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 05-12-2004, 18:17   #436
dk8cb
Roland
 
Join Date: Mar 2004
Location: Munich, Germany
Posts: 2,748
Hi,

a strong signal again.

Roland
Attached Images
File Type: png 7320_041205_dk8cb.png (26.8 KB, 142 views)
dk8cb is offline   Reply With Quote
Old 06-12-2004, 11:00   #437
dk8cb
Roland
 
Join Date: Mar 2004
Location: Munich, Germany
Posts: 2,748
Hi,

excellent reception with the exception of the first minute, when the well known P23S problem with no audio occured. OK after that.

Roland
Attached Images
File Type: png 7320_041206_dk8cb.png (26.1 KB, 144 views)
dk8cb is offline   Reply With Quote
Old 07-12-2004, 11:08   #438
dk8cb
Roland
 
Join Date: Mar 2004
Location: Munich, Germany
Posts: 2,748
Very good reception today.

Roland
Attached Images
File Type: png 7320_041207_dk8cb.png (26.5 KB, 160 views)
dk8cb is offline   Reply With Quote
Old 07-12-2004, 20:12   #439
simone
simone
 
simone's Avatar
 
Join Date: Jan 2003
Location: Bad Salzschlirf, Hessen, Germany
Posts: 10,144
Hi all,
not so good in my location, see attachment
73, Simone
Attached Images
File Type: png 7320_041207_simone.png (27.2 KB, 152 views)
simone is offline   Reply With Quote
Old 08-12-2004, 11:14   #440
dk8cb
Roland
 
Join Date: Mar 2004
Location: Munich, Germany
Posts: 2,748
100%

100%, as it should always be.

Roland
Attached Images
File Type: png 7320_041208_dk8cb.png (25.4 KB, 140 views)
dk8cb is offline   Reply With Quote
Old 08-12-2004, 15:39   #441
simone
simone
 
simone's Avatar
 
Join Date: Jan 2003
Location: Bad Salzschlirf, Hessen, Germany
Posts: 10,144
Hi all,
pretty good results today, see attachment

73, Simone
Attached Images
File Type: png 7320_041208_simone.png (26.0 KB, 171 views)
simone is offline   Reply With Quote
Old 08-12-2004, 21:09   #442
jbriggs
James Briggs
 
jbriggs's Avatar
 
Join Date: Aug 2002
Location: Somerset, UK
Posts: 148
Simone, Roland, notice anything different today? Have a close look at your attachments.
Roland, I am interested to see the difference between DReaM & DRM in your file: a bug?
James
jbriggs is offline   Reply With Quote
Old 08-12-2004, 21:18   #443
simone
simone
 
simone's Avatar
 
Join Date: Jan 2003
Location: Bad Salzschlirf, Hessen, Germany
Posts: 10,144
Hi James,
the bitrate was different to normal, but there is always the difference between DReaM and DRM thatīs the bitrate (80bps) for the text.
73, Simone
simone is offline   Reply With Quote
Old 08-12-2004, 21:34   #444
jbriggs
James Briggs
 
jbriggs's Avatar
 
Join Date: Aug 2002
Location: Somerset, UK
Posts: 148
Hi Simone, I did not know about the 80 bps difference: thanks. The reduction in bitrate is due to the use of UEP, so I am really interested to hear comments on any differences you can see/hear.
James
jbriggs is offline   Reply With Quote
Old 08-12-2004, 22:20   #445
dk8cb
Roland
 
Join Date: Mar 2004
Location: Munich, Germany
Posts: 2,748
Hi James,

for a discussion of the difference in indicated bitrate have a look here.

It is difficult to judge the effect of UEP under nearly perfect reception conditions such as today. I did not notice any dropouts or moments with reduced audio quality today.

This UEP test might have been interesting a few days ago, when quality of reception was among the worst ever encountered on this frequency. (But I rather prefer it the way it is now.) There are rumours that a different antenna was used for a few days. Can you give us some enlightenment on that?

Roland
dk8cb is offline   Reply With Quote
Old 09-12-2004, 07:39   #446
corrados
Registered User
 
corrados's Avatar
 
Join Date: Nov 2003
Location: Darmstadt
Posts: 10
Quote:
Originally posted by jbriggs
...The reduction in bitrate is due to the use of UEP, so I am really interested to hear comments on any differences you can see/hear.
James

Hi James,

I was listening to your UEP tests yesterday. I had a lot of dropouts so that the effect of "graceful degradation" should be clearly noticable. But it wasn't the case... The dropout behaviour was just the same as using EEP.

The reason for that could be that the length of the higher protected part of the audio block is too short so that it does not completely cover the block where the 8-bit CRC is calculated (unfortunately, this length varies with time depending on the audio material and AAC encoder algorithm but the length for UEP in DRM is fixed. So a certain "headroom" should be considered when setting the length).

Another thing might be that the difference in protection level was too little. You set the higher protected part to a rate of 0.5 and the lower protected part to a rate of 0.6 (correct me if I'm wrong...). It would be very interesting to see how the system performs if you would choose a rate of 0.5 for the higher protected part and a rate of 0.78 for the lower protected part so that the difference is larger.

I put a recording from yesterday on our homepage so you can try it out yourself. Since the file size is approx. 30 MB, I'll remove the file soon (it will be available at least until next weekend). Here is the direct link:
http://www.tu-darmstadt.de/fb/et/uet..._20_14kbps.wav

Volker
corrados is offline   Reply With Quote
Old 09-12-2004, 09:04   #447
ollieh
Registered User
 
Join Date: Dec 2004
Location: South London, UK
Posts: 1
Hi, Ollie here from BBC R&D. (My first ever posting).

Thanks for the responses so far to our UEP tests, and please keep them coming!

We did some laboratory tests of UEP, and found that we could get a modest improvement in degradation behaviour by protecting 20% of the audio data at the stronger rate, a parameter suggested by the audio coding experts.

It gets better if you protect more at the stronger rate, but then there is a significant reduction in the bit-rate, and so the audio quality is sacrificed for the listeners who have a good signal.

However, I'm aware that the proportion needing higher protection is programme-dependent, and it might be that the BBC World Service material would benefit from different settings to the material we used in our lab tests.

I suggest that we keep the parameters the same over the weekend, and then try increasing the part A length next week, to see if that improves things.

Our current feeling is that we would be reluctant to go any higher than 0.6 for part B in a shortwave channel. Going to 0.71 would mean that the audio would start to fail earlier, albeit with a wider region of "graceful degradation".

Thanks again for your helpful feedback.
ollieh is offline   Reply With Quote
Old 09-12-2004, 10:10   #448
dk8cb
Roland
 
Join Date: Mar 2004
Location: Munich, Germany
Posts: 2,748
Quote:
Originally posted by ollieh
It gets better if you protect more at the stronger rate, but then there is a significant reduction in the bit-rate, and so the audio quality is sacrificed for the listeners who have a good signal.

Hi Ollie,

to be able to contribute anything to these tests, I should perhaps sacrifice my good reception on the dipole antenna by using a small telescopic arial instead. Perhaps I'll do that tomorrow.
Today, I have perfect reception again.

But, all in all, I have to point out, that reception on this frequency is usually very good if there are no problems originating from the transmitting site, something I do notice much too often. The overall decoding result, when summed over many days could be much improved not so much by UEP but rather a lot more by putting out a PERFECT SIGNAL! This means that one should concentrate on efforts to solve the occasional transmitter audio feed dropout problems, problems of sometimes inaccurate transmitter tuning(?), and the old problem of skipping audio. Has the BBC already looked into the question, why the audio on this frequency is always delayed by half a minute, something that cannot be noticed on any other frequency of any other broadcaster, not even on Orfordness' 1296 kHz transmissions? Imagine someone setting his clock according to the GMT time signal.
All these problems have been there for a long time now and albeit occuring not as often as they used to, these are the problems that still limit the overall result (when integrated over a month or so) at least for my location and my equipment.
I'd really like to see more efforts in trying to get rid of these "old" but still unsolved problems.

It doesn't help, if reception quality could in theory be perfect but when the practical application of this theory fails.

Roland
dk8cb is offline   Reply With Quote
Old 09-12-2004, 11:17   #449
dk8cb
Roland
 
Join Date: Mar 2004
Location: Munich, Germany
Posts: 2,748
100%

Hi,

again perfect today.

Has the encoder software been udated? My subjective impression is that the SBR on the "Lily Bolero" shortly before the news sounds a bit better, but perhaps I still have the sound from 1296 kHz in my ears...

Roland
Attached Images
File Type: png 7320_041209_dk8cb.png (25.1 KB, 136 views)
dk8cb is offline   Reply With Quote
Old 09-12-2004, 11:56   #450
corrados
Registered User
 
corrados's Avatar
 
Join Date: Nov 2003
Location: Darmstadt
Posts: 10
Quote:
Originally posted by ollieh
...We did some laboratory tests of UEP, and found that we could get a modest improvement in degradation behaviour by protecting 20% of the audio data at the stronger rate, a parameter suggested by the audio coding experts...

I have attached a figure showing the number of CRC protected bytes in the AAC stream used in your UEP transmission on 7320. As you can see, a value of approx. 40 bytes for the higher protected part should be ok. How many bytes did you use?


Quote:
...It gets better if you protect more at the stronger rate, but then there is a significant reduction in the bit-rate, and so the audio quality is sacrificed for the listeners who have a good signal...Our current feeling is that we would be reluctant to go any higher than 0.6 for part B in a shortwave channel. Going to 0.71 would mean that the audio would start to fail earlier, albeit with a wider region of "graceful degradation"...

Ok, lengthening the higher protected part will decrease the bit rate but increasing the rate in the lower protected part will increase the overall bit-rate. So if one has a very good reception he will have the benefit of high quality audio. The one who has bad reception can still hear something (which might be still ok to listen to).

I would really like to hear how it sounds if UEP protection is used with a large difference between the higher and lower protected part rate in the case when reception is not so good.

Volker
corrados 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 17:39.


Powered by vBulletin Version 3.5.3
Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.