HF+.
The AGC will trigger at -34 dBm and change the
input impedance (and degrade the NF by about 5 dB.)

Have a look at these videos (They were intended to
be a single video....)

https://www.youtube.com/watch?v=6kldqiwHCK8

https://www.youtube.com/watch?v=iHqMxxuoU7E

As it turns out, the Airspy ONE is a trifle less
affected than the HF+ by signals in the 88 to 108 MHz
range when one listens to 144 MHz. That is true only
for a small range of interference levels however.
The HF+ has a graceful degradation due to the front
end AGC so receive works even when very strong signals
are present in the FM band while the Airspy One has
a disgraceful degradation (One would have to turn
down the gain manually which I did not do.)

The Airspy input impedance on the VHF input changes
in steps. It switches at 117.9 MHz so my testing
on 144 MHz is done with 88-108 probably a bit outside
the input filter. The next change comes when the frequeny
reaches 1199.9 MHz. There is hysteresis on the switching
however, it switches back when the frequency goes below
279.9 MHz. The HF+ is however dead in the range 279.1
to 1200 MHz so it does not matter which one of the
filters is in line.

For my unit I find

Lower VHF 60 to 118 MHz
Upper VHF 118 to 279 MHz
UHF 1200 to 1674 MHz.

The UHF range is useless because of VERY strong sideband
spurs. Looks like a defective polyphase mixer because the
spur levels depend strongly on the offset from the center
frequency. If I place the signal close to the LO I see no
spurs, but if I move 20 kHz away from the center, the
spurs about ±23 kHz on both sides of the signal show up
and they grow stronger as the difference between signal
frequency and LO increases.

The sidebands at ±550 kHz are equal in strength
with the main signal. There are MANY spurs, separated by
about 23 kHz. Presumably this is a production error in
the chip, if it were a design error, it should have been
well known by now...


73

Leif



> I do not have the VNWA3, but use the HP 8753C VNA with the S-Parameter Test
> Set, HP 85047.  I've found that just connecting the source port to the
> input of a radio, or, more commonly, the input of a preamp (DUT - device
> under test), an attenuator must be inserted between the source energy and
> the DUT.  This attenuator can be calibrated out of the measurement setup.
> Problem is that the source power is usually too much for the 'front end' of
> most DUTs to handle in a linear manner.  I usually try to hit an active DUT
> with no more than -30 dBm in an attempt to assure the DUT front end is not
> overloaded and remains in its linear operating range.  When attempts are
> made to measure S11 with the DUT at or beyond the 1 dB compression point,
> the reading of S11 will not represent the small signal input impedance.
>
> I do not have an AirSpy HF+, or I'd measure using my VNA for the group.
> Possibly the designers of the AirSpy HF+ could relate what the 1 dB
> compression point is so we could make a good measurement of S11 while
> assuring nothing is in the non-linear operating region.
>
> Dave - WØLEV
>
> On Sun, Jan 7, 2018 at 6:26 PM, Dave GW4GTE <dave@...> wrote:
>
> > Hi all,
> >
> > Some test results as promised - see 2 attachments:
> >
> > S11 measurements using VNWA3 via 15cm of sma cable into the relevant HF+
> > port.
> >
> > I make no comment other than to invite others to repeat a similar test if
> > anyone's interested.
> >
> > Dave GW4GTE
> >
> >
> >
>
>
> --
>
> *Dave - WØLEV*
> *Just Let Darwin Work*






--
Dave - WØLEV
Just Let Darwin Work


Re: HF+ Latency

Simon Brown
 

John,

 

The main factor is the software, whether the software is engineering for lowest possible latency. There’s something a bit odd with the figures you achieved below.

 

Simon Brown, G4ELI

 

www.dxgalaxy.com
www.sdr-radio.com
www.sdr-satellites.com

 

 

From: main@airspy.groups.io [mailto:main@airspy.groups.io] On Behalf Of John Fonseca
Sent: 10 January 2018 17:26
To: main@airspy.groups.io
Subject: [airspy] HF+ Latency

 

I was just curious if anyone has compared latency on the HF+ to any other SDR receivers they have.

I did a non-scientific test comparing it to a Cloud-IQ by recording the output of my mixer with an Icom 7600 panned hard left and SDR panned hard right and then comparing the distance between the same peak on each track in Audacity with SDR Radio v3 as the SDR software.  The connected SDR receiver is split from the 7600's RX loop.

Of course there are many contributing factors to overall latency (PC/software/soundcard/DSP) but all was the same the same for both SDR receivers.

What I saw:
HF+ (BW 786 kHz):  426ms
CLoudIQ (BW 1.2288 MHz):  251ms

Has anyone else looked at this?  

Thanks,
John
AC2SV


Re: HF+ Latency

Gregory Schippers
 

I don't need to look.... I can hear it. My guess is ethernet connected will always give less latency. Interested to see input on this as well.... 
Greg, KC8HXO 


On Wednesday, January 10, 2018, John Fonseca <fonsecanyc@...> wrote:
I was just curious if anyone has compared latency on the HF+ to any other SDR receivers they have.

I did a non-scientific test comparing it to a Cloud-IQ by recording the output of my mixer with an Icom 7600 panned hard left and SDR panned hard right and then comparing the distance between the same peak on each track in Audacity with SDR Radio v3 as the SDR software.  The connected SDR receiver is split from the 7600's RX loop.

Of course there are many contributing factors to overall latency (PC/software/soundcard/DSP) but all was the same the same for both SDR receivers.

What I saw:
HF+ (BW 786 kHz):  426ms
CLoudIQ (BW 1.2288 MHz):  251ms

Has anyone else looked at this?  

Thanks,
John
AC2SV


AirSpyHF+ macOS Server

Chris Smolinski
 

I've just released my macOS server app for the AirSpyHF+. This app lets you use your AirSpyHF+ on the Mac with the SdrDx SDR app, without any need for Windows. Works with 10.6 through 10.13.

Instructions and the download link here: https://blackcatsystems.com/software/airspyhf_macos_server.html

Chris Smolinski
Black Cat Systems
http://www.blackcatsystems.com


HF+ Latency

John Fonseca <fonsecanyc@...>
 

I was just curious if anyone has compared latency on the HF+ to any other SDR receivers they have.

I did a non-scientific test comparing it to a Cloud-IQ by recording the output of my mixer with an Icom 7600 panned hard left and SDR panned hard right and then comparing the distance between the same peak on each track in Audacity with SDR Radio v3 as the SDR software.  The connected SDR receiver is split from the 7600's RX loop.

Of course there are many contributing factors to overall latency (PC/software/soundcard/DSP) but all was the same the same for both SDR receivers.

What I saw:
HF+ (BW 786 kHz):  426ms
CLoudIQ (BW 1.2288 MHz):  251ms

Has anyone else looked at this?  

Thanks,
John
AC2SV


Re: Spyserver as a Windows Service

PMM
 

On 10 Jan 2018 16:04, "Paul Mclaren MM0ZBH" <paul.mcl@...> wrote:
Thanks for the replies. Looks like it will be a 3rd party solution.....I just wanted to make sure there wasn't something built in to Spyserver I was missing.

Regards

Paul

On Wednesday, 10 January 2018, Daniel Fox <foxd@...> wrote:
> https://www.howtogeek.com/50786/using-srvstart-to-run-any-application-as-a-windows-service/
>
>  
>
> Dan
>
>  
>
> From: main@airspy.groups.io [mailto:main@airspy.groups.io] On Behalf Of Tatamare
> Sent: Wednesday, January 10, 2018 8:53 AM
> To: main@airspy.groups.io
> Subject: Re: [airspy] Spyserver as a Windows Service
>
>  
>
> ·         To add or remove an app from the Startup tab, press the Windows Logo Key  + R, typeshell:startup, and then select OK. This opens the Startup folder. Select Start . Find the app you're looking to add or remove, and press and hold (or right-click) it. Select More > Open File Location. In the file location folder, press and hold (or right-click) the app and select Copy. Paste the shortcut to your app into the Startup folder to have it run at startup.
>


Re: Airspy Mini USB issues

Bart AA7VA
 

Nick wrote "FWIW, the physical hardware seems fine -- despite the fact that my Mini does get plugged/unplugged quite a bit.  I do notice the Mini gets hot during normal use but that seems relatively harmless."

Agreed until the point of the excess heat being "harmless - It only took 2 months for mine to burn out dead as can be - and this is a common problem mentioned on other blog sources.  Go for the R2 or another brand, the Mini is a problem unit.

--
Bart Van Allen
AA7VA



Re: Spyserver as a Windows Service

Paul Mclaren MM0ZBH
 

Thanks for the replies. Looks like it will be a 3rd party solution.....I just wanted to make sure there wasn't something built in to Spyserver I was missing.

Regards

Paul


On Wednesday, 10 January 2018, Daniel Fox <foxd@...> wrote:
> https://www.howtogeek.com/50786/using-srvstart-to-run-any-application-as-a-windows-service/
>
>  
>
> Dan
>
>  
>
> From: main@airspy.groups.io [mailto:main@airspy.groups.io] On Behalf Of Tatamare
> Sent: Wednesday, January 10, 2018 8:53 AM
> To: main@airspy.groups.io
> Subject: Re: [airspy] Spyserver as a Windows Service
>
>  
>
> ·         To add or remove an app from the Startup tab, press the Windows Logo Key  + R, typeshell:startup, and then select OK. This opens the Startup folder. Select Start . Find the app you're looking to add or remove, and press and hold (or right-click) it. Select More > Open File Location. In the file location folder, press and hold (or right-click) the app and select Copy. Paste the shortcut to your app into the Startup folder to have it run at startup.
>
> _._,_._,_
> ________________________________
> Groups.io Links:
>
> You receive all messages sent to this group.
>
> View/Reply Online (#26782) | Reply To Group | Reply To Sender | Mute This Topic | New Topic
>
> Change Your Subscription
> Group Home
> Contact Group Owner
> Terms Of Service
> Unsubscribe From This Group
>
> _._,_._,_


Re: Spyserver as a Windows Service

Daniel Fox
 

From: main@airspy.groups.io [mailto:main@airspy.groups.io] On Behalf Of Tatamare
Sent: Wednesday, January 10, 2018 8:53 AM
To: main@airspy.groups.io
Subject: Re: [airspy] Spyserver as a Windows Service

 

·         To add or remove an app from the Startup tab, press the Windows Logo Key  + R, typeshell:startup, and then select OK. This opens the Startup folder. Select Start . Find the app you're looking to add or remove, and press and hold (or right-click) it. Select More > Open File Location. In the file location folder, press and hold (or right-click) the app and select Copy. Paste the shortcut to your app into the Startup folder to have it run at startup.


Re: Spyserver as a Windows Service

David J Taylor
 

Thanks I will try that but I think that will only work after a user logs on (but I may be wrong). I need it to start when the PC turns on.

Regards

Paul
===============================

Well, there is a common startup as well. But you could use:

- Task scheduler - set the .exe to run when Windows starts (might be best to give it a 60 second delay in case it depends on e.g. networking being available) Be aware of what context the program is run, i.e. as Administrator it may need different settings to as a regular user. An elevated program may have reduced levels of network access.

- Use SRVANY.exe which I've seen recommended but not used myself:

https://support.microsoft.com/en-us/help/137890/how-to-create-a-user-defined-service

Experiment and see!

Cheers,
David
--
SatSignal Software - Quality software written to your requirements
Web: http://www.satsignal.eu
Email: david-taylor@blueyonder.co.uk
Twitter: @gm8arv


Re: Spyserver as a Windows Service

Paul Higginson
 

Try one of the 3rd party server starter tools, I've no particular recommendation as I've never tried this on a 'small' pc.

https://docs.microsoft.com/en-us/dotnet/framework/windows-services/introduction-to-windows-service-applications is useful reading

Paul GW8IZR


On 10/01/2018 14:51, Paul Mclaren MM0ZBH wrote:
Thanks I will try that but I think that will only work after a user logs on (but I may be wrong).   I need it to start when the PC turns on.

Regards

Paul

On 10 January 2018 at 13:53, Tatamare <yo7hhi@...> wrote:
  • To add or remove an app from the Startup tab, press the Windows Logo Key  + R, typeshell:startup, and then select OK. This opens the Startup folder. Select Start . Find the app you're looking to add or remove, and press and hold (or right-click) it. Select More > Open File Location. In the file location folder, press and hold (or right-click) the app and select Copy. Paste the shortcut to your app into the Startup folder to have it run at startup.




moderated Re: HF+ Delayed Airmail Delivery?

Ruben Navarro Huedo (EA5BZ)
 

Yes...ordered 11/11 and still waiting.

Air mail here.


Re: Spyserver as a Windows Service

Paul Mclaren MM0ZBH
 

Thanks I will try that but I think that will only work after a user logs on (but I may be wrong).   I need it to start when the PC turns on.

Regards

Paul

On 10 January 2018 at 13:53, Tatamare <yo7hhi@...> wrote:
  • To add or remove an app from the Startup tab, press the Windows Logo Key  + R, typeshell:startup, and then select OK. This opens the Startup folder. Select Start . Find the app you're looking to add or remove, and press and hold (or right-click) it. Select More > Open File Location. In the file location folder, press and hold (or right-click) the app and select Copy. Paste the shortcut to your app into the Startup folder to have it run at startup.



moderated Re: HF+ Delayed Airmail Delivery?

Chris DL5NAM
 

Have ordered via German dealer WIMO - still waiting till Pony Express arrive WIMO.

 

Next time i order in China direct = faster and cheaper ! Will wait till friday then i cancel my order.

 

Chris

15501 - 15520 of 42211