TAP for Multi-Event INF

[DeltaMikeCharlie]

Greetings from Australia and apologies for not posting in German.

I have written a TAP called INFplus that can save multiple EPG events associated with a single recording. This is unlike the existing Topfield recording information that can only store the information for 1 event.

INFplus monitors recordings in progress as well as the current EPG event for the channel being recorded, when the current EPG event changes, INFplus saves the new event information into an INF+ file for use later on.

When you playback one of these recordings and press "Info", the standard Topfield information box is hidden and a new screen is displayed listing all of the EPG events that were broadcast whilst the recording was in progress.

A sample of what the INFplus screen looks like is attched.

As well as viewing the event information, you can also jump to near where an event is supposed to start and also "Remember" the starting location for each event, sort of like a bookmark.

The TAP is still in the development stage and I am looking for someone in Europe to help test my TAP in Europe and to provide a translation of the on screen interface.

If you would like to help test, you need to have a good level of technical literacy with the Topfield for troubleshooting purposes. If you would like to help translate the interface, then a good knowledge of English would also be necessary.

If you are interested, please send me a private message via this forum.

Regards,
DeltaMikeCharlie.

[Twilight]

wellcome on board!

a good start thanks!

will it be available on taptodate?

twilight

[xyzzyx]

[quote=""DeltaMikeCharlie""]
A sample of what the INFplus screen looks like is attched.
[/quote]

I think the attachment was missing.

That seems to be a solution for the idea of jkit that was mentioned here to improve ScanForDD51. That means to combine the Dolby Digital 5.1 detection with the recognition of the event change.

I would be interested how you recognize the change of the event. I think you do it during the recording. Would this solution be also possible during playback of the file or by reading the recorded file from disk?
Or what is the format of the INF+ file so that I could read out the bookmarks you set for the event changes. If I would read the standard bookmarks I would not know whether they are set by your logic.

If I have some time left I could also offer the testing and translation.

[DeltaMikeCharlie]

Twilight,

Thanks for the welcome. I do plan to release INFplus on TAPtoDate once it has had some more testing. It has only been tested in Australia by 2 people on 3 PVRs. It could still be dangerous to release "into the wild" just yet.

xyzzyx,

I have attempted to attach the screen shot again. If this does not work, I'll provide a hyperlink.

Detecting event changes is very basic, I poll with TAP_Hdd_GetRecInfo() periodically and then I use FireBird's new EPGInfo functions to get the detailed EPG information where I look for a number of field changes.

Your question about playback is very interesting, but I am not sure if the EPG data is written into the MPG TS file on the PVR. In the Australian PVRs, it does not appear to be. As far as I know it is only Video, Audio and Subtitles that are saved.

INFplus does not create standard Topfield bookmarks. It writes EPG data into a file called INF+. There is at least 1 record for each event as there can be duplicates when broadcasters update the EPG at the last moment. When an INF+ record is written, I save the TYPE_RecInfo->recordedSec as well as the MPG_file_size / 9024 and all of the EPG data. The MPG file size is used for the block offset for TAP_Hdd_ChangePlaybackPos().

In early experiments, I found that I could actually modify the standard INF file so that it could contain the INF+ data without interfering with the standard operation of the PVR. I decided to abandon this method and use the INF+ because then it would be an add-on, not a modification to the native system.

DeltaMikeCharlie.

[Wolfman]

Hi,

I am only a user, not a developer or coder but concerning EPG-Data in the MPG-TS Stream:

On the PC I use "TS Doctor" to cut the rec Files.

This Software has an option to find cutting marks by using the EPG Data which works quite fine with the Toppy Recordings and so I guess the Data is stored in the TS-Stream on german Toppys at least.

[FilmFan]

[quote=""DeltaMikeCharlie""]Your question about playback is very interesting, but I am not sure if the EPG data is written into the MPG TS file on the PVR. In the Australian PVRs, it does not appear to be. As far as I know it is only Video, Audio and Subtitles that are saved.[/quote]
I am a developer and I can confirm what Wolfman wrote: The .rec files recorded by the german models contain any EPG data sent on the current transponder or channel.

By the way I am interested in the detailed data structure of your INF+ files, because it would be nice to be able to show the extended information with my TFInfo program on a PC.

[xyzzyx]

Yes, would be also nice for me to read the already created INFplus to jump directly to the EPG changes positions and look additionally or at first near them for the DD51 change.

[DeltaMikeCharlie]

FilmFan,

I will be happy to share the INF+ format with you once I have finished the development and testing of the TAP. I see that having the INF+ data available on the PC too as being very useful.

At the moment, the extra EPG data is kept in a separate INF+ file on the PVR and does not modify the native INF file. This was so that INFplus acted more like and add-in than a modification. To get to this data, you need to know the iNode number for the INF file of your recording.

I did some early research into the INF and the behaviour of the (Australian) firmware and I found that I could modify the INF so that it retained compatibility with the native firmware and was able to contain the INF+ data.

I'm thinking about adding a "Prepare to Archive" feature to INFplus that will take the INF+ data and modify the INF so that when it is archived to external media, the INF+ is retained. INFplus, or any PC program, just needs to recognise this bigger INF and then read the INF+ data from it.

Would you be interested in testing the compatibility of the extended INF on European firmware?

xyzzyx,

I do not currently keep track of the DD51 information in the INF+. As far as I know, in Australia, our channels do not change audio modes between events, a channel is either permanently Dolby or it's not (although I may be wrong).

This is not a big problem though. When I find a recording in progress, I can check the TYPE_TapChInfo and save the dolbyTrack into the INF+. There are also still a number of "Unknown" fields in FireBird's struct for the EPG data, perhaps one of those contains the Dolby information for the event.

All,

INFplus has also been designed to integrate with other TAPs and I have worked with the author of TMSArchive on integrating with that TAP. Another TAP can call INFplus, passing it the correct parameters and INFplus will pop up and display the INF+ for the required file.

Regards,
DeltaMikeCharlie

[xyzzyx]

[quote=""DeltaMikeCharlie""]
xyzzyx,

I do not currently keep track of the DD51 information in the INF+. As far as I know, in Australia, our channels do not change audio modes between events, a channel is either permanently Dolby or it's not (although I may be wrong).

This is not a big problem though. When I find a recording in progress, I can check the TYPE_TapChInfo and save the dolbyTrack into the INF+. There are also still a number of "Unknown" fields in FireBird's struct for the EPG data, perhaps one of those contains the Dolby information for the event.
[/quote]
No, I meant I will care about finding the exact position of the DD change with my ScanForDD51 tap. I would only need to have the position of the event change to narrow down the possible positions in the file to look for a DD change.
I use AudioTrackInfo from FireBirds lib to get the general information whether it is DD (AC3) or not and then the stream is read from the file to find the dolby version.

All,

INFplus has also been designed to integrate with other TAPs and I have worked with the author of TMSArchive on integrating with that TAP. Another TAP can call INFplus, passing it the correct parameters and INFplus will pop up and display the INF+ for the required file.

Yes, or instead of the popup (or additionally in another call interface) you could give back the data in a struct. For me it would be enough to get the real positions of the event changes in the file (e.g. as data type long64, in amount of bytes) and would use a fseeko64 command to find the position in the file.

[DeltaMikeCharlie]

xyzzyx,

Now I understand.

I do not save the actual file size, I divide by 9024 before saving. This provides the block number for the TAP_Hdd_ChangePlaybackPos() when I jump to an event. You can easily take this number and multiply by 9024, but then you will loose some accuracy.

Also, INFplus polls for event changes and it can be configured to poll for changes from between 10 seconds to 5 minutes. In the most extreme situation, the INF+ event may be written 4 minutes 59 seconds after the actual event changed. This would also decrease your accuracy.

For each INF+ record, I store the number of seconds recorded so far plus the block size mentioned above. To narrow down your fseeko64 offset, you could look for the 2 closest INF+ records and then interpolate the offset you need. I use a similar method when working out exactly which block to jump to.

The INF+ can also contain duplicate EPG event data if a broadcaster updated the EPG data for an event in progress. INFplus flags duplicate data when it loads prior to display, you would also need to know how to identify duplicate data and which ones to keep.

I am happy to share the INF+ structure with you, but please understand that the structure may change once the European testing is complete. To obtain some INF+ data samples, naturally you would need to get a copy of INFplus for testing.

If you would like a copy of INFplus to test, send me a private message with your e-mail address and I'll send you a copy with some instructions.

[Mucki]

Is it possible that InfPlus doesnt work on external Storages ?

[DeltaMikeCharlie]

Hi Mucki,

I posted my answer in the main INFplus thread.

https://forum.tms-taps.net/viewtopic.php?t=47260