Which online Kiwi sdr is worth visiting.
in General Chat
I have been interested for some time in what can be heard on the other side of the world on shortwave so I started looking into the KIWI SDR phenomenon. After visiting several sites featuring this SDR service, I was somewhat disappointed. Out of over 600 SDR receivers online, some operators of these online SDRs don't have their act together. A large proportion have poor reception or have no reception signal at all. To avoid disappointment, among potential listeners looking for that one perfect SDR, I decided to set up this website for them.
I am independent, have no affinity with KIWI sdr or any operator.
I maintain this site beside my ham-radio hobby (PD4P)
Comments
I forogt to mention the website: www.online-sdr.info
I have 3 systems online and am never happy with any of them. RFI, CMI, and overload can be tough enough on a narrow BW system but the 0-30 MHz nature of a Kiwi can make it even harder.
jimlill.com:8073 and 8075 are here at my home QTH whereas I have a remote system in Maine rx2.wa2zkd.net:8073
-Jim
Agreed only about 10% of all the KiWis are performing at anything like their full potential.
Use this dynamic list of KiWi's, but only bother with the first 100.
If you sort by the second column, Internal HF new, by clicking on the header, you will get an even better idea if which ones are the best.
BTW I'm frequently at the number one position during early night time in here the UK, and others I'm involved with are usually somewhere towards the top. But it's not been easy or cheap to achieve it.
http://wessex.zapto.org:8073/
Regards,
Martin
Hi Jim, thanks for the info. My website is not set up to give an indication whether or not an online SDR is good. I (try) to visit as many SDRs as possible per day just to make a screen print. In other words, what can users expect when they log in to a particular Kiwi. I want to visit all 600+ before I actually listen. And trust me, I've visited your SDRs and they are by no means the worst.
Regards Peter (PD4P)
Have a look at the better WSPR/FST4W reporters to get additional insight into what is working well.
again, not a perfect metric because it's dependent upon geolocation., MUF and latitude,but still very useful.
Glenn n6gn
Hi Peter,
Nice initiative. The snapshots taken of course will change a lot over time due propagation, varying local QRM and other factors including the operator switching antennas. Perhaps having 4 smaller scale vertically stacked snapshots at 6 hours interval would give a better overall picture.
Related to a visual indication of the waterfall another indication of performance is the SNR measurements that Martin G8JNJ mentioned.
You are probably aware that you can use search critera on http://kiwisdr.com/public/ and with a minimum SNR value can weed out the poorly performing HF Kiwis with persistent low SNR's over time.
Included plots show SNR varying over time and frequency band for some of the better Kiwis.
Marco IS0KYB, which I believe came up first with defining a SNR measure, mentioned that looking at the difference between max and min values over 24 hour periods in such plots might give a better qualifier for the Kiwi receivers.
That this SNR measure is actually far from perfect is discussed in: https://forum.kiwisdr.com/index.php?p=/discussion/comment/14645#Comment_14645
PS: There is also a similar overview of on-line SDR receivers without indicating when the snapshots were taken at https://rx-tx.info/
Ben
In addition to Top Spotters site there is also http://www.jimlill.com:8088/index.html
You may wish to break your USA sites down by State
An interesting initiative, and anything that can be used to provide a better indication of the better KiWi's is to be encouraged.
There are lots of previous SNR discussions on this subject that can be found in this forum.
Some time ago I also asked the priyom folks behind the
rx.linkfanel.net.
SNR rating website to consider some different methods of evaluating the SNR performance
https://github.com/priyom/dyatlov/issues/5
But I never got any form of response from them.
I still believe that comparing day / night values is the best indicator, but looking at your waterfall snapshots. I wondered if it may somehow be possible to perform some sort of pattern matching or other visual processing, in order to pick out or compare good performing KiWi's.
Good luck with your initiative.
Regards,
Martin
I find a measurement of diurnal (propagated) noise variation to be the single best indication of station quality. It's influenced by latitude , MUF and solar cycle but does correlate well with WSPR spotting performance.
Looking for a single metric to identify a complex subject is probably always going to be fraught with problems.
Glenn n6gn
Thanks to all for the info but I don't want to pay too much attention to things like signal-to-noise scores etc. As an example: Number 1 in this top 200 is the SDR from VK3KHZ, Croydon, Victoria, Australia. Someone ignorant would think "he's on top so that's probably the best". But when they click on the link, they see, just like me, very black areas across the entire spectrum from 0 to 30 megahertz. I might then think that this top 200 is not entirely accurate.
I've also come across several SDRs that have very poor reception and aren't afraid to put a link at the top of their page for listeners to donate. I have to admit, I donated 25 euros once to an operator of a Kiwi sdr, namely Emerald SDR in Carlow, Ireland. This owner used the FX Megaloop of the NTi (Bonito) and was the only one in Europe to receive the Alpha or RSDN-20 signals from Russia.
The screenprints on my website are a snapshot. There is indeed a difference between reception during the day and at night. And then I leave out the other conditions such as rain, snowfall, thunderstorms.
Any help is welcome, if a Kiwi SDR operator sees that I haven't visited their station yet for whatever reason, there are a handful of sites where I get "error 404", "WSPR only" or "password only", then they can make a screen print themselves and send it to me and I will post it on my site. This screen print must contain the following information: station name, grid, city, used antenna and the date and time of the screenshot.
"Thanks to all for the info but I don't want to pay too much attention to things like signal-to-noise scores etc. As an example: Number 1 in this top 200 is the SDR from VK3KHZ, Croydon, Victoria, Australia. Someone ignorant would think "he's on top so that's probably the best". But when they click on the link, they see, just like me, very black areas across the entire spectrum from 0 to 30 megahertz. I might then think that this top 200 is not entirely accurate."
There is a bit of a problem with the current algorithm, in that if any frequencies are masked on a particular KiWi, the minimum signal level across that part of the spectrum is used as part of the SNR measurement, and the masked frequencies distort that value.
However it is generally fair to say that those KiWi's in the top 10% (typically when sorted less than position 50) on the list are generally pretty good.
Once I've found decent KiWi's in different parts of the world it tend to bookmark them anyway, and I also have a list of predefined URL's including good KiWI's that I use for TDoA runs.
As you have pointed out some KiWi's are particularly good for specific tasks such as VLF or NDB monitoring, even though they may not be much good on other frequencies. In which case it's a good idea to keep you own list of those, rather than the ones that are the best for say general HF band listening.
To be honest personally I'm not entirely sure in the value of having to plough through pages of screen shots to try and find a decent KiWi, when choosing one from say the top ten listed HF SNR scores would do pretty much the same job but faster. However I'm willing to be persuaded otherwise, and I hope your project evolves into something that is genuinely useful to the listening community.
Regards,
Martin
There is a bit of a problem with the current algorithm, in that if any frequencies are masked on a particular KiWi, the minimum signal level across that part of the spectrum is used as part of the SNR measurement, and the masked frequencies distort that value.
In the case of the SNR values shown on the Kiwi UI, and returned by the
/snrinterface, this hasn't been true for quite some time. Masked bins are simply disregarded and the number of bins considered in the SNR calculation smaller.Hi John,
Ah OK, I wasn't aware of this.
So in that case, I don't understand why KiWi's with large masked chunks of spectrum still get such good SNR scores.
Is it because the average is across a smaller number of bins that are included in the calculation, or is it some other factor ?
Regards,
Martin
Don't know. Looking carefully at the data from /snr might be helpful.
@jks
Thanks John for appending the measured SNR value to the Kiwi host name on the TDoA map display. Very helpful for quickly finding potential Kiwis that can be used for the process.
How does any SNR test differentiate a "real" signal from a spurious carrier?
That was a quick hack I added at Martin's request. It's the HF value computed by the periodic SNR measurement routine. If measurements are not enabled then no value is shown. The value also appears in the mouse-over popup for each sampling station.
For anyone wondering: this change isn't part of a general release yet. You have to manually rebuild to get it.
"How does any SNR test differentiate a "real" signal from a spurious carrier?"
It doesn't, but it's better than nothing.
My request was in order to save time and effort when selecting KiWi's from the map for TDoA runs.
In my experience any KiWi with an SNR of <=15dB can generally be ignored.
Regards,
Martin