Jump to content
Derrick

TransEdit Discussion

Recommended Posts

Derrick

..no thx, it's you who deserves the credits (w00t)

 

This is the best tuning tool for SS2 I've tried so far..

It's equally the state of the art for users who to update

their Astra or Hotbird settings as well as for dxer's who

want a fast scan of narrow scpc carriers :o

 

Cheers

 

Derrick

Share this post


Link to post
boborg

Thanks from here as well. Always nice with new things to play with.

Share this post


Link to post
crypto

Hi Griga,

 

thank you very much for this great tool. Channel scanning is a complex task and I haven't found anything that comes as close to what I need as your scanner. DVBViewer definitely needs an update, dvbportal has no own scanner, because I know how much work needs to be done <_<, and setup4pc is ok, but very slow and has a duplicate transponder problem. Even though the scanner is close to perfect, please let me make some comments on it.

 

You mention, that the transponder descriptions are a non-standard feature. Why don't you use the provider as a description? Why rely on third party transponder lists? The dvb standard has defined NIT parsing for that. For non-listed transponders and feed hunting a step-scanner as in setup4pc would do as an advanced feature. I also don't like the duplicate entries for each audio pid. A single entry would do. The audio tracks need to be parsed each time a channel is tuned in, because many channels have changed audio tracks from transmission to transmission.

 

Keep up the good work.

Share this post


Link to post
Griga
Why don't you use the provider as a description?

 

I could do that... however, sometimes there is more than one provider on a transponder, and I don't want my personal description to be overwritten. Maybe something like

 

IF description is empty AND provider present THEN description := comma delimited list of providers

 

I will think about it...

 

Why rely on third party transponder lists? The dvb standard has defined NIT parsing for that.

 

NIT... yes I saw these three letters here and there in the specs. <_< But I don't know yet how it works and how reliable it is. Looks like I have to read a bit more...

 

For non-listed transponders and feed hunting a step-scanner as in setup4pc would do as an advanced feature.

 

I already thought about it and discussed it with Derrick - some kind of "Update Transponderlist" or "Create Transponderlist" function.. However, there are some tricky problems, especially concerning the symbolrate. The new SS2 hardware rev. doesn't support SR auto-detection anymore. On Astra it's sufficient to try 22000 and 27500, but on other satellites some more symbolrates may be used. With, let's say, six different SRs you get the following calculation:

 

6 SRs x 2 (H/V) x 2 seconds tunerlock-timeout = 24 seconds per trial

 

Quite a lengthy thing - and how can the accurate frequency of a transponder be determined? By checking the frequency range with 1 MHz steps and then take the medium frequency - even more time-consuming. Any idea how this can be shortened?

 

I also don't like the duplicate entries for each audio pid. A single entry would do. The audio tracks need to be parsed each time a channel is tuned in, because many channels have changed audio tracks from transmission to transmission.

 

Well, I tried to keep it DVBViewer-compliant :P personally I prefer a separate entry for each audio PID, because my ol' TechniSat receiver does it just the other way round, thus making it difficult to handle multi-radio services (several radio channels sharing one SID) and to program timer-recordings for a certain audio channel. I think both ways of "mapping" channels have their (dis-)advantages.

 

Griga

Share this post


Link to post
crypto

Hi Griga,

 

> IF description is empty AND provider present THEN description := comma

> delimited list of providers

 

> I will think about it...

 

Yes, something like that would be fine.

 

> Quite a lengthy thing - and how can the correct frequency be determinded? By

> checking the ...

 

I see your point. I did not know, that the new tuners do not have a auto SR anymore. How could they cut such a useful feature?

 

> NIT... yes I saw these three letters here and there in the specs

 

This is how STBs do the scan. I have recorded the NIT from one transponder (ARD). It contains all transponders and all data needed (even SRs). The spec even allows links to other satellites.

post-23-1071057635_thumb.png

Share this post


Link to post
Griga
I have recorded the NIT from one transponder (ARD). It contains all transponders and all data needed (even SRs).

 

Hmm, seems to be something I should consider... thanks! Just read the NIT and create a transponderlist out of it... looks so simple.... I suspect there is a pitfall somewhere. Do all transponders send this NIT? Can it be found on every satellite position?

 

Griga

Share this post


Link to post
Derrick
Do all transponders send this NIT? Can it be found on every satellite position?

 

Griga

no, NIT search is only possible on satellites like astra and hotbird. All search methods that are only based on NIT search fail on most satellites..

 

The NIT can be missing or contains wrong entries in many cases.

Share this post


Link to post
crypto

> Hmm, seems to be something I should consider... thanks! Just read the NIT and

> create a transponderlist out of it... looks so simple.... I suspect there is a pitfall

> somewhere. Do all transponders send this NIT? Can it be found on every

> satellite position?

 

I agree. It seem to be to simple. All transponders carry the NIT, but there are as always exceptions of this scheme. The software in the STBs have defined so called home channels. These are one or two channels per satellite where to begin the NIT parsing.

 

The transponder list are organized in a hierachical way. Some transponders contain only other transponder of the same network. There are linkage descriptors where to find a transponder from the next network and so on. To get a complete NIT it may be necessary to tune into some more transponders. On Astra it is simple. The table I recorded carries 63 transponders, astra.ini has 59 transponders. It seem to be complete.

 

Here is how the table is parsed, if you like to do your own experiments 0x43 is the interesting tag:

 

      NIT ACTUAL 
     SECTION: 0 
      table_id: 0x40 
      section_syntax_indicator: 0x01 
      section_length: 0x3fc(1020) 
      network_id: 0x0001 
      version_number: 0x16 
      current_next_indicator: 0x1 
      section_number: 0x00(0) 
      last_section_number: 0x01(1) 
      network_descriptors_length: 0x4c 
      network_descriptors 
       descriptor_tag: 0x4a => linkage_descriptor 
       descriptor_tag: 0x4a => linkage_descriptor 
       descriptor_tag: 0x4a => linkage_descriptor 
       descriptor_tag: 0x40 => network_name_descriptor 
       descriptor_tag: 0x4a => linkage_descriptor 
       descriptor_tag: 0x4a => linkage_descriptor 
       descriptor_tag: 0x4a => linkage_descriptor 
       descriptor_tag: 0x4a => linkage_descriptor 
      transport_stream_loop_length: 0x3a3 
      transport_streams 
       transport_stream_id: 0x0001 
        orignal_network_id: 0x0085 
        transport_descriptors_length: 0x0d 
        transport_descriptors 
         descriptor_tag: 0x43 => satellite_delivery_system_descriptor 
          descriptor_length(byte): 11 
          descriptor_data(hex): 01 20 70 50 01 92 81 02 75 00 03  
           frequency: 12.07050 ghz 
           orbital_position: 19.2 degree 
           west_east_flag: 0x1 => eastern position 
           polarization: 0x0 => linear-horizontal 
           modulation: 0x01 => QPSK 
           symbol_rate: 27.5000 msymbol/s 
           fec_inner: 0x3 => 3/4 conv. code rate 
       transport_stream_id: 0x0002 
        orignal_network_id: 0x0085 
        transport_descriptors_length: 0x0d 
        transport_descriptors 
         descriptor_tag: 0x43 => satellite_delivery_system_descriptor 
          descriptor_length(byte): 11 
          descriptor_data(hex): 01 17 97 50 01 92 81 02 75 00 03  
           frequency: 11.79750 ghz 
           orbital_position: 19.2 degree 
           west_east_flag: 0x1 => eastern position 
           polarization: 0x0 => linear-horizontal 
           modulation: 0x01 => QPSK 
           symbol_rate: 27.5000 msymbol/s 
           fec_inner: 0x3 => 3/4 conv. code rate 
       transport_stream_id: 0x0003 
       transport_stream_id: 0x0004 
... 61 more stream ids follow ... 13 of them are in the next section ...

 

cu,

 

crypto

Share this post


Link to post
crypto

@Derrick

 

> no, NIT search is only possible on satellites like astra and hotbird. All search

> methods that are only based on NIT search fail on most satellites..

 

I disagree. If it were so, a most commercial STBs couln't do a scan on the satellite. Do you have examples of such satellites?

 

> The NIT can be missing or contains wrong entries in many cases.

 

Hm, the network provider chooses the contents of the NIT table. There may be transponders missing, if the provider wants to hide some channel from public access. This is done sometimes. Sometimes the are also entries of channels, that are not always on air.

 

Anyway this is the DVB standard and I would never say "wrong entries in many cases". Do you imply that NIT parsing is not useable for channel scans?

Share this post


Link to post
Derrick

>All transponders carry the NIT, but there are as always exceptions of this scheme...

 

there are more exeptions than you think.. NIT search is nice as an option and was already more than once mentioned by me <_<

 

..here an example of what I'm talking about:

 

Transponder 16E 11190H

{

Program Association Table [PAT] at PID 0x0000

 

programs

{

program_number 0x0000 -> PID = 0x0010 (is network_PID)

program_number 0x0001 -> PID = 0x1079 (is program_map_PID) (Scopus Network Technologies - CBS-NY MAIN)

}

 

Network Information Table [NIT] at PID 0x0010

{

no NIT found

}

 

Service Description Table [sDT] at PID 0x0011

{

transport_stream_id = 0x00B0 (00/00)

{

CRC32 = 0x14D0047E (OK)

table_id = 0x42 (service_description_section - actual_network)

section_syntax_indicator = 1 (OK)

section_length = 0x003C

version_number = 0x07

current_next_indicator = 1 (is applicable)

section_number = 0x00

last_section_number = 0x00

original_network_id = 0x00B1 (TPS - La Télévision Par Satellite)

service_id = 0x0001

{

EIT_schedule_flag = 1 (present)

EIT_present_following_flag = 0 (not present)

running_status = 4 (running)

free_CA_mode = 0 (all components not scrambled)

Descriptor entries

{

descriptors_loop_length = 0x002B

descriptor = 0x48 (service_descriptor)

{

descriptor_length = 0x29

service_type = 0x01 (digital television service)

service_provider_name_length: 0x1B

character_code_table (provider_name) = 00 - Latin alphabet (ISO/IEC 6937)

service_provider_name = Scopus Network Technologies

service_name_length = 0x0B

character_code_table (service_name) = 00 - Latin alphabet (ISO/IEC 6937)

service_name = CBS-NY MAIN

 

 

no NIT on one of my favorite transponders and the ONIT can be very misleading as well, or why do you think do programs like progdvb like to overwright your precious settings? :P

Share this post


Link to post
crypto

Hi Derrick,

 

I see where you are going <_< . This is one of the babies, they don't want you to watch. I think they even switch pids randomly.

Share this post


Link to post
Griga

OK, ok... if I find some time for it I'll study the NIT and try to parse it. Then we'll see... anyway, a very helpful discussion. Thanks @Crypto & Derrick!

 

Griga

Share this post


Link to post
Griga
descriptor = 0x48 (service_descriptor)

{

descriptor_length = 0x29

service_type = 0x01 (digital television service)

service_provider_name_length: 0x1B

character_code_table (provider_name) = 00 - Latin alphabet (ISO/IEC 6937)

service_provider_name = Scopus Network Technologies

 

P.S. @Derrick: How did the character_code_table get there? The ETSI specs do not mention it... any clue?

 

Griga

Share this post


Link to post
Gioxy

@ Griga

 

...try to take a look to the "Refresh" function in SkyView's "Transponder info" -> "Cache" window (you have to order the providers "by Transponders" before): it refresh in some way the transponder list, I'm not sure if is by NIT, but it may give you some indication for your great TransEdit...

 

Cheers

 

Gioxy

Share this post


Link to post
Derrick
descriptor = 0x48 (service_descriptor)

{

descriptor_length = 0x29

service_type = 0x01 (digital television service)

service_provider_name_length: 0x1B

character_code_table (provider_name) = 00 - Latin alphabet (ISO/IEC 6937)

service_provider_name = Scopus Network Technologies

 

P.S. @Derrick: How did the character_code_table get there? The ETSI specs do not mention it... any clue?

 

Griga

isn't it in the ETS 300 468? <_<

 

A.2 Selection of character table

Text fields can optionally start with non-spacing, non-displayed data which specifies the alternative

character table to be used for the remainder of the text item. The selection of character table is indicated

as follows:

- if the first byte of the text field has a value in the range "0x20" to "0xFF" then this and all subsequent

bytes in the text item are coded using the default character coding table (table 00 - Latin alphabet)

of figure A.1;

- if the first byte of the text field has a value in the range "0x01" to "0x05" then the remaining bytes in

the text item are coded in accordance with character coding tables 01 to 05 respectively, which are

given in figures A.2 to A.6 respectively;

- if the first byte of the text field has a value "0x10" then the following two bytes carry a 16-bit value

(uimsbf) N to indicate that the remaining data of the text field is coded using the character code

table specified by ISO Standard 8859-1 to 9 [5].

Values for the first byte of "0x00", "0x06" to "0x0F", and "0x11" to "0x1F" are reserved for future use.

Share this post


Link to post
crypto

Hi Griga,

Hi Derrick,

 

the character_code_table entry seems to be an artefact. Here is what I get:

 

     SDT ACTUAL 
     SECTION: 0 
      table_id: 0x42 
      section_syntax_indicator: 0x01 
      section_length: 0x3c(60) 
      transport_stream_id: 0x00b0 
      version_number: 0x07 
      current_next_indicator: 0x1 
      section_number: 0x00(0) 
      last_section_number: 0x00(0) 
      original_network_id: 0x00b1 
      services 
       service_id: 0x0001 (1) 
        eit_schedule_flag: 0x1 
        eit_present_following_flag: 0x0 
        running_status: 0x4 
        free_ca_mode: 0x0 
        descriptors_loop_length: 0x2b 
        descriptors 
         descriptor_tag: 0x48 => service_descriptor 
          descriptor_length(byte): 41 
          descriptor_data(hex): 01 1b 53 63 6f 70 75 73 20 4e 65 74 77 6f 72 6b 20 54 65 63 68 6e 6f 6c 6f 67 69 65 73 0b 43 42 53 2d 4e 59 20 4d 41 49 4e  
           service_type: 0x01 => digital television service 
           service_provider_name_length: 27 
           service_provider_name: "Scopus Network Technologies" 
           service_name_length: 11 
           service_name: "CBS-NY MAIN" 
      crc_32: 0x14d0047e

 

Ah and BTW there is a NIT on that transponder

 

     NIT ACTUAL 
     SECTION: 0 
      table_id: 0x40 
      section_syntax_indicator: 0x01 
      section_length: 0x15(21) 
      network_id: 0x00eb 
      version_number: 0x00 
      current_next_indicator: 0x1 
      section_number: 0x00(0) 
      last_section_number: 0x00(0) 
      network_descriptors_length: 0x08 
      network_descriptors 
       descriptor_tag: 0x40 => network_name_descriptor 
        descriptor_length(byte): 6 
        descriptor_data(hex): 53 63 6f 70 75 73  
         network_name: "Scopus" 
      transport_stream_loop_length: 0x00 
      crc_32: 0x88a29865

 

but the transport stream loop is empty. So a "simple" parser might report "no NIT" <_<:

Share this post


Link to post
Derrick

..maybe I'll stream pid 0x0010 later again without the tool :P but with the character set it's exactly as written in the ETS 300 468, isn't it?

 

>- if the first byte of the text field has a value in the range "0x20" to "0xFF" then this and all subsequent

bytes in the text item are coded using the default character coding table (table 00 - Latin alphabet)

 

>descriptor_data(hex): 01 1b 53...

 

0x53 -> S (Latin alphabet) <_<

 

btw. you should consider the table repition rates as well. For me it's difficult to measure but some are apperantly not within the limits as described in dvb.

Share this post


Link to post
crypto

Ahh, tricky! It is an implicit entry. You did your home work. <_<

Share this post


Link to post
Griga
but with the character set it's exactly as written in the ETS 300 468, isn't it?

 

Aaah, now I've got it! I've only read the descriptor description, but not Annex A...

 

Sometimes I saw strange characters in the channel- and provider names and wondered why... finally I simply let them pass through a filter procedure to get rid of the unwanted characters and forgot about it.... <_<

 

Griga

Share this post


Link to post
Derrick

a NIT_search won't never ever yield channels like these <_<

 

10E_transedit.jpg

Share this post


Link to post
Griga

Hey, that's far out... never saw something like this... especially the SID, they are all one, in the true sense of the word. <_<

 

And I'd really like to know in which country they speak aaa and aab...

 

@Derrick: Are you still using an old version? The final release should show a SR column besides the frequency.

 

Griga

Share this post


Link to post
Derrick

..no, it's not that far out. Actually it's closer to astra than hotbird (10E) ^_^

 

yep, it's not the latest version.. but I can switch to the one you released :wacko:

 

@Crypto, you can't blame any parser not to find the NIT (or what's left of it) on the mentioned transponder (16E, 11190H). I've done a complete transponder recording (SkyView) of about 100MB. There were only 12 occurences of pid 0x0010 :P

 

40 -> table_id: network_information_section - actual_network

F015 -> length 21 -> Bytes

00EB -> network_id

C1 -> reserved (2 Bits): 3 -> version_number (5 Bits): 00 -> current_next_indicator (1 Bit): 1

00 -> section_number

00 -> last_section_number

0008 -> network_descriptors_length 8 bytes

40 -> network_name_descriptor

06 -> 6 bytes

Scopus (Latin alphabet)

0000 -> transport_stream_loop_length 0 bytes

 

..and here's what TSReader says (only found 11 :D )

 

16E_11190H_50.jpg

 

ps. nice time and date <_<

Share this post


Link to post
crypto

Awesome! Thank you very much for the update. I really like your program.

Share this post


Link to post
Griga

Well, as it always goes... after having uploaded version 1.5, I got the a hint from Derrick... had a look at the code and the specs again, and found a bug, that in rare cases prevents TransEdit from detecting a channel.

 

So I just re-uploaded it. I think the scanner engine now is - no, not perfect. That would be boring. :bounce: Working quite well. That's it.

 

@Crypto: Sorry, I will reply to your mail soon. I was quite busy with other things, especially the DVBSource filter.

 

Griga

Share this post


Link to post
crypto

Hi Griga,

 

no rush! I noticed what is going on here the last days and you all did very well.

Share this post


Link to post
Derrick

Hiho :D

 

I just made test on 42E and now everything seems almost perfect :D ..more transponders and imho faster as well :bounce:

 

The one transponder that gave me so much trouble was read in no time ;)

Share this post


Link to post
crypto

Hi Derrick,

 

Good, that you know the specs so well.

Share this post


Link to post
Derrick

..not by heart. It's all in the ISO but you'll have to find it :bounce:

Share this post


Link to post
damjang

I report here, but I don't know if this is a TransEdit problem or provider problem. When I scan the Trasponder Mediaset on Hotbird (11388 H 27500) sometimes TE found 2 channels, sometime nothing (all red!).

 

Regards

damjang

Share this post


Link to post
KGM11

I have a similar problem, but it's quite random and not limited to a single transponder. The scan process sometimes just doesn't find anything on a 'known good' transponder (PAT, PMT, SDT turn red), but if scanned again it works fine. Not a problem if I want to scan single transponders, but when doing a full satellite scan it's quite annoying because you never know if all transponders have been scanned properly. My other card (Technotrend DVB-s 1.3) doesn't exhibit this problem. I've also switched the cables between the two cards (one has a slightly better signal level because the other one is hacked up out of two connected runs of cable), but the TT card works fine on both cables while the Skystar (2.6B) shows the described behaviour on both.

 

I don't know if it's a hardware or software problem, but maybe it would be possible to add an option to TransEdit that causes it to re-scan transponders on which it didn't find a signal/any services on the first try?

Share this post


Link to post
Gioxy

Sometimes Mediaset's channels are encripted, because they don't have a international permission to transmit some material (like some film and some sport event).

Moreover they have a large bandwidth consumption (I don't know why...) and are the slowest channel to visualize on my PC... maybe changing some TransEdit scanning parameter will do the trick...

 

Cheers

 

Gioxy

Share this post


Link to post
Griga
maybe changing some TransEdit scanning parameter will do the trick

 

Probably... if PAT, PMT and SDT turn red, I'd try to increase the tuner lock and PAT timeout, as described in the ReadMe. The default values have been tested with Astra 19° East. However, on other satellite positions the conditions may be different. Some broadcasters have a idiosyncratic way to implement the DVB standard... :idiot:

 

Griga

Share this post


Link to post
Derrick

The problem that you have to rescan some transponders isn't solved yet. IMHO nobody knows whether this is a driver related problem or something else. Sometimes the tuner simply won't lock but on a rescan the tuner in some cases will lock immediately. Alas a good indicator for the qpsk lock is missing..

Share this post


Link to post
Guest Oliver

I encountered this problem once when scanning for Euro1080. Sometimes 2 channels were found, sometimes only one and once nothing.

 

Bye, Oliver

Share this post


Link to post
Griga

@Everybody who encounters this problem: Which driver version / SkyStar2 hardware rev. do you use?

 

Griga

Share this post


Link to post
Derrick

the latest driver (4.28) and rev. 2.6B

Share this post


Link to post
Job48

Possibly not very helpful my rmark, but it seems to me, we are dealing with a nature constant here. I encountered the same problem with the pinnacle hardware, I used before. Looks like computers have the same problems as humans (myself) not finding thing on the first run. One has to look again... :idiot:

 

So the update function is of importance, adding new channels on rescan, without deleting the ones already found. One can do fine this way, I think.

Share this post


Link to post
Derrick

Apparently there are 2 different not related problems.

 

1. the scan of a transponder is successful but sometimes not complete

 

2. scanning a transponders fails but a rescan is succesful

 

Maybe the 1st case can be improved by increasing the time out for the tables slightly (@Griga could comment on this case)

 

The 2nd case could be more complicated. I have a theory but no proof of course :idiot: IMHO it could be related to the symbol rate. If the symbol rate is almost constant like on astra, no problems will occur. The bandwidth of a a modulated qpsk carrier depends on the symbol rate. For low SRs filters have to be changed in the frontend (and for high BWs as well). The FE only gets parameters for frequency, pol, and SR. The filters have to be automatically changed by the FE. If a I2C command was known to enforce this, the qpsk lock could be achieved faster and more reliable in case of different SRs. The list is sorted according to frequency but not to SR so on other satellites the SR jumps all the time.

Share this post


Link to post
crypto

As Derrick has pointed out, there is a large mix of SRs on hotbird and others. As far as I know, TransEdit relies on Auto FEC. I think by supplying the FEC found in the NIT, the tuner can lock faster and more reliable.

Share this post


Link to post
Griga

I would like to get some feedback from TransEdit users....

 

Usually a complete TransEdit scan of Astra 19° East takes about 4 minutes on my PC, with hardware rev. 2.3 and the 4.22 drivers. Yesterday I installed the latest TechniSat drivers 4.28. Now it takes 6 minutes :( anybody else experienced that? How long does it take on your PC?

 

Griga

Share this post


Link to post

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...