DRM Software Radio Forums  

Go Back   DRM Software Radio Forums > DRM Software Radio - User Forums > Reception Results - Recent Posts
User Name
Password
FAQ Members List Calendar Search Today's Posts Mark Forums Read

Reply
 
Thread Tools Search this Thread Display Modes
Old 15-06-2018, 13:21   #61
PhilipOneL
Registered User
 
Join Date: Oct 2014
Location: Eastern Newfoundland
Posts: 192
Friday morning, 15/6/18, in downtown St. John's, Newfoundland. I finally heard some audio from WINB DRM! I had turned on both my MorphyRichards 27024 (with an 8m wire outside) and my Avion DR1401 (with a Sony AN-LP1 loop antenna inside the house, pointing SW). I was looking at the two radios' responses to the WINB signal. On the MR it varied between 0 dB MER and 16 dB. On the Avion it was from zero to 12 dB.

Suddenly a burst of audio came out of one of them. It only lasted about a half-second, long enough to recognise a child's or a woman's voice in what sounded like acted anguish. Then it was gone.

I had the two radios stacked at that point and, to be honest, I don't know which one produced the audio!

But I do know that one of them *can* decode whatever WINB is producing, at least sometimes.

A day or two ago, I had noticed that the AM-decoded audio on 15670 is a complex tone rather than the usual audio whoosh of DRM. Right now, at 1155 UTC, I hear the tone-audio (on my Satellit 750), while seeing low MER rates and the WINB label on the two DRM sets.

--------------
EDIT, five minutes later.

At 1202, the Avion leapt into audio action and produced about twenty seconds of audio from the WINB DRM signal. It does work. . .

Last edited by PhilipOneL : 15-06-2018 at 15:25.
PhilipOneL is offline   Reply With Quote
Old 15-06-2018, 13:35   #62
PhilipOneL
Registered User
 
Join Date: Oct 2014
Location: Eastern Newfoundland
Posts: 192
By the way, *very* momentarily, a few minutes after that audio spurt, the Avion showed a number 2 in its programme indicator, suggesting it was picking up a second programme, at least for a moment (well under a second). I wonder if this was simply a glitch in the Avion, or an indication of a data stream.
PhilipOneL is offline   Reply With Quote
Old 15-06-2018, 15:00   #63
AF4MP
Registered User
 
Join Date: Apr 2006
Location: Roswell, Georgia, USA
Posts: 9,351
Quote:
Originally Posted by PhilipOneL
I wonder if this was simply a glitch in the Avion, or an indication of a data stream.

Hi Philip,

It was probably a momentary glitch, because the data signal that WINB is simulcasting is not part of the actual DRM encoding.

As mentioned before the technical specifications of DRM do not recommend simulcast with DRM on HF.

So far, judging from all reports, the WINB DRM reception is marginal at best.
__________________
73,

-Zyg- AF4MP
3402' N, 8424' W
EM74ta
AF4MP is offline   Reply With Quote
Old 15-06-2018, 15:52   #64
PhilipOneL
Registered User
 
Join Date: Oct 2014
Location: Eastern Newfoundland
Posts: 192
It is just coming up to 1430 Friday (15/6/18) now and I've been able to sit by the radio most of the past three hours, catching little snippets of speech from the WINB DRM on 15670 kHz. None of the snippets lasted more than twenty seconds and they are separated by ten and fifteen-minute segments of DRM silence, so it's not exactly a great listening experience. . . .

I am struck, once again, by how poor this DRM audio is. It does not have to be so, because I do hear good sound from, say, RRI. But it often seems highly compressed, making it difficult to pick out the speech when someone's accent is different from my own. It sometimes seems slewed towards bass frequencies (like V Nigeria) but here, with WINB, it seems EQ'd towards a tinny, higher-frequency sound.

I'd like to be able to analyse that audio to figure out what they are doing. I have an old but well-functioning copy of Sony Sound Forge. It can do some spectrographic analysis, but not precisely what I would like. Does anyone have anything to suggest for analysis of broadcast sound to be able to make clear comment on the EQ?
PhilipOneL is offline   Reply With Quote
Old 15-06-2018, 17:41   #65
tpreitzel
Registered User
 
Join Date: Apr 2012
Posts: 2,341
Quote:
Originally Posted by MARCEL
Bonjour tpreitzel,

Sorry, but the problem it is my English which is very bad, I do as I can, thus the I use a translator.
I decoded WINB from 11:42 am UTC approximately 10 at 13 minutes, with more or less long cuts of in the important fading, from 12:20 am the signal begins to fall and at 12:21 am more by audio,

At 12:12 am the emission was in Spanish language,

Marcel,

I know that English can be difficult so I thank you for your explanation. Your English is far, FAR better than my French.

Merci
tpreitzel is offline   Reply With Quote
Old 15-06-2018, 17:44   #66
tpreitzel
Registered User
 
Join Date: Apr 2012
Posts: 2,341
Jun 15th, 2018 partial log ...

Codec: -
Modulation: -
Frequency: +
Attached Images
File Type: png 15670-180615-1340-1530-TPR.png (40.6 KB, 51 views)
tpreitzel is offline   Reply With Quote
Old 15-06-2018, 17:46   #67
AF4MP
Registered User
 
Join Date: Apr 2006
Location: Roswell, Georgia, USA
Posts: 9,351
Quote:
Originally Posted by PhilipOneL
I am struck, once again, by how poor this DRM audio is.

I checked the audio bandwidth the other day and it was only 3 kHz.

WINB is running their Simulcast DRM service in half the normal bandwidth at 5kHz using 64-QAM in the most robust mode..

This only allows 8.28 kbps of DRM data. Even if they switched to 16-QAM with the same robust signal coding then the available DRM data would only be 5.5 kbps and the resulting audio would be far more limited. (Reference: EBU Tech 3330 page 15).

However, as the main aim of WINB is data transmission and not the DRM broadcast, we should not be surprised!
__________________
73,

-Zyg- AF4MP
3402' N, 8424' W
EM74ta
AF4MP is offline   Reply With Quote
Old 15-06-2018, 17:53   #68
tpreitzel
Registered User
 
Join Date: Apr 2012
Posts: 2,341
Quote:
Originally Posted by PhilipOneL

I'd like to be able to analyse that audio to figure out what they are doing. I have an old but well-functioning copy of Sony Sound Forge. It can do some spectrographic analysis, but not precisely what I would like. Does anyone have anything to suggest for analysis of broadcast sound to be able to make clear comment on the EQ?

Although quite complex, you might try Sonic Visualiser or Audacity for Linux. Hopefully, other members can recommend something similar for Windows.

Yeah, the quality of audio is poor due to the poor choice of 64 QAM modulation at a low bit-rate. The use of higher bit-rates for HE-AAC+ like RRI renders decoding unreliable. The HE-AAC+ codec is really only viable on the MW band, but can be used on SW reliably with 16 QAM modulation. The muffled audio is probably due to the lack of SBR, but can be due to poor production, e.g. poor quality of the remotes in VoN's broadcast. The use of 16 QAM modulation would lower the threshold for decoding while still allowing sufficient robustness and quality, but this fact has been known for more than a decade.

I'll forgo any opinion at the moment on WINB's potential application for this broadcast, but 64 QAM modulation is certainly NOT the right choice for reliable decoding of audio in ANY configuration on shortwave bands.
To achieve a good balance between quality and reliability on the SW bands, digital broadcasters in 2018 should certainly be using 16 QAM modulation WITH the xHE-AAC codec. I should add that even with the latter combination, a broadcaster can even screw-up that excellent combination with poor production.

Last edited by tpreitzel : 15-06-2018 at 20:12.
tpreitzel is offline   Reply With Quote
Old 16-06-2018, 02:18   #69
PhilipOneL
Registered User
 
Join Date: Oct 2014
Location: Eastern Newfoundland
Posts: 192
TP, thanks for those suggestions. Forty-odd years ago, I used to use a (then-old!) Kay Spectrograph for analysing short speech segments. It was very limited but very elegant in its output, on little sheets of coated and point-burnt paper. I have never seen software quite as simple and useful. But Sonic Visualiser may be exactly the thing I was looking for.
PhilipOneL is offline   Reply With Quote
Old 17-06-2018, 20:52   #70
MARCEL
Registered User
 
Join Date: Mar 2010
Location: Northeast France
Posts: 351
Audio extract from WINB June 14.18 :

Voir le Fichier : DRM_WINB_14618_15670_11H42.mp3
__________________
DRM:
Mixer Platine MF9 Mhz with Dream

Antenna:
Mantova Turbo 5/8 wave for 27 mhz
Long Wire 80 meter

Last edited by MARCEL : 17-06-2018 at 21:04.
MARCEL is offline   Reply With Quote
Old 18-06-2018, 07:45   #71
Digger
Registered User
 
Digger's Avatar
 
Join Date: Mar 2005
Location: Shizuoka City, Japan
Posts: 9,652
Thanks for the WINB 8,28 kbps DRM sound sample, Marcel. It does not sound *so* terrible after all. I recorded the audio with Audacity.

And thanks for the TDA 891 kHz 12 kHz IF file from April last year. I played it on one PC and modulated a signal generator set to f kHz, fed the RF to the GR-216 which was tuned to f+12 kHz. I recorded the xHE-AAC audio (JIL FM, Ch.2 with 6.94 kbps, sample rate 32 kHz, No SBR) on a notebook with Audacity installed.

Enclosed are two screenshots from Audacity, one of WINB and the other from TDA with xHE-AAC. There is a significant increase in audio band width with the xHE-AAC signal.
Attached Images
File Type: jpg WINB_DRM_8k8.jpg (37.0 KB, 67 views)
File Type: jpg TDA-JIL-FM_DRM_xHE-AAC6k94.jpg (45.0 KB, 70 views)
__________________
Regards,
Terje

http://www.hobbyradio.se/en/drm/webp....index_en.html
Digger is offline   Reply With Quote
Old 18-06-2018, 16:25   #72
AF4MP
Registered User
 
Join Date: Apr 2006
Location: Roswell, Georgia, USA
Posts: 9,351
Attached is the Dream audio spectrum display received via the KiwiSDR receiver in Weiser, Idaho, USA just minutes ago.

The audio does not remain for more than a couple of seconds or so and then disappears.

The audio quality, when its there (as I had mentioned before) is "quite good".
Attached Images
File Type: png WINB Idaho Audio BW.png (8.0 KB, 70 views)
__________________
73,

-Zyg- AF4MP
3402' N, 8424' W
EM74ta
AF4MP is offline   Reply With Quote
Old 18-06-2018, 16:52   #73
AF4MP
Registered User
 
Join Date: Apr 2006
Location: Roswell, Georgia, USA
Posts: 9,351
WINB with FLdigi

Quite easy to look at the WINB datacasting portion of the signal with FLdigi (see Attachment - small size because of Forum's file size restriction).

The data signal is there even when the DRM is not.

You cannot, of course, know what type of data signal is being used.
Attached Images
File Type: png WINB FLdigi.png (67.8 KB, 87 views)
__________________
73,

-Zyg- AF4MP
3402' N, 8424' W
EM74ta
AF4MP is offline   Reply With Quote
Old 21-06-2018, 20:17   #74
tpreitzel
Registered User
 
Join Date: Apr 2012
Posts: 2,341
Jun 21st, 2018 partial log ...


Frequency: +
Codec: -
Modulation: -
Attached Images
File Type: png 15670-180621-1400-1600-TPR.png (37.5 KB, 58 views)
tpreitzel is offline   Reply With Quote
Old 22-06-2018, 21:11   #75
tpreitzel
Registered User
 
Join Date: Apr 2012
Posts: 2,341
Jun 22nd, 2018 partial log ... a bit longer than usual.


Frequency: +
Codec: -
Modulation: -
Attached Images
File Type: png 15670-180622-1359-1653-TPR.png (41.3 KB, 45 views)

Last edited by tpreitzel : 25-06-2018 at 20:12.
tpreitzel is offline   Reply With Quote
Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is Off
HTML code is Off
Forum Jump


All times are GMT. The time now is 11:22.


Powered by vBulletin Version 3.5.3
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.