Re: Upcoming DVB-T channel changes for HH (Hamburg)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

On Wed, 28 Jan 2009, Tobias Stoeber wrote:
> > > should (ideally) use a 8 MHz width space from 559.25 MHz to 567.25 MHz for> > > Ch 32.> > > > Is this correct, or should the range be from 558 to 566MHz,> > apart from locations (such as the UK and Australia) where an> > offset may be used?> > Well, you may be right ... I recalled that from former norms (analogue) and> the fact, that the digital channels were expected to use the existing> boundaries.
Well, the thing is, that where I came from and where I wasworking as a broadcast technician many decades ago, where agrandfathered norm was in use that makes a good PAL signalappear to be practically high-definition in comparison...
The reason for the offset was due to the use of analoguemodulation on the video carrier, which causes a sidebandto be created on either side of this carrier, correspondingto the frequency of the modulating signal.
For simplicity, I'll say that this blurry-o-vision wouldhave a bandwidth of about 3MHz, with some vague colour infosquezed above this, and then a separate audio carrier withan offset of 4,5MHz from the video carrier.
So, *very* roughly, in ASCII graphics, the frequency spectrumof the modulated carrier could be seen like this...      _____ I _____     /     |I|     \    ignoring the colour and sound.  __/      \I/      \__   F-3      F      F+3
This would require twice the bandwidth of the actualinformation, or 6MHz for 3MHz baseband video, and 9MHzwhen you add in the sound.
Strictly seen, the carrier F - modulation frequency fcarries the same information as F + f, so one cantransmit only the upper sideband without losing anyinformation, and filter out the lower sideband.
Because the filter is not perfect, and I've long sincedrank the precise values from memory, there still willremain a bit of the lower sideband...
          _ I _____         / |I|     \  ______|  \I/      \__   F-3      F      F+3
Thus the 1,25MHz offset of F from the lower bound ofthe frequency range.
And the upper bound will not be 8MHz higher, as that'sthe step to the next carrier frequency here.  Those1,25MHz belong to the leftovers of the lower sidebandof the next channel up.  So the maximum availablebandwidth will be based on 8MHz minus the 1,25MHz, minusa bit for additional services -- I've never graspedthe details of the differing norms for audio, as I'venever had the need, only having picked up the aboveas a side-effect from my work outside of television.

>> it seems, that you are correct (or how to you read the info in the pdf?).
I'll have to get to this later with a better netconnection...  But my basic understanding is thatthe CODFM signal can be vaguely seen as some 8192carriers packed within that 8MHz bandwidth (or 2kin early-adopter parts of the UK), and there isnot the waste of a leftover bit of sideband thatneeds the 1,25MHz offset -- instead you could viewthe carrier as the center of the 8MHz range, as isdone.
Now there is no guarantee that any of this is right,as I haven't attempted to absorb the multitude ofinformation I've come across to reach the `Aha!'moment of enlightenment that I need.

> > Maybe.  It's better, in my mind, than the existing case of> > individual sites, which again, may or may not cover the case> > of nearby areas.> > Well, the old style of de-transmitter_region scan file had the charme, that it> is easier (at least for me) to select the transmitter sites in my direct> surrounding and I've no real use of de-federal_state.
You also have the advantage of being familiar with the sitesyou can receive, as well as your local geography.  If I wereto suddenly be plopped into your general area, I'd have todig out a map to try and see what might be nearby.  I'm anoutsider and don't have the background of a native, exceptin a few places where I've spent more time than I shouldhave.  I've only once passed through Braunschweig on theway to Danmark from the sunny south, and it won't be untilI take the time to do the research, thus familiarisingmyself with a region which does not immediately bring tomind Weisswurst or Spätzle, that the individual site files(when they even exist, or when they are even accurate)would be as useful to me.
Take a look at fr-*.  That's (in my out-of-date mirror)over 100 sites.  One reason this is needed is due to thefact that they chose not to make use of Single-FrequencyNetworks, and so nearby towns are assigned differentfrequencies, rather than the reuse that I've noted in mycomments for B-W.
Counting the number of ZDF sites on the ten pages ofteletext, I come up with over 130 transmitters, most ofwhich are shared with the ARD and others.  But this isnot always true, as there are a handful of sites, mostin edge cases, where only the ARD/Dritte stations arebeing sent, so I'll estimate that Christoph will needsome 130 to 140 different files just to cover all thesites, or regions if you prefer.
Then there is the problem that the sites do not alwaysmatch a large metro area, and unless you're working asa TV installer, chances are you're not going to knowthe names and coördinates of the actual transmitters,though it is far better than the analogue case ofseveral thousand sites, including filler/repeatertransmitters...

> Take a large geographical area like Niedersachsen, so you come to the> conclusion, that in most areas you won't need the overwhelming majority of> entries in de-Niedersachsen (because transmitters are to far away), but you> will in most cases need entries from the [regional?] [ de-federal_state ]files.
I presume you meant to say you'd need regional sitefiles.
It's a tradeoff.  Do you want the 32 de-* files in myout-of-date mirror, including one obviously-wrongfile, the regions Ostbayern and Ruhrgebiet, or do youwant maybe around five times as many, but up-to-dateand correct, or do you want, say, 20 files, with thelargest cities and separate or perhaps combinedBundesländer, automagically generated?  Or would yourather have about 20 files total, with the Bundeslandfiles at least partly massaged by me to add more info,and applicable out-of-area info?
The thing which may not be immediately obvious is howa SFN simplifies combined lists.  The B-W file youreviewed replaces 13 (soon 14) individual region files,most with three frequencies.  This is, however, cutto a fifteen-line scanfile, rather than 39 total lines.Depending on area, one could need to pull in threeor four site files, and that's not even includingout-of-area broadcasts -- Grünten, as eventually youcan see from its elevation and mast height, blastsinto areas one might not expect.

> > I've posted in the past my suggestion for a de-BW file, made> > by hand, which tries to address this issue, as well as provide> > an overview for anyone trying to make sense of the frequencies> > and broadcast policies, as well as to help with antenna> > orientation, towerspotting, or anything else that might interest> > me, in a single location.
> Well, I had to look in may list archive, but did find it (your posting from 02> Dec 2008). First, I looks like a lot of wrok, especially because  this area> not only benefits from other federal staes but also from France and> Switzerland.
Thanks for looking at it.  It could be seen as true, that itwould be a lot of work for someone coming from outside andbeing told to do it, but in reality, I was already doingthis work, or had been, but into about five or six differenttext files on three different computers.
Putting all the info into one place was a way to make it lesswork for me, in the long run.
The thing is, I'm an outsider.  Just as others have theirown reasons for following linux development, there are anumber of things I like to do without considering the amountof work, just because I'm interested, and these often arerelated to technology and other technical matters.  Not tomention my background of climbing antenna masts to cook mybrain, and guarantee I'll never be the father of my children.
As I was not familiar with the details that are made availablevia teletext (a convenient offline medium when I haven'thad Internet access for remaining somewhat with-it), Ineeded to do research to answer the questions I had, whichraised more questions, and on and on...  This includedtrying to locate the antenna sites by name on printedmaps, then becoming intrigued by the alien (to me) conceptof SFNs, with my analogue background.
Anyway, the result was that I had made a bunch of differentsets of notes, in the unlikely event that someone mightconsider hiring me for my areas of interest, but I thoughtit might interest others in case they had the same view,as an outsider, if I made it public -- particularly asthe BaWü scanfiles were largely missing, or inaccurate.
So why not try to transfer my mental overview into amore permanent form, that I won't be able to drink awayso easily, and to see if I actually have a clue as towhat I think I know (which failed miserably in mydisinterested-outsider overview of german DVB-T  ;-)
The one thing I'd really like to include would be asimple ASCII map that one can use as orientation fortransmitter locations.  In the absence of this, I'vetried to partner each site with a somewhat-large nearbyDorf that can be seen on most maps that use most of apage to cover the Bundesland -- which has not always metmy satisfaction, as an outsider, intimately familiarwith areas I've bicycled through, but not always havinghad years of familiarity with, say, the PTB or otherlocations that immediately ring bells with natives.

The thing is, the benefit from other areas is a resultof also understanding those areas.  That is, I addedBayern without really understanding it, simply becausea BR coverage map pointed me at potential overlap.Work to get an overview of Bayern will come later.  Thatneeds an overview of part of Austria (though info fromBayerischer Rundfunk helps, but, um, do signals fromLinz really reach into Bayern?) where the switchover iswell underway, and the Czech Republic, which ispresently in the process of switching, but has notreached the final step as in Germany.

Actually, the details of Switzerland may not be so easy,but the overall view today is easy to describe in a fewwords, quite the opposite of Germany.  I think I didthat in the file:  one multiplex, if you're around here,this frequency, and a bonus DVB-H to go with it, and noplans to make further use of the eight or so UHF GE06frequencies available in each area that I've heard of.
France was easy too, because I had done my homeworkin the hopes of receiving these channels when making aholiday in the Kaiserstuhl, but it was too early forthat.

> > A Bundesland-based set of files is a region-based set, or can> > you better describe the regions you are thinking of?  In any
> I actually meant scan files for specific DVB-T region like de-Berlin,> de-Munich, de-Nuernberg etc.
I will argue that Berlin (also being its Bundesland) shouldremain, and likewise Muenchen, for the poor out-of-area soulswho never make it anywhere else, and who would stare blanklyat me if I told them they were in Bayern, dude, you know, Ba-vay-ree-yah.  There are things to see.
I'll need to make my overview of Bayern before I can suggesthow to handle Nuernberg.  It's an early adopter, needs someexplanation of the reception situation and such, and so farI really don't have a good idea...

> > But that's just my idea, and really, I would like to hear> > what you think of the contents of my de-BW file, as the> > single bit of feedback I got on it was negative.> > Well, as a (mostly?) complete collection of all transmitting sites that could> possibly be received, it's good work, that I would appreciate.
I've since realized that I need to add more from Hessen.  Imade this file based only on looking at maps, and trying tofit my limited experience in the field with what someonemight expect to see...
Ideally it will be complete.  The switchover is complete ingermany (save for a few frequency tweaks), switzerland, andthe nearby area of austria, though I'm not as up-to-date onplans there for additional multiplexes, so the informationshould not go out-of-date anytime soon -- though the swisscould do an about-face and start a second DVB-T2 multiplexwith HD Suisse, just to spite me.  Or a second multiplex toprovide the missing second programmes and give more bandwidthto improve the picture quality of the existing channels.Anything is possible within the GE06 agreement, which is notcompletely covered by my handcrafted scanfile.

> On the other hand, I am not so good in geography, so that I would have to use> some sort of map to find out, what distances are between my location in> Baden-Wuerttemberg (lets for instance say Ulm or Biberach) and the out of area> sites. I would say, that for Biberach additonally sites from Bavaria (Bayern)> or Switzerland may be correct.
No worries, you are no different from me.  Of course, I wouldnot need a map to tell you that in Ulm, you are practicallyin Bayern (unless you are in Neu-Ulm) and due to the flatnature of the land in that direction, you can expect toreceive signals from into the Alps or thereabouts.
The thing about the signals from Switzerland is that whileearlier the Säntis was known for clean strong signals thatwould penetrate well into germany, hundreds of km away,in part due to using one VHF frequency, as part of theswitch to DVB-T, the radiation pattern was greatly changedto send as little signal as possible towards Muenchen andother areas which got either a good or watchable signal.
I have the following rather-old and from-unknown-sourceprediction of the radiation pattern...29.2 27.2 25.2 24.2 22.2 20.2 16.2 16.2 16.2  6.2 14.2 24.2 28.2 28.2 24.2 14.2  9.2 17.2 17.2 17.2 20.2 22.2 24.2 26.2 29.2 33.2 36.2 35.2 17.2 38.2 44.2 46.2 44.2 39.2 25.2 27.2which shows it is far from omnidirectional these days, asis confirmed by reception reports.
The signals do travel well from Basel northwards alongthe Rhein, as I had no problems receiving them while onholiday in the Kaiserstuhl with a simple portable receiverand stub antenna, both french and german bouquets.
The geography of the Schwarzwald, as well as the planningto use a large number of lower-powered sites as part ofthe SFNs, mean that people who used to receive theprogramming from, say, Zürich (also VHF) nowadays canreceive nothing, in spite of being a few minutes fromthe border.  In other words, apart from the Oberrheinfrom Basel to Freiburg im Breisgau and beyond, unless oneis on a hilltop or pretty close to the border, one maynot receive the swiss multiplex at all, while the non-directional high-power german transmitters may comein clearly.

Of course, the easiest solution is to uncomment all theout-of-area frequencies, as well as those used withinthe area, and do your scan based on that.  Then, if Ihave done my homework, you will receive everything.
And do enjoy your stay ;-)  Be sure to take your DVB-Treceiver with you as you climb to the top of the UlmerMünster, if you are not afraid of heights.  And as youhead to Biberach (which one, by the way?) do not forgetto take the long way, stopping at a Hausbrennerei,trying a good home-cooked meal, sampling the local wines,relax in a thermal spa, suffer through the AlemannischeFasnacht, and promise to return, eh?

> I like it, could imagine to do a similar file at least for my Bundesland.
Allow me to offer to volunteer.  First I need to pull outmy maps, and try to get an overview of your area...
But do not expect me to have it complete by tomorrow.As always, I have too many things to and too little timeand too many other things getting in the way...

> Interestingly you also list the height above see level and antenne height of> the BW sites.
Like I said, I wanted to put it all in one place.  Suchtechnical data is easy to find, and quite helpful.  Forexample, I used to think that the `Hochrhein' antennawas the impressive concrete structure that one seesprominently when heading out of Zürich.  Likewise thelisting for de-Loerrach, the imposing antenna is actuallyin Switzerland, St.Chrischona, and according to my oldplans, transmits the swiss-german multiplex, not evenlisted in this scanfile.
Having this info at hand, I can guess at coverage overhills in the area of a certain site, when studying a mapwith some topographic information.  A single hill of acouple hundred metres can make the difference betweena signal I receive with just my finger, and no receptionat all even with a directional antenna, when placedbetween me and a transmitter.

> reception borders for indoor / outdoor / roof antenna) and about the> topography, so that one could estimate for a given geographical location> (retrieved by GPS or from a map oder Google), which transmitters may be in> reach. ;o)
The reception estimates are, in my opinion, highlysuspect, particularly if you download those from SWRfor B-W, due to the nature of the deep valleys thatcause that map to appear like some modern art withgreen ink on a white background for a large area, butthen, I haven't wired my portable receiver for batteryoperation to let me climb a few hundred metres and seeif some sites that might be in range are able to bereceived clearly, or if there is some directionalityat the transmitting site.
That's another thing I don't mention.  I could besitting along the Bodensee now, dipping my toes intothe water in an attempt to get frostbite, staring atthe red glow atop the Säntis and wondering why I'm notgetting a strong signal, while much further away alongthe Kaiserstuhl, without sight to the antenna, I'mhardly able to miss the comparable mux.  Perhaps Icould add this to my comments, if I appear to be leadingpeople to think that coverage figures will be similar.

> There are indeed so many factors in real life, that play a role regarding> DVB-T reception (and scanning), like type and placement of antenna,> sensitivity of the tuner etc. ... which are in fact not a problem of either> auto scan nor a scan file  :-/
As far as the antenna is concerned, I'm hoping that thecomments within are helpful.  For example, the situationin switzerland is that now transmissions are vertical;before they were horizontal, and there are still manyrooftop antennae with the wrong polarisation for thatdirection -- similarly in B-W, a good number of areaswere served by vertically polarised filler transmittersthat now receive signal from a horizontal Grundnetzsender,and antenna orientations are mostly unchanged.
That actually isn't so much an issue with DVB-T, apartfrom fringe areas, as it's capable of delivering nearlya perfect signal where analogue had visible multipath,but it could also be that these rooftop antennae havebeen unused for years, as pretty much every house hassprouted a satellite dish.

Well, that has kept me from actually doing research toadd info to your überregional scanfile  :-)
barry bouwsmalongwinded
_______________________________________________linux-dvb users mailing listFor V4L/DVB development, please use instead linux-media@xxxxxxxxxxxxxxxxxxxxxxxx@linuxtv.org

[Linux Media]     [Video 4 Linux]     [Video Technology]     [Asterisk]     [Photo]     [Samba]     [Xorg]     [Xfree86]     [Devices]     [DVB Maintainer]     [Linux USB]