Stuart, On 28/11/05, Stuart Auchterlonie <[EMAIL PROTECTED]> wrote: > On Sat, Nov 26, 2005 at 02:11:08PM -0800, Joe Votour wrote: > > > > The EIT is the "Event Information Table" (if I > > remember my TLAs correctly). The DVB specification > > supports sending lots (and I mean lots) of information > > within it (the basics such as show start/end times, > > program name and genre, but also things like complete > > cast lists, MPEG format information, etc.). > > > > The way it works is that within the MPEG-2 stream that > > is received by your DVB card, there are video, audio > > and PCR (timing) PIDs for one or many "channels", > > there are extra PIDs that are added in by your > > provider (or may come from the original satellite feed > > - regardless of how they got there, they are there). > > A DVB card can listen on these PIDs and capture/parse > > the data, and turn it into program guide entries. > > Most of it is TLV-type (type, length, value) data with > > a fair amount of ASCII text. > > This is the best explanation of EIT transmission > I've seen lately :-)
Agreed. But I do have to ask what's a PID? A programme identifier? No? > > As I said, I don't know the specifics of the EIT > > processing within MythTV, but I would suggest that you > > stick with the tv_grab programs if you get accurate > > data from them. The reasons that I suggest this are: > > 1. With the DVB streams that I have seen, the program > > guide information within the DVB stream is only good > > for 1-2 hours (typically, you get the "now" and > > "next", the currently showing program, and the next > > program). This means that it will be difficult to > > plan far ahead in the future for scheduling, since the > > data just won't be there. > > Currently Myth doesn't use the information in the now > & next descriptors, it uses information from the future > event descriptors. This contains information anything > up to about 7 days in the future. This is Myth in SVN, I take it? > > 2. In order to get complete EIT data, your DVB card > > will likely need to hit multiple frequencies to > > capture the EIT data on those frequencies. If you > > have a single card and aren't recording, this won't be > > an issue, but if you're recording back to back shows > > and only have a single card, you've got nothing to > > tune to other frequencies to find the EIT data. In > > that case, you're missing program guide listings. > > Few comments to this. > 1. To get a full spread of infomation you need to > visit all the transports. That is what the EIT crawler > does. You will see it operating with the > > "now looking for EIT data on channel x" > > This is known as Active Scan mode. > > 2. When watching a program the EIT scanner goes into > Passive Mode where it collects information while > you are recording / watching. You will still end up > visiting all the transports but it might take a bit > longer.... This sounds interesting. > > Again, this information doesn't come from the MythTV > > implementation (I've implemented this in the past, and > > it's quite painful), but comes from my knowledge in > > the area. YMMV, depending on what kind of hardware > > you have, and how much program data you want (I'm > > quite happy with DataDirect, and having about 10 days > > worth of listings). > > > > The major thing to know is the quality of the transmitted > EPG data depends a LOT on where you live. > > UK DVB-T has great 7day EPG. > German DVB-T has great 7day EPG. > European DVB-S has usable EPG, though YMMV. I'd like to use this, because otherwise I'd have to use many different grabbers (for UK and German TV), and would have to associate the channel IDs from the grabbers with the channel IDs from the channel database table. BTW, how do people do this association? Because for me the association of Guide channel IDs and channel IDs doesn't work automagically. I have to manipulate the Mysql channel table by hand, and fill in the xmltvid field manually. > Aussie EPG is almost useless. > > No idea about ATSC EPGs. > > > So depending on where you live the over the air EPG is > the best source or the worst source of information. > > > Stuart thanks, Stuart. -- cheers, Jakob. _______________________________________________ mythtv-users mailing list mythtv-users@mythtv.org http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users
