Video on ACM GS streams

[midwestmac]

Hi I was wondering if anyone was interested in the possibility of maybe viewing Video
on a ACM GS continuous streams.
I have a friend who might be able to make a program for this to sort out all the garbage. He understands TS packets and streams very well.
I started a post on BBheaders and GSE on another forum but haven't got that far with it.
The problem we are having is the need for some 60 sec recording samples That I can't always
get here in N.A.
Like these:
- a true MIS GS stream (with at least two different ISI
values present).(Where the ISSYI would be on or enabled)
- ACM transport stream MIS (Where the ISSYI would be on or enabled)
- CCM-only generic streams (either SIS or MIS)
- Packetized GS (as opposed to continuous GS)

Thanks

 

[midwestmac]

To add to my other post, I have been able to get recordings of ACM GS cont streams
but for some reason my friend says the format of what I'm recording here in N.A. does not seem to match the specs from the "ETSI" .pdf

But if I could get another like this type that might help to compare.

I also have one recording of a Cont CCM which helps a little to understand the flow


I'm using the latest TBS recorder and 6925
Any of the above TS, GS, SIS, MIS, packetized would help
Thanks

 

[satesco]

Hi midwestmac,

If you want, I can give the GS recorded files of different sizes.There is no application or player to restore the contents of a GS stream.
A packetized stream(11120 V,2750,2/3) with Sky Grabber from 27.5W.

 

[midwestmac]

Hello Satesco,
A packetized stream would be great we don't have a recording of one. 60 seconds or 100mb would
probably be perfect. I can't find a packetized stream to record.
If you can record a ACM cont GS MIS so we could compare that recording to what I'm recording.
Getting down to the nitty gritty the bytes and bits with a hex editor, need to compare the frag ID and the protocol type.
with what I have recorded.
As far as recording I'm having better luck with the Tbs recorder and I believe if nothing was selected in the "input stream identify"
It should record the whole raw stream untouched?
This will take some time for my friend to do but, to put it simple toss out all the garbage and get down to a video, if there is one.
If anyone can get a recording of anything with the "ISSYI" and/or "NPD" enabled that would
be great too. I haven't seen one.
Thankyou very much

 

[midwestmac]

I have been reading about BBframes and GSE for about a year. I know it probably doesn't make sense
right now because of the topic I started. But to be able to have all these types of stream samples will make it easier to understand the stream flow
of whats different and why. Then make a program to sort out all the BBheaders and GSE headers.

 

[satesco]

I recorded a GenericPacketized this morning of different sizes, one of 7 mb,and another 100 mb.What puzzles me is that you want to find out the content of this stream when in fact by SkyGrabber we can see this content.
But if you want to look at it, here's your GP files .
Problem is not with this type of stream,because it is transmitted,I think, in the same modulation,and therefore can be analyzed more easily,but with Generic stream!
I hope I am not mistaken,but I think ISSYI and NPD function cannot made enable with TBS or any commercial card with the tools that we use , from the simple reason that they lack an important component of the hardware,which has only DekTec card.
The records that I've done with Generic stream are of course without ISSYI and NPD enable,but I delude myself into thinking that you may be able to learn something important about this stream and thus eventually ,who knows,we can see clearly the content.
You can download them from here .I hope to help

 

[Anonymous_1]

@midwestmac: check your PM for some files, this can help you.

@satesco: As you know, I have a Detktec DTA2137CSXP and I'll be glad to know how to decode continuous streams, please tell me.
According to Dektec, this is not possible (with this card, maybe with an other model ?) I'm ready for testing

 

[midwestmac]

Thank you everyone a million times this is great. I finally have some different recordings we can compare.

The ISSYI and NPD (Null Packet Deletion) flags are two bits found in the first MATYPE byte.
A recording of this type would be useful to study and have to compare the data flow.
Even if it might be a Transport stream.

I wish I had a Dektek card but, With TBS card and dealing ACM Cont GS I would think that some of the switching modulations might be able
to be dealt with, there will probably be some errors though. We will see, and I will forward these files to my friend

Thanks again.

 

[satesco]

Hi Feedsat,
I assure you that I would be very happy if I knew how to decode these Continuous streams!
So far no programmer developed such an application, even those from DekTec, as you know.Not sure that they would disclose the source code if someone would like to make such a program.In this regard, it is to know if CrazyCat would venture to make such an application if he'd have this card?!
But no matter what is the situation now, I am convinced that Generic stream is a challenge for programmers,those who make dvb-s2 cards.For now, the market does not require such a card, but if there are going to be more like us, interested in such streams,surely such GS players will appear on the market.The only condition is that those that want this thing to reach a critical threshold.The rest comes by itself.But our passion is ahead of the technological development,more precisely on the market

 

[CrazyCat]

What you mean when say 'decode GS' Generic continous stream can contain anything, this only info for demod - alway output stream in frame-mode.
DVB-S2 standard propose GSE for incapsulating any traffic in GCS frames. For example DVB IPTV (UDP/RTP IPv4/6) over GSE - in this case you can extract some TS packets from "raw frames garbage" produced by TBS 6925/6983.

Another case - generic packetized stream. Some incapsulated packetized traffic (fixed size packets) in BBFrames. For example IP V4/6 (49160 bits) over normal size frames or Ethernet (12284 bits) over short frames. In this case also possible transmit DVB IPTV and extract some TS packets from "raw frames garbage"

P.S. Btw STV0910ADB have on-chip programmable TS demuxer (PID filtering and remaping) + IP v4/6 and Ethernet filtering for generic packetized (STV0910ADA not have this block).

 

[CrazyCat]

@satesco: As you know, I have a Detktec DTA2137CSXP and I'll be glad to know how to decode continuous streams, please tell me.
According to Dektec, this is not possible (with this card, maybe with an other model ?) I'm ready for testing

Dektec not provide any application for GS processing. Only provide support BBFrames receiving in DTAPI for your applications in L.3 frames format + possible output to ASI port:
DTA-2137 - Retrieving Baseband Frames
But first need upgrade firmware to v2

If you have old DTA-2137 (not 2137C) - this not support this feature (i think only software restriction - firmware updater not allow upgrade to v2). But main disadvantage is old STV0900AAB (with some bugs in MIS, ACM reception).

 

[midwestmac]

Hi Crazycat, I haven't given up on this GSE yet. I think its the unknown that I can't see that keeps me going I guess.

@satesco: Thanks for the recordings. They were close to not having errors. But every DFL (data field length) doesn't line up with the next BBheader. I don't know? Could I get another recording of
7.oE-11180H ? But this time select the "input stream Identify" if its (0) or (21) or whatever and then
"Apply" see if that makes a difference. Or maybe it has something to do with the MODCOD button with the 6983 TBSrecorder that effects the recording.
With hex editor I can find all the 10 byte BBheaders but for example
42 00 00 00 2d 88 co 00 00 95
The 5th & 6th byte DFL 2d 88 = 11656 bits divided by 8 = 1456 bytes until the next BBheader
The Data field length is over lapping or short of finding the next BBframe So there not lining up right for some reason.

Here's a little tool my friend made that checks recordings for errors.
Its a command line program. And you put this .exe program in the same Directory as your recording.
Then from command line
For example: (note: it doesn't like spaces in file name)
C:\BBAnalyzer myrecording.ts > myrecordinglog.txt

It spits out a .txt file
Like this one below "G1712000log.txt"

If you or anyone wants to try another recording that is Thanks

 

[satesco]

Hi midwestmac,

You have two GS files from 11182 H , one with ISI-0 apply, and another one without.Vpid-256.
I hope to help you to analysis and of course, await the outcome of these analyzes.

 

[midwestmac]

Yes! thankyou Satesco, for going out of your way and getting that so quik. This will help big time.
I know we are probably pushing these cards to there limits sometimes, but we can only hope and keep trying.

With a good recording it will help to sort all the headers and look at the whole stream.
And hopefully it will be different than what I'm recording here as far as what the GSE specs. say and help
looking at inconsistencies. Just to read all the different scenarios that can happen in GSE ,WOW what a mess.

We were thinking that maybe if nothing was selected in the input stream it might record another stream if one was
to startup in the middle of a recording, but it looks like it corrupts the recording if nothing is selected
I've tried it both ways sometimes I get lucky and get a "some what" half way good recording with nothing selected there.
As far as recording ACM GS, I posted on TBS website that I can record up to 300mbs without errors,
of course all depends on the signal and that was with (0) selected in the "input stream identify". Then other times I have a hard time recording 30mbs.
It can be frustrating.
I'll forward these recordings to my friend and will let you know .

Thankyou!! very much for doing this and if you ever need anything let me Know.

 

[satesco]

Hi midwestmac,

There are two situations when GS recording is difficult: when the signal is weak,but especially when Data Rate is very low.And there is yet another (little)problem, when you select ISI:when you click apply,at that moment the signal is UNLOCKED even if the signal is strong.Must be left so till the signal is lock up again and recording begins.In this case it is possible that errors appear during recording stream.
I find it interesting the experiment that you do and your friend, so please ask if you need GS files.Only to know what size you need.
In fact, this type of transmission is to be expected in the future.Technically speaking,it is cheaper for program providers.At least for those who use satellite feeds.Yet for us, ordinary consumers,it is difficult to purchase dedicated receivers for these transmissions.Hopefully soon,maybe next year(!?), at least with dvb-s2 cards, to succeed to see Generic Stream without problems.
Best Regards.

 

[BlindFaith]

Did you had a look at the video parameters of the fragments you've extracted? It's mostly low resolution low quality video from internet streaming. The whole thing reminds me of extracting jpegs and and other items in the early days of dvb data transmission ;)

 

[satesco]

Did you had a look at the video parameters of the fragments you've extracted?...

I hope about these parameters is all about, isn't it?

StreamReaderEx-BDA 1.2.2.1257 (Beta)...

A big thank for update StreamReader.With both cards can now make Continous stream via VLC.{It is true without image,but has sound and is jerky,sometimes indistinct}

 

[midwestmac]

A big thank for update StreamReader.With both cards can now make Continous stream via VLC.{It is true without image,but has sound and is jerky,sometimes indistinct}

Nice! I like the new additions to streamreader have not had much time to play
with it yet. Thanks for the post!

Have a good one talk to you soon.

 

[midwestmac]

Hello,
I haven't given up on this yet. I like being able to record with crazyscan now.
I have recorded both ways modcod "all" or sometimes a specific modcod like say QPSK 3/4
For me it depends on the signal I'm looking at.
If I record a signal with modcod 'ALL" I get a few more BBframes That look like this
42 00 00 00 00 80 00 00 00 f7
The Dfl 00 80 (Data field length) is only 16 bytes which I think, is when the modcod is at idle or waiting for the receiving end to say "it wants more". Not sure? I have seen these with the TBS recorder too, not a big deal they could probably be garbage anyway and thrown away,

Lately I've been going over recordings with a hex editor trying to figure out more, to make it
easier for my friend to piece this puzzle together.
The BBframes are the easy part but the GSE headers are a little more trickey
I was told these streams will have to be sorted by either there Fragment ID or Protocol Type

So I have been searching looking for info on protocol types that make sense in are recordings
A GSE header is always right after the 10 byte BBheader here's an example of a GSE header with a popular protocol type "00 02" and fragment ID "0c" that I see many of, and are easy to find

Fragment ID is the 3rd byte and the protcol type is the 6th & 7th byte with a 6 byte label
82 80 0c 05 52 00 02 00 ff fd c1 00 00
And thats following the GSE specs.
The protocol type can also be at the 3rd & 4th byte depends if its a Concatenated PDU

But its the other GSE headers that have different Protocol types then this "00 02" that need to be sorted
to make sure everything is extracted from a recording.
And I can't find any specs yet on the protocol types that are being used all I know is this "00 02" that usually has Pid 256 behind it, comes up a lot.
I have been able to find the Pid 256 associated with different Protocol types but I'm not sure what
they are yet?
Thought I'd share what I have so far. Happy New Year