[linux-dvb] DVB-S in the US.

[Chris Worley <cworley@liberate.com>]

I'm using the Hauppauge DVB-S card with the DVB software, in the U.S..

My dish is aimed at Echostar 6:

	http://www.lyngsat.com/echo6.shtml

(watch out for the pop-up adds.)

In using the libdvb/stest utility, I setup my TEST file as:

LNB ID 0 TYPE 0 DISEQCNR 0
   SAT ID 119 NAME "EchoStar 6" LNBID 0 FMIN 974000 FMAX 1411000
     TRANSPONDER ID 0002 SATID 0119 TYPE 0 FREQ 989000 POL H SRATE 
20000000 FEC 2
     TRANSPONDER ID 0003 SATID 0119 TYPE 0 FREQ 1003000 POL H SRATE 
20000000 FEC 2
     TRANSPONDER ID 0006 SATID 0119 TYPE 0 FREQ 1047000 POL H SRATE 
20000000 FEC 2
     TRANSPONDER ID 0008 SATID 0119 TYPE 0 FREQ 1076000 POL H SRATE 
20000000 FEC 2
     TRANSPONDER ID 0009 SATID 0119 TYPE 0 FREQ 1091000 POL H SRATE 
20000000 FEC 2
     TRANSPONDER ID 000a SATID 0119 TYPE 0 FREQ 1105000 POL H SRATE 
20000000 FEC 2
     TRANSPONDER ID 000b SATID 0119 TYPE 0 FREQ 1120000 POL H SRATE 
20000000 FEC 2
     TRANSPONDER ID 000c SATID 0119 TYPE 0 FREQ 1134000 POL H SRATE 
20000000 FEC 2
     TRANSPONDER ID 000d SATID 0119 TYPE 0 FREQ 1149000 POL H SRATE 
20000000 FEC 2
     TRANSPONDER ID 000e SATID 0119 TYPE 0 FREQ 1164000 POL H SRATE 
20000000 FEC 2
     TRANSPONDER ID 0011 SATID 0119 TYPE 0 FREQ 1207000 POL H SRATE 
20000000 FEC 2
     TRANSPONDER ID 0012 SATID 0119 TYPE 0 FREQ 1222000 POL H SRATE 
20000000 FEC 2
     TRANSPONDER ID 0014 SATID 0119 TYPE 0 FREQ 1251000 POL H SRATE 
20000000 FEC 2
     TRANSPONDER ID 0015 SATID 0119 TYPE 0 FREQ 1266000 POL H SRATE 
20000000 FEC 2

I figure we don't need the dual LNB baselines in the states, so I set 
all that to zero, and work directly off the LNB frequency.

The FEC is reported to be 3/4.  I don't know which dvb code this 
corresponds to, but FEC code "2" and "3" work; "2" seems to work a 
little better.  I'm not sure what the "type" field is, it doesn't seem 
to effect results anyway.

Running stest proves somewhiat unpredictable.  Sometimes it finds 
PAT/PMT data for a frequency, other times it doesn't.  Sometimes the 
channel ID's are all printed as "FFFFFFFF" for a given frequency, 
sometimes the printout is fine.

I did find one error, where DVB.cc got confused and try to look at 
cable data.  In "DVB::scan_tp":

		if (!prog->prog_nr) {
		        dvb_nit_t *nit = (dvb_nit_t *)calloc (1, sizeof (dvb_nit_t));
			if (GetNIT(nit) >= 0) {
			        if (prog->prog.nit)
				        free(prog->prog.nit);
				prog->prog.nit = nit;
			} else
			        free(nit);

			//if (fdvb >= 0) osd.Text(64, 120, 2, 1, "NIT");

			cerr << "NIT\n";

			// FIXME: fix NIT parsing
			//if (front.type!=FRONT_DVBC)
		      tp.freq=(uint32_t) (1000000*nit->freq+0.5);

I commented out that last line, and it quit printing the wrong 
frequency for found PAT/PMT's.

The output data from stest looks something like this:

LNB ID 0 TYPE 0  DISEQCNR 0
   SAT ID 119 NAME "EchoStar 6" LNBID 0 FMIN 974000 FMAX 1411000
     TRANSPONDER ID 0002 SATID 0119 TYPE 1 FREQ 989000 POL H SRATE 
20000000 FEC 2
       CHANNEL ID 0 NAME "BET" SATID 0119 TPID 2 SID 7c TYPE 1 VPID 
1022 APID 1023 PCRPID 1022
       CHANNEL ID 1 NAME "HBO-E" SATID 0119 TPID 2 SID 12c TYPE 1 VPID 
1422 APID 1423 PCRPID 1422
       CHANNEL ID 2 NAME "GOLF" SATID 0119 TPID 2 SID 191 TYPE 1 VPID 
1822 APID 1823 PCRPID 1822
       CHANNEL ID 3 NAME "EVENT" SATID 0119 TPID 2 SID 1c7 TYPE 1 VPID 
1022 APID 1023 PCRPID 1022
       CHANNEL ID 4 NAME "CNNFN" SATID 0119 TPID 2 SID ce TYPE 1 VPID 
1422 APID 1423 PCRPID 1422
       CHANNEL ID 5 NAME "GLVSN" SATID 0119 TPID 2 SID 110 TYPE 1 VPID 
1822 APID 1823 PCRPID 1822
       CHANNEL ID 6 NAME "TECH" SATID 0119 TPID 2 SID bf TYPE 1 VPID 
1022 APID 1023 PCRPID 1022
       CHANNEL ID 7 NAME "COURT" SATID 0119 TPID 2 SID cc TYPE 1 VPID 
1422 APID 1423 PCRPID 1422
       CHANNEL ID 8 NAME "BRAVO" SATID 0119 TPID 2 SID 81 TYPE 1 VPID 
1022 APID 1023 PCRPID 1022
       CHANNEL ID 9 SATID 0119 TPID 2 SID 8f01 TYPE 0 VPID 1422 APID 
1423 PCRPID 1422
       CHANNEL ID a NAME "AUD01" SATID 0119 TPID 2 SID 39b TYPE 1 VPID 
1822 APID 1823 PCRPID 1822
       CHANNEL ID b NAME "AUD13" SATID 0119 TPID 2 SID 3a8 TYPE 1 VPID 
1022 APID 1023 PCRPID 1022
       CHANNEL ID c NAME "AUD16" SATID 0119 TPID 2 SID 3ab TYPE 1 VPID 
1422 APID 1423 PCRPID 1422
       CHANNEL ID d NAME "METRO" SATID 0119 TPID 2 SID 3a4 TYPE 1 VPID 
1822 APID 1823 PCRPID 1822


Note how the vpid/apid data keeps repeating (for example, it says the 
only audio PIDS for these 14 channels are 1022, 1422, and 1822).  The 
real PIDS can be seen here:

	http://www.lyngsat.com/dig/dish119.shtml

(Again, sorry for any pop-up adds.)

Furthermore, stest doesn't find anything at the frequencies for about 
half of the transponder ID's.

I found two flags pertaining to NTSC/PAL in the driver, but switching 
them to what should have setup NTSC output didn't work, so I switched 
them back, and got a PAL-syncing TV.

Switching to VDR, I setup my channel.conf file with a clear (no CA) 
channel:

	DISH1:11340:h:0:20000:4386:4387:12:0:101

(1134 MHz, 200 M/sym/sec, vpid/apid 4386/4387, transponder 12, no CA, 
SID 101.)

Those PIDS are as reported at the above web site for this channel 
(rather than what the stest scan found).

In playing with this line in channel.conf, there were a few times when 
I could get VDR to display the channel information data in the stream, 
but I have not been able to repeat that (it allways comes up with 
"DISH1" and the time, but a few times it also came up with the channel 
information data describing the channel).

I can't get it to display any video data.

Any ideas on what other tests I can do or get it to work?

Thanks,

Chris




-- 
Info:
To unsubscribe send a mail to listar@linuxtv.org with "unsubscribe linux-dvb" as subject.