Digital ATV standards

[g8ajn]

Greetings from Bournemouth! As you may be aware we are currently final (?) testing the 'SQ digital repeater for the Poole, Bournemouth and Christchurch area. Having improved the analog input reception it has been disappointing to see the 'soft' low-res pictures received on the digital output using 4Ms/s. We have decided to try moving to 6Ms/s on a regular basis. Whilst some of this may be down to the commercial receivers which are probably happier at high data rates,it is hardly a profligate use of bandwidth, and should allow a reasonable 3Mhz video to be received. I notice that other groups, such as GB3VR, have now moved to 6Ms/s for the same reason.
It occurred that we should be agreeing a national standard by now, to at least give a chance of resolving a dx digital signal. Common settings(easy to remember PIDs?) for any digital stations accessing digital repeaters should make initial contacts easier. I dont propose an enforced set of parameters for all digital signals, only a base guideline for repeaters, which should be able to re-transmit the incoming signals at something like the original input quality.
Another point is that with the likely imminent bandplan changes, it may be better not to be seen coping too well with a low bandwidth tv signal, or we may lose even more band space!
Your thoughts and comments would be much appreciated.
Dave, G8AJN
g8ajn.tv@gmail.com
http://www.gb3sq.org.uk

[M0DTS]

Interesting ideas..

I have not played with anything over 3Ms/s at 1/2 here yet but did see some 6Ms/s from Terry G1LPS some time back and it was very sharp picture so i can see the advantages. I would try a test run with 6Ms for the next few weeks and see if it stops anyone seeing the repeater.

If there was a list on the forum somewhere that shows all of the UK digital repeater settings it would be just as good i guess... as you would never be sure the dx repeater was using standard settings or not...
Maybe it is better to have a slightly less sharp picture and more SNR for the dx stations... there's always a compromise!

There has only been one dx report of a digital repeater (HV) as far as i know of... hope SQ is the next.

73

Rob

[G4GUO]

Being Selfish and proving the back row is listening....

Generating DATV using a SDR that has a fixed final D/A sample rate
I have to use interpolation/decimation to get the required final
sample rate so if there were only a few symbol rates say 2,4,6 M Symbols/s
I would be very happy. Please no weird telecoms rates like 2.048M!

As far as FEC rate is concerned I just have to calculate the
video/audio bitrates for each different Symbol rate / FEC puncture mask
combination so that is less of an issue for me.

Shame your repeater is on the wrong side of Portsmouth and I am so low down.

- Charles G4GUO

[G8ADM]

Hi
Here in GB3BH and GB3HV land in the Home Counties we have standardized on the following:

70cm 2 M Symbols / sec. FEC = 3/4 Frequency = 436.00 MHz. 1 stereo audio pair.
(This uses a bandwidth of about 2 MHz)

All other bands:
4 M Symbols / sec. FEC = 3/4. Audio pairs as available.

I am not sure why you have poor resolution, many broadcast channels only use 4 M S/s

Regards,

Dave.
G8adm

[G4EWJ]

As well as standard SR / FEC, I think it would be useful to have a standard set of PIDs.

Brian

[g8ajn]

Thanks to all above who took the trouble to reply. If I can answer a couple of points made; Rob M0DTS has the right idea of a detailed list. He is currently running an excellent site at http://www.M0DTS.co.uk . I believe that the RSGB or BATC are best places for such a list. Current the Repeater pages don't even include basics such as erp etc.
BATC President Peter G3PYB has agreed to raise the matter in the future. I will be running some comparative tests locally to see if there is really is any visible advantage/disadvantage of using 6Ms rather than 4Ms. I will report back here later with the results.
Charles G4GUO is clearly well embedded into digital and the good news is that the Isle of Wight repeater GB3IV has now been upgraded to include digital signals and does cover Portsmouth and Southampton areas. More details on their website http://www.scartclub.co.uk . With the planned 3cm link from 'IV to Bournemouth we should be able to QSO before too long!
Dave G8ADM brings up the valid point that commercial broadcasters often use 4Ms, I like the 3/4 FEC option, we will have to try that one here I think.I will go through all the encoder settings and see if I can improve the results. Has anyone actually tried all these different parameters on a repeater and written down the comparative results anywhere?
Yes, Brian,G4EWJ, I think it would be helpful as the various repeaters across the country move to 'digital out' to have guideline parameters of FEC,Ms/s,PID,IBP etc. Bearing in mind that I am discussing 23cms (and up) and particularly Repeater inputs and outputs.
I am starting to try various settings on the SR Systems encoder/modulator and will attempt to get a decent digital picture that will do justice to the excellent analog pictures we get into repeater. I will report back here later on when we have had time to evaluate the different settings.
Thanks to everyone for their comments, they are most welcome.
73 Dave G8AJN

[g8ajn]

Here is an update following tests of various digital parameters as described above. We initially tested with 4Ms/s and 1/2 FEC. Our tests stations were at Wimborne and Ringwood at about 12km and one nearer station that is in a dip and gets only a p3 signal into the repeater. Using a good quality DVD as the video source and with fast moving picture changes (motor sport), there was a tiny amount of hiccuping on fast action, but no pixelating was visible. Only when the signal strength was dropped a few dB did pixelating start to occur. Picture quality was quite good overall on 'low detail' pictures but the colour resolution was soft and gave a low resolution colour fog at the edges of the mono parts of the picture, giving the whole picture a 'VHS' look.
We then moved to 6Ms/s still with 1/2 FEC and ran the same clips again. Picture quality improved markedly, the whole dynamic range seemed better, with the colour now better defined.
No movement/pixel problems, but a drop in received signal level was noticed of about 1 or 2%. This would seem reasonable given the extra bandwidth involved.
We then moved to 2/3 FEC but remaining with 6Ms/s. No signal level change was noticed compared to 1/2 but with a power drop of ~ 2dB there was some occasional brief freezing reported from several of the test stations.
Used as a repeater the quality of the re-transmitted analog signals is very good (even noticeable on the streaming pictures)and really does do justice to the quality of the incoming analog signals.
Some of these differences may well be down to the types of receivers used being poor at handling the lower rates and possibly the encoder too, the final settings of which I will list here later for those newcomers to DATV who are unsure what settings to use for their SR Systems encoder which comes with a rather unhelpful instruction sheet. Bearing in mind that we would need up to four times this bandwidth to get comparable pictures from an analog repeater, we are maintaining a reasonable quality level and giving a faithful representation of the incoming signal.
Extra detail is important at times when a piece of equipment is being discussed, or a circuit diagram or even text is being shown and as I have said before,to me a repeater is just that.. a repeater. Not a quality squasher. Not digital because we can do it, but because it is capable of near analog quality with a fraction of the precious bandwidth.
( we briefly tested 8Ms/s and were rather relieved to see no discernable improvement over the 6Ms/s!)
I do understand that there might be repeaters in more widely scattered districts where the extra output from using 4Ms/s may be preferred to the better quality of higher data rates. Hopefully other digital repeaters as they come on will be able to try these tests for themselves and decide accordingly which is best for them.
By using extra care with the video quality input to the modulator (SR Systems) and settings on the SR parameters, quality is just about right with 4Ms/s and 3/4 FEC. Take care with the SR setup, for example HD stands for half-definition!
Finally thanks to everyone for their comments and help.
Best wishes
Dave G8AJN