Twice every year, the High Frequency Coordination Conference (HFCC) is held to decide on frequency allocations for HF frequency users (i.e. shortwave broadcasters, utility stations, etc) for half of the year (i.e. schedule A or B, sometimes referred to as summer and winter). It is also an event for preeminent members of the broadcasting community to meet together and learn of the latest technologies, emerging issues or ideas and conduct demonstrations and experiments. The latest conference, B15, was held on 24th – 28th August 2015 in Brisbane, Australia.
While I wasn’t invited, and I had no reason to be invited, this event proved to be interesting to me because of broadcasting test transmissions from partnered broadcasters to the delegates at the conference. In the last few years, interest in DRM (Digital Radio Mondiale, or digital shortwave) broadcasting has picked up, as a way to transmit higher fidelity audio with multimedia and ID capabilities without relying on internet or satellite connectivity which may not be available to prospective listeners. This year, broadcasters were demonstrating various transmission mode parameters with different audio quality trade-offs.
Originally, I was not aware of the conference dates or of the test transmission until I was contacted by a member of the conference after previously having monitored their transmissions. They were requesting my help to monitor their transmissions, and in return, they were a great contact that gave me some information about the test transmissions that would be running during the conference. The information provided was as follows:
RNZI 0030-0100 UTC 17675 kHz RNZI 0200-0230 UTC 17675 kHz KTWR-Guam 0630-0730 UTC 15450 kHz Vatican Radio 2230-2300 UTC 11890 kHz
As RNZI is practically “across the pond” from us, their signals are very frequently seen (and sometimes heard, depending on the SNR) from Sydney. Looking at the two listed test transmission times, it seemed odd to me. I checked the regular RNZI schedule, which appears to be:
0255-0400 UTC 17675khz 0651-0758 UTC 11690khz 1551-1745 UTC 7330khz 1746-1950 UTC 11690khz 1951-2000 UTC 15720khz 2255-0200 UTC 17675khz
Paying close attention, the 0030-0100 schedule overlaps with the 2255-0200 regular transmission. The 0200-0230 schedule might make some sense, as it means they leave their transmitter on for a little longer after the end of their 2255-0200 transmission, but when I first monitored it, the carrier did not go down at 0230 as expected, and instead, sailed all the way down to 0458!
As the conference ended over a month ago, it seems they made a change to the website, and the actual broadcast schedule is available without log-in. A bit late, but it confirms my suspicion that the RNZI information provided was at least partially incorrect. Checking their schedule now, it seems that they opted to basically run their transmissions continually with some antenna bearing changes on 17675khz from 2151-0458, before going to 13730khz during 0459-0650, then to 9890khz through 0651-1258. Sadly, I wasn’t aware of these other frequencies, and from what I was seeing, I came to the incorrect conclusion that they were merely extending their regular transmissions.
The details for KTWR and Vatican Radio were correct, so I did put in a lot of effort to keep an eye out for their signals – but as usual, propagation is fickle, and the Vatican Radio was early morning, with KTWR at the late afternoon, both periods where propagation can change quickly. Worse still, the weather was not being forgiving, and local thunderstorms were wreaking havoc with reception at times.
Radio New Zealand International
My slackness in monitoring RNZI can be attributed to several reasons, not least the fact that they just seem to be “always around”, thus there is little novelty factor to it. They also don’t issue physical QSLs unless you pay for them, and I have received electronic QSL from them before.
Regardless, just leaving my radio on was enough to come up with something interesting. DRM signals in Australia are quite hard to get, especially in suburban areas. Even with my Wellbrook loop, I’ve rarely ever seen the SNR eclipse 20dB, nor have SNR above 15dB for extended periods of time. Audio has always been choppy, although a minute of good audio sometimes does make it through. It comes down to deep fading due to the long paths, and other signal corrupting effects (smearing of impulses from multipath/doppler).
Just leaving the decoder on, at 0300UTC, I was able to record a record 23.7dB SNR from RNZI for a very short period as the multipath became “additive” rather than destructive. But you can see from the SNR trend just how painfully difficult it is to get any audio whatsoever.
Radio Vaticana is a station which I’ve received stable DRM from before, partly because of their use of a more “stable” 16QAM service mode encoding at the time. It’s also a station which I’ve gratefully received QSL cards from.
Their transmission was interesting, as it was clearly targeted to the conference in the radio text message. They did change modes on occasion, mid-way through the broadcast, to assess the effects. In the more robust mode, they were using Mode C, 4QAM for SDC, 16QAM for MSC, with 9.18kbit/s for audio in regular aac, resulting in not-so-clear audio. A slightly less robust mode using Mode B, with 11.64kbit/s in regular aac was also tested which sounded markedly better. The less robust mode was 16QAM for SDC, 64QAM for MSC, with 13.78kbit/s for audio in aac+ but audio was only very rarely received.
Some of the received audio can be heard in this cut sample where some intermediate sentences had been cut due to DREAM decoder issue causing stuttering. The SNR trend plots in the appendix at the end show just how difficult and fickle propagation can be. Unfortunately, I didn’t record them every day of transmission – the last day was so poor that the decoder wasn’t locking at all (0-4dB SNR), so a plot wasn’t created. But many days, the SNR was 8-12dB.
I dutifully filed my reception reports, and true to their word, I received a QSL for this special transmission.
I also received some collectibles – a few stickers and a fridge magnet. My thanks to Sergio Salvatori for this wonderful gift!
Trans-World Radio, Guam (KTWR)
Of course, there was another station running trials at the time – that of Trans-World Radio. This is another station I had previously received a QSL card from, but in the AM mode. They never broadcasted DRM before, until this trial.
Their service was received on Monday to Thursday, with nothing heard on Friday. Reception was patchy, with very little audio on Thursday especially. They did try several modes with different bitrates and capabilities – the above was on the last day of transmission where they were audacious enough to try aac+ parametric stereo and a journaline style multimedia service. Sadly, reception stability wasn’t enough to see the MM part of the transmission.
Received text included “Hello and Welcome to the TWR-Guam DRM transmission”, “We demonstrate the power of the Ampegon Content Server”, “With latest technology and features”, “Hope you have best reception and enjouy the demo”, “Thank you very much”, “Your TWR-Guam team”, “We wish you a nice week”.
Generally, the test DRM station received with SNR varying between 10dB to 20dB due to deep fading. Only fragmented audio was received, the percentage varying by day, although FAC and SDC decodes were relatively solid.
- Mon: Mode C, 16QAM SDC, 64QAM MSC, PL 1. 16.52kbit/s AAC+ Mono
- Tue/Wed: Mode B, 16QAM SDC, 64QAM MSC, PL 1. 20.96kbit/s AAC+ P-Stereo
- Thur: Mode B, 16QAM SDC, 64QAM MSC, PL 1. 19.04kbit/s AAC+ P-Stereo + MM
They really took full advantage of the DRM capabilities, broadcasting a special program with music and voice segments to exploit the abilities of AAC+ with parametric stereo. For example, here is their identification at the end of the broadcast, and a music segment. You can hear what happens when packets aren’t decoded properly – where it affects the SBR data, the treble is lost and only the bottom 0-4.5khz “base” encoded audio is decoded. You can’t get that sort of quality via regular analog! That being said, reception was exceptionally difficult to achieve stably, the trend plots are shown in the Appendix.
I dutifully reported my reception via their website at twr.asia using their Reception Report form (quite extensive), but to my dismay, if we leave detailed program information, the form seems to get too big and “hangs” at submission. I tried several times without knowing for sure whether the reports were submitted.
As it turns out, they had little problem receiving (most of) my reception reports, and this arrived just a few days ago.
Thanks to Kathy Gregowske for this QSL letter – one of my first QSL letters, in addition to a QSL card as below.
The card is very much appropriate, given the president of TWR is talking about DRM transmissions. I also like how the card gives you some indication of a challenge you can set yourself to collect them all.
As is traditional with QSL mail-outs, a program schedule was attached, although it seems they have not run any regular DRM transmissions yet. I hope their next schedule has some DRM on it!
There is also a leaflet about the power of religion. Indeed, a lot of media is used to spread religion, but I’ll leave the discussion at that. I’m very much more interested in the technical side of things.
The End? Not Quite
So there were three test transmission broadcasters, so that’s the end of the post, right? Not quite. As it turns out, the Voice of America was promoting their VoA Radiogram service at the conference, and I stumbled across their leaflet in the online files. The service utilizes regular AM transmitters to send MFSK32/Olivia “ham digimodes” as the service, thus allowing inexpensive non-SSB capable radios to receive and decode text and images.
Because of the error correcting capabilities of digimodes, and the coding gain, it becomes possible to decode readable text and images where the signal to noise ratio is down to even -10dB in some cases. This means the noise is stronger than the signal by 10-fold. Indeed, in Australia, the reception of VoA radiogram (that weekend) was hovering between -4dB to -10dB SNR, so an audio program would have had no chance to make it.
While I haven’t received a QSL card as such from them, I have been in contact with Dr. Kim Andrew Elliott in sending routine reception reports for their programs and some of my reports have been featured as tweets on VoA Radiogram’s Twitter account.
— VOA Radiogram (@VOARadiogram) September 19, 2015
— VOA Radiogram (@VOARadiogram) September 12, 2015
It’s quite exciting to think that an HFCC conference was held in an adjacent state, about a month ago, and broadcasters were putting their efforts to sending shortwave test broadcasts our way. Australia generally has a smaller shortwave listenership than other countries, and there are few strong signals that are beamed here regularly. Any new signal can be quite an exciting thing to chase down and receive. I certainly did quite a bit of work in turning my loop from time to time to try and optimize the signal, but the harsh QRN and the QRM from the neighbours preparing for work in the morning/returning home in the evening made things challenging.
Regardless, I thank the broadcasters for their timely QSL card issuance, and I definitely enjoyed being a part of your test broadcasts. Next time, maybe the HFCC should let us radio listeners know, so that they can extract the maximum benefit from running those test broadcasts.
Appendix: Vatican Radio SNR Trend Plots
24th August 2015
25th August 2015
26th August 2015