Open andimik opened 4 years ago
Seems a small hardware modification (22uF capacitor) solves some issues with my TBS 6903X. But I can't modify it on my own.
Check the screenshots, after the mod the stream will be recognized correctly as generic continuous stream.
https://www.satellitescommunity.de/forum/index.php?thread/1684-tbs6903x/&postID=39361#post39361
😮 😮 😮 👍 👍 👍
@newspaperman Before you write anything for TBS 6903X and 6909 cards, please wait till I get my card modified, but this might take weeks.
I already got a stream from a modified tbs0903x, but it's still corrupted :(
So for now on DigitalDevices cards are working. We need to ask Crazycat to set FORCE_CONTINUOUS and TSOUT_NOSYNC in windows, these are hw registers on the card, so they should be available on windows too.
Regarding NRK:
Now I tried under Ubuntu MATE 18.04 with the latest TBS drivers from Sept 2020 (http://www.tbsiptv.com/download/common/tbs-open-linux-drivers_v20200923.zip) and latest dvblast 3.4 from https://www.videolan.org/projects/dvblast.html.
dvblast -s 5400000 -v 13 -f 10720000 -m psk_8 -u -3 -a 0 --multistream-id-is-id 171 > /tmp/tbs6903x_linux_nrk_test_mis171.gse
Recorded with a 6903x v1 without hardware modification:
I fear these two recordings still contain errors, as the ensemble names still are not correct.
There is even no difference if I use MIS 171 or not.
Which tool and command do you use under Linux?
thanks for the recording. Can you please repeat this under Linux for the bundesmux. Now it looks very promising to me.
Well, it seems it's not possible to set a specific MIS in Linux as well. No idea if this could be solved after the hardware mod.
I tried -1 2
, then -9 2
and also --multistream-id 2
grep Radio
looks for a text which is used in both muxes:
$ dvblast -s 1500000 -v 13 -f 12641000 -m psk_8 -3 -a 0 -1 1 -u | strings -16 | grep "Radio"
DVBlast 3.4 (release)
...
debug: MULTISTREAM
...
debug: frequency 12641000 is in Ku-band (higher)
debug: configuring LNB to v=13 p=0 satnum=0 uncommitted=0 lnb-type=universal bis_frequency=2041000
debug: tuning DVB-S frontend to f=12641000 srate=1500000 inversion=-1 fec=999 rolloff=35
modulation=psk_8 pilot=-1 mis=1 /pls-mode: ROOT (0) pls-code: 0 is-id: 1 /
...
lock status: 1
...
debug: - Bit error rate: 0
debug: - Signal strength: 40672
debug: - SNR: 45264
...
TOGGO Radio
Radio Horeb
PDRadio Daten
...
Surprisingly, I got lots of (but not all) frames out (under Linux) with even a TBS5927. Much more than under Windows.
But it re-syncs regularly (LED flashes) each around 10 seconds.
I checked the content, MIS 1 and MIS 2 are correct, so it's not a software issue.
$ cat /tmp/test5927_mis2.gse | strings -16 | grep "Radio" | uniq
TOGGO Radio
4OEJ :tGGO Radio
S~]GGO Radio
PFIGGO Radio
nKGGO Radio
:QGGO Radio
--GGO Radio
$ cat /tmp/test5927_mis1.gse | strings -16 | grep "Radio" | uniq
@DRadio DokDeb
Radio Horeb
PDRadio Daten
Radio Horeb
@DRadio DokDeb
PDRadio G_MM
PDRadio Daten
@DRadio DokDeb
Radio Horeb GG{
Radio Horeb
@DRadio DokDeb
PDRadio Daten
Radio Horeb
PDRadio Daten
@DRadio DokDeb
PDRadio Daten
@DRadio DokDeb
PDRadio Daten
@DRadio DokDeb
PDRadio Daten
@DRadio DokDeb
PDRadio Daten
@DRadio DokDeb
PDRadio Daten
Deutschlandfunk Radionacht
@DRadio DokDeb
This was a very good idea to try the TBS6903x under linux . The good news: The encapsulation format looks exactly like the one from my CineS2 V7a The bad news: Also here frames are missing.
I don't use dvblast, I use tune-s2 for tuning and in a separat window I use "dvbstream -o 8192 > output.ts" Perhaps this makes the difference or what's more likely, it's in the driver.
Fortunately the linux driver is opensource. You could try to set all occurences of
ChipSetField(pParams->handle_demod, FLD_FC8CODEW_DVBSX_HWARE_TSSTATE1_TSOUT_NOSYNC(Demod), 0);
to
ChipSetField(pParams->handle_demod, FLD_FC8CODEW_DVBSX_HWARE_TSSTATE1_TSOUT_NOSYNC(Demod), 1);
and in every if /else if/ else path where the TSOUT_NOSYNC is set also make sure that FORCE_CONTINUOUS is also set to 1 error |= ChipSetField(pParams->handle_demod, FLD_FC8CODEW_DVBSX_PKTDELIN_PDELCTRL2_FORCE_CONTINUOUS(Demod), 1);
the file is located here: media_build/linux/drivers/media/dvb-frontends/stid135/stid135_drv.c Don't forget to rmmod and modprobe the module (or reboot) after make install I can send you the patched if you like.
I hope the patch was successful, although some instances already were set to 1. Otherwise, please send me your file. stid135_drv.c.zip
tbs6903x_nrk_after_stid135_drv_patch.gse.zip
This is NRK DAB+ after the patch, because I still cannot select the correct MIS for BuMux.
I also tried your new tool here on Github, but I have no idea how to proceed afterwards. Wireshark cannot import this in order to get the UDP addresses ... Maybe you can give me a hint.
As far as I understood you then just use it in odr-dabmod
?
stid.zip Thanks for the capture again, same problem still exists. Please try again with my file + please do a reboot.
Glad you found my new tool. However it's only extracting the bbframes out of the pseudo TS. What's missing is a tool that extracts the UDP payload out of the bbframes, which can be sent to edi2eti. It will be a combined tool gseedi2eti, but it's not 100% finished.
I will publish my modified gsextract tool, then you can at least see something in wireshark. With the corrupt stream however you won't get much data. From your 13 MB file, I get a pcap file sized only few kB
One info about MIS: The card is now in BBframe mode. MIS is just one single byte in the bbframe and can be filtered later by the bbframe processing.
B8 52 AB 00 B8 = BBSync Byte inserted by card 52 = MATYPE1 AB= MATYPE2 == MIS 171
tbs6903x_nrk_after_stid135_drv_patch_try2.gse.zip
New try, better now?
I exchanged that file (which really was different from mine) and executed make
and sudo make install
and rebooted.
But the names are still mixed up now.
$ cat /tmp/tbs6903x_nrk_after_stid135_drv_patch_try2.gse | strings -8 | grep "NRK Reg" | sort | uniq
NRK Reg1 GA
NRK Reg1 OsloVik
NRK Reg2 VeTeVik
NRK Reg3 S
NRK Reg4 Vest
NRK Reg5GA
NRK Reg5 Innland
NRK Reg6
NRK Reg6 Tr
NRK Reg7 NoTrFi
NRK Reg7 NoTrFiGA
Unfortunately problem is not solved. I published my modified gsextract version on github. Small installation instruction:
git clone https://www.github.com/newspaperman/gsextract-b8
pip3 install ./gsextract-b8
export PATH=$PATH:/home/[user]/.local/bin # whereas [user] is your username
you can then run it directly:
gsextract input.file output.pcap
intput.file is the file that was processed by pts2bbf (see https://github.com/newspaperman/bbframe-tools )
Usage of pts2bbf:
cat input.ts | pts2bbf | input.bbf
You will then see that the output is very corrupted. The *.pcap
file can be viewed in Wireshark
I have edited your posting as two paths were not correct ...
Well, I cannot believe that gsextract
finds errors in every second BBFrame.
I even cannot stop it with ^c
and it is rather slow.
And at the end I get errors:
Traceback (most recent call last):
File "/home/andreas/.local/bin/gsextract", line 11, in <module>
sys.exit(cli_runner())
File "/home/andreas/.local/lib/python3.6/site-packages/gsextract/gsextract.py", line 13, in cli_runner
gsextract()
File "/home/andreas/.local/lib/python3.6/site-packages/click/core.py", line 829, in __call__
return self.main(*args, **kwargs)
File "/home/andreas/.local/lib/python3.6/site-packages/click/core.py", line 782, in main
rv = self.invoke(ctx)
File "/home/andreas/.local/lib/python3.6/site-packages/click/core.py", line 1066, in invoke
return ctx.invoke(self.callback, **ctx.params)
File "/home/andreas/.local/lib/python3.6/site-packages/click/core.py", line 610, in invoke
return callback(*args, **kwargs)
File "/home/andreas/.local/lib/python3.6/site-packages/gsextract/gsextract.py", line 10, in gsextract
gse_parser.gse_parse(file=input_file, outfile=output_file, stream=stream, reliable=reliable)
File "/home/andreas/.local/lib/python3.6/site-packages/gsextract/gse_parser.py", line 101, in gse_parse
raw_packets = parse_gse_packet_array([],0, cleanup=True,reliable=reliable)
File "/home/andreas/.local/lib/python3.6/site-packages/gsextract/gse_parser.py", line 179, in parse_gse_packet_array
extracted_ip_packets = extract_ip_from_gse_data(entry[1], high_reliability=reliable)
IndexError: index out of range
You said you've modified it? But your fork is even with the original master.
ok sorry you are absolutely right, I commited my change locally only and forgot to push it to github. Please execute: cd gsextract-b8 git pull origin master cd .. pip3 uninstall gsextract pip3 install ./gsextract-b8
I uploaded one correct capture of bmux for you (and others): https://mega.nz/file/1MgGTLJB#cDUZBgUlJe99PrcpWLoSld5PUmSeFXoz69W38ypt18c
You will see the difference between my and your capture. Now there are many registers you can try to set and see if it makes any difference... more to follow
You can insert the following 2 lines after line 5004 and see if there is some improvement, if you have only 1 error at the beginning it's normal. error |= ChipSetField(pParams->handle_demod, FLD_FC8CODEW_DVBSX_HWARE_TSCFG1_TSFIFO_MANSPEED(Demod), 3);
error |= ChipSetOneRegister(pParams->handle_demod, (u16)REG_RC8CODEW_DVBSX_HWARE_TSSPEED(Demod), 0x18);
If that's not improving replace the 0x18 by 0x16, 0x10 and 0x28 and try again.
If that's not gonna change anything we try other registers.
I've to try in the next days on a different PC. There the card is working fine, but I have to install Ubuntu on that PC.
Because suddenly the mainboard on my linux PC does not recognize the card any more and under Windows (on the same PC) there is an exclamation mark in the hardware manager.
But thanks for the modified version. I tried with a recording from some entries above and got 7 addresses 239.x.x.x out for the 8 streams.
did you try to boot your pc once without card?
Yes, but never mind, this PC is more than 15 years old and was very loud at the end. So no big loss ... The new one runs much much faster, on the other hand.
I tried to change this register setting according to your description, but only in line 5006 (+ make
and sudo make install
+ reboot)
What about line 5059 (I guess it was formerly 5057)? There is the same 0x18 as well.
For each 0x number you gave me I've recorded exact 1 minute from the NRK DAB+ Transponder with
timeout 60 dvblast -s 5400000 -v 13 -f 10720000 -m psk_8 -3 -a 0 -1 171 -u > /tmp/nrk.ts
then processed the bbframes
cat /tmp/nrk.ts | ./bbframe-tools/pts2bbf > /tmp/nrk.bbf
and then wrote a pcap file.
gsextract /tmp/nrk.bbf /tmp/nrk.pcap
0x | ts | bbf | pcap | remark |
---|---|---|---|---|
0x28 | 82.1M | 77.8M | 1.2M | index out of range in gsextract |
0x10 | 12k | lots of warning: transport_error_indicator on pid | ||
0x16 | 82.1M | 77.8M | 1.1M | no index out of range, but corrupt data each 2nd bbf |
0x18 | 82.1M | 77.8M | 1.2M | no index out of range, but corrupt data each 2nd bbf |
Beside of the 0x10 I've saved all files for possible future investigation.
Maybe this screenshot helps as you can see that the Oslo ensemble (Number 1) has much more output in pcap than the others.
In the 8th stream there is the private national Norvegian mux on 12D of Riks, so the Readme has been updated.
And a SDR in Windows might handle the NRK Transponder as well, as the frequency is very low and is in the range of most cheap RTL dongles.
https://forum.amsat-dl.org/index.php?thread/101-software-dvb-s-demodulator/&pageNo=35#post16808
Ok fine, then keep the 2 lines and add a third one either with 2,1 and 0:
error |= ChipSetField(pParams->handle_demod, FLD_FC8CODEW_DVBSX_HWARE_TSSYNC_TSFIFO_SYNCMODE(Demod), 2);
here the results
u16)REG_RC8CODEW_DVBSX_HWARE_TSSPEED(Demod), 0x18
0x | ts | bbf | pcap | remark |
---|---|---|---|---|
0x2 | 80.7M | 76.4M | 1.2M | corrupt data each 2nd bbf |
0x1 | 80.7M | 76.4M | 1.4M | corrupt data each 2nd bbf |
0x0 | 13.2k | only warnings: transport_error_indicator on pid xxx |
with 0x28
0x | ts | bbf | pcap | remark |
---|---|---|---|---|
0x2 | 82.1M | 77.8M | 1.1M | |
0x1 | 82.1M | 77.8M | 1.2M | corrupt data each 2nd bbf |
0x0 | 82.1M | 77.8M | 1.2M | corrupt data each 2nd bbf and index error: Out of range |
with 0x20
0x | ts | bbf | pcap | remark |
---|---|---|---|---|
0x2 | 9.6k | only warnings: transport_error_indicator on pid xxx |
What does strings -6 | sort | uniq | grep "NRK Reg"
show with your Cine card? 7 lines or more like me?
As promised check my new bbfedi2eti tool https://github.com/newspaperman/eti-tools
Here is a capture from 1°West https://mega.nz/file/MUZkWYSD#RW7fZUR-us4924MxG9jwjdApIkMqTrT1gaImM9Gse8I
Confirmed! This works!
Assuming the capture is from your DD Cine, right? As the capture is from 2018.
correct :)
This is captured with my 6903X from upper mentioned https://github.com/piratfm/eti-tools/files/5565897/tbs6903x_nrk_after_stid135_drv_patch_try2.gse.zip some days ago.
I'll test later if the gsextract had a bug. It seems that dablin_gtk had no problem to play it.
Very good news.
It works nearly live even with my TBS 6903-X, I'm currently using
u16)REG_RC8CODEW_DVBSX_HWARE_TSSPEED(Demod), 0x28
and
FLD_FC8CODEW_DVBSX_HWARE_TSSYNC_TSFIFO_SYNCMODE(Demod), 2);
.
Let me test it with the original settings as well.
I have to write a ts file with dvblast.
dvblast -s 5400000 -v 13 -f 10720000 -m psk_8 -3 -a 0 -1 171 -u > /tmp/nrk.ts
And in a 2nd console I play it
cat /tmp/nrk.ts | ~/apps/bbframe-tools/pts2bbf | ~/apps/newsy_eti-tools/bbfedi2eti -dst-ip 239.199.2.8 -dst-port 1234 | dablin_gtk
But it won't work with a combination with sleep
as it takes some seconds to lock (could you give me a hint if my command is wrong???)
dvblast -s 5400000 -v 13 -f 10720000 -m psk_8 -3 -a 0 -1 171 -u | sleep 10 && ~/apps/bbframe-tools/pts2bbf | ~/apps/newsy_eti-tools/bbfedi2eti -dst-ip 239.199.2.8 -dst-port 1234 | dablin_gtk
Great news, so you don't get parsing errors in gsextract with these settings?
u16)REG_RC8CODEW_DVBSX_HWARE_TSSPEED(Demod), 0x28 and FLD_FC8CODEW_DVBSX_HWARE_TSSYNC_TSFIFO_SYNCMODE(Demod), 2);
Usually this should work: dvblast -s 5400000 -v 13 -f 10720000 -m psk_8 -3 -a 0 -1 171 -u | ~/apps/bbframe-tools/pts2bbf | ~/apps/newsy_eti-tools/bbfedi2eti -dst-ip 239.199.2.8 -dst-port 1234 | dablin_gtk
did you try this already?
dvblast -s 5400000 -v 13 -f 10720000 -m psk_8 -3 -a 0 -1 171 -u | (sleep 10 && ~/apps/bbframe-tools/pts2bbf) | ~/apps/newsy_eti-tools/bbfedi2eti -dst-ip 239.199.2.8 -dst-port 1234 | dablin_gtk
I use tune-s2 to tune separately and dvbstream will stream immediately.
I also use tune-s2 for usals. However you have to tell -long -lat on every call.
I published a new tool in my bbframe-tools, you can try to play a stream on vlc. It's on 33°East 12544 V,1703,5/6,8PSK 224.229.100.96:1001 and 224.229.100.97:1001
If this video is also played correct, I think you solved the driver issue :+1:
Thanks for the hint, yes the brackets ( ) now work together with sleep
. Remark: If I use less than 10 seconds it may happen (= sometimes it works, sometimes not) that the EOF is reached.
Success also with Bundesmux 1 and 2!
No need to separate MIS, it's done via their different IP-Addresses (which I already found out before).
dvblast -s 1500000 -v 13 -f 12641000 -m psk_8 -3 -a 0 -1 2 -u | (sleep 10 && ~/apps/bbframe-tools/pts2bbf) | ~/apps/newsy_eti-tools/bbfedi2eti -dst-ip 239.128.72.10 -dst-port 50010 | dablin_gtk
and
dvblast -s 1500000 -v 13 -f 12641000 -m psk_8 -3 -a 0 -1 1 -u | (sleep 10 && ~/apps/bbframe-tools/pts2bbf) | ~/apps/newsy_eti-tools/bbfedi2eti -dst-ip 239.128.43.43 -dst-port 50043 | dablin_gtk
(I even did not know, ABBA was singing that in German! No joke!)
Remark: for me, -1 1
or -1 2
makes no difference as the card (currently??) cannot handle MIS.
But for Bundesmux 2 I get this warning around 42x in a second in bbfedi2eti
(the output file plays well, BTW)
[11/22/20 18:30:21] EDI: Unknown TAG Fptt
So there is no need to modify my card at the moment.
so you don't get parsing errors in gsextract with these settings?
Sure, I still get these errors (see above, I have tested it already with various settings). Well, gsextract was just used to check the IP addresses only once. So let's come to the conclusion that there is a bug in gsextract or - not really logical - the lost frames are not needed indeed.
My movable dish cannot receive 33 East (12W and 26E are its maximum), I could try with another one in the garden where I have free view to all others. Maybe I'll try next weekend at sunlight.
Due to this new milestone I should close this issue and open part 4 soon.
As soon as the parsing errors disappear you can send a PR.
Seems the driver needs not to be modified because if you play the file I've recorded on 18th Nov (http://andimik.bplaced.net/temp/tbs6903x_linux_nrk_test_mis171.gse ) then you also can hear it! But the other is not working.
Maybe it's a kind of offset?
cat /tmp/tbs6903x_linux_nrk_test_mis171.gse | pts2bbf | ./apps/newys_eti-tools/bbfedi2eti -dst-ip 239.199.2.1 -dst-port 1234 | dablin_gtk
Let me test with the original open source TBS driver (= not modified) in the next days.
yes I also see that unknown tag error with my Cine S2 V7a.
@piratfm Do you know what this unknown tag could mean?
[11/22/20 18:30:21] EDI: Unknown TAG Fptt
Let's see if it's really working with the original driver, I don't know how long those set registers stay active.
Perhaps there's even more interesting GSE stuff on other satellites.
If someone wants to try with a SDR (software defined radio) solution: Use a DVB-T (!!) Stick with E4000 Tuner (due to the very high frequency above 2 GHz) or HackRF or other and http://www.pabr.org/radio/leandvb/leandvb.en.html and then please report if you can see and use (!) the BB frames.
Next week a friend of mine will give me his NooElec NESDR SMArt XTR with E4000 tuner. I'll try it.
Seems that
wget -q -O - https://rbbedi.irt.de:443/services/x | dablin_gtk -f edi -I
wget -q -O - https://edistream.irt.de/services/x | dablin_gtk -f edi -I -1
and
wget -q -O - https://bredi.irt.de:443/services/x | dablin_gtk -f edi -I
don't work any more, there is only a bad gateway error.
The only one remains VRT, see overview for SubChannels at http://edihttp.vrt.be:8187/services/
Are this errors related to this text that appears on dabplus.de website ???
Am 21.12.2020 wird auf Digitalradio (DAB+ oder IP) umgestellt: 21 Tage 23 Stunden.............
Do you know what does it mean ? (I can translate it, but I don't know what refers to...)
This is a EU directive, even for those countries not having any DAB on air like Finland or Portugal.
In Italy all cars sold even from January 1st 2020 have to be equipped with DAB+. For all others it's December 21st.
A similar law is expected for UK for 2021.
Obviously there is not such a law in Norway, as World DAB reported, that (only) 98% of a new cars are equipped with DAB.
So it's just a warning for potential car buyers and their sellers.
You could look for ECC or EECC
https://en.wikipedia.org/wiki/European_Electronic_Communications_Code
So it has nothing to do with existing DAB services and of course not affecting satellite DAB reception.
https://avancedigital.gob.es/es-es/Participacion/Paginas/DetalleParticipacionPublica.aspx?k=346
This I found for Spain.
Ok. Thanks. I already knew this EU Directive. I got confused when translating..... I translated "DAB+ oder IP" as "DAB+ on IP" , and as those missing EDI services where really DAB+ on IP (internet), I thought it may have something to do......
The ones that don't work are all from IRT servers. I heard some time ago that IRT was to be closed, following ZDF and other decisions to leave it.... ????
Regarding 33 East, I have some dropouts. no idea if this is good enough.
$ tune-s2 12544 v 1703 -adapter 0 -system DVB-S2 -modulation 8PSK -fec 5/6 -lnb UNIVERSAL
and in a 2nd console
$ dvbstream -o 8192 | ~/apps/bbframe-tools/pts2bbf | ~/apps/bbframe-tools/bbfudpdecap -dst-ip 224.229.100.96 -dst-port 1001 | vlc -
This is the recording under Linux: 33ost_udp.zip
Hi,
PID 1061 also works on 28.2E now with the BBC Radios :)
Is it this Transponder?
https://www.digitalbitrate.com/dtv.php?mux=11425&liste=1&live=69&lang=en
It is on the UK beam ...
Yes, it's on 28.2E, 11425 H 27500, I'm able to receive this frequency in France, near Nantes. I don't have room for installing an lnb on 4.9E so I thought I'd give another try on 28.2E since there was a service with the right name and bitrate that was not present on the list.
DAB feeds at 4.8 ° E have now been switched off.
Thank you for the information.
All three UK DAB streams on 4.8E have left and can now only be received in parts of Western Europe as this is the UK beam.
See https://github.com/piratfm/eti-tools/commit/1b9209fb122188eb011699794d2aedb166f46a62
or https://www.digitalbitrate.com/dtv.php?mux=12303&liste=1&live=71&lang=en
SID is 10580 for BBC DAB on 28.2E.
Bye Bye BBC DAB.....
There's a transponder on 16°E that might contain Dutch DAB+ multiplexes. It's listed as DVB-S2 32APSK so I can't check it myself as I haven't got a compatible card.
11072 H 1497 32APSK 3/4
It's listed on Flysat as carrying 'Data' - maybe DVB-GSE or DVB-IP on a single PID? The TS bitrate would be ~5.5Mbps so enough room for the 2-3 national Dutch ensembles.
The surrounding frequencies are all low symbol rate feeds carrying Dutch radio stations, which are I assume the feeds for the FM transmitters in the Netherlands. That's what's making me wonder if this data frequency is carrying DAB+!
Well, I have already checked every satellite between 62 East and 30 West in the past months.
As one of my dishes is permanently pointing to 16 East I wondered if I would have missed such a signal.
But there are some data transponders which I could not lock, in deed.
I can’t receive anything on 11072 H.
I receive 11071,4 H, DVB-S/QPSK, 547, ½ , so I think i’ts not possible to have another carrier al 11072 with SR aprox 1500 KS/s
There is another 32APSK carrier on 11054 H, SR 1497, FEC ¾., but I have a lot of errors, I have only SNR 7 dB at this frequency, impossible to handle 32APSK....
:-(
https://en.m.wikipedia.org/wiki/Trap_street
I fear this is either a fake entry at the flysat page to protect their copyright.
Or it's a copy&paste error (see some lines above, similar data).
The Bundesmux 1+2 Transponder is still listed as DVB-S only.
So please don't be surprised if you can't lock in a transponder which is listed there.
I've edited previous comment. I don't know what happened, I didn't quote anything........ 😮
Hello, i need an dump of 12518 H, 12518 V, 12519 H, 12530 V and 10810 H in Hispasat 30W.
Summary of closed issue #28 (= Satellite sources of DAB feeds - part 2)
wget -q -O - https://rbbedi.irt.de:443/services/x | dablin_gtk -f edi -I
wget -q -O - https://bredi.irt.de:443/services/x | dablin_gtk -f edi -I
wget -q -O - http://edihttp.vrt.be:8187/services/x | dablin_gtk -f edi -I -1
to listen to, whereas x is the subchannel ID, see https://www.fmlist.org/sendertabelle/dab-ww.php and http://search-hradio.irt.de/ Combining them to a single DAB ETI stream is not 100% possible, but there is a workaround with dropouts