Subtitles out of sync on recorded programs on ABC

[Teddles]

The next series of test drivers are now available for you to try/test.

Ian, my V2 units with stock f/w are already on the second most recent driver version on @Captain's list - 20201218.
The newest version listed - 20250218, was trialed with experimental image version 7.6-alpha which I run some days back without success. So only the 20201218 version to try? I've got a vanilla version of image 6.4 installed; OK to use that for the test or would latest 7.5.1 be preferable?

Teddles

[Captain]

I understood it as: Live TV works, but recordings are not okay.

You sent recordings to Ian, and he tested them on the T4 and U4 but couldn’t reproduce the issue. However, this isn’t a direct comparison since the hardware and drivers are different, even though the Enigma2 version is the same.

I also understood that IanB’s recordings on his V2 from the same channels work fine. Have you checked his recordings on your setup? It might be useful if both of you record the same broadcast at the same time and then swap the recordings to see if the behavior changes.

Additionally, it could help to analyze the recordings with TS-Doctor to check for errors in the DVB file. Issues could arise due to reception, the box itself, or even the hard drive being used.

Here in Germany, we tested recordings with DVB subtitles as well as teletext subtitles, and everything worked fine.

To answer your question: Since 6.4 is EOL, the text would be advisable with 7.5.1 or 7.6.

[Captain]

tsanalyze_1.0-r0_cortexa15hf-neon-vfpv4.ipk

(96.25 KiB) 9-mal heruntergeladen

tsanalyze "/media/hdd/movie/20241017 2200 - ABCTV HD - Fake Or Fortune_.ts"

PAT
section_length: 77
transport_stream_id: 547
version_number: 16
active: 0x1
program_number @ PMT_PID
0 @ 0x10 (16)
672 @ 0x102 (258)
673 @ 0x100 (256)
674 @ 0x101 (257)
676 @ 0x104 (260)
675 @ 0x103 (259)
677 @ 0x107 (263)
678 @ 0x105 (261)
679 @ 0x106 (262)
680 @ 0x108 (264)
681 @ 0x109 (265)
682 @ 0x10a (266)
683 @ 0x10b (267)
684 @ 0x10c (268)
685 @ 0x10d (269)
686 @ 0x10e (270)
687 @ 0x10f (271)

active PMT
program_number: 0 => pmt pid 0x100
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 0 => pmt pid 0x101
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 0 => pmt pid 0x102
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 0 => pmt pid 0x103
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 0 => pmt pid 0x104
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 0 => pmt pid 0x105
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 0 => pmt pid 0x106
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 677 => pmt pid 0x107
version_number: 17
PCR_PID: 0x906 (2310)
maximum_bitrate_descriptor: 0x0e len 3
maximum_bitrate_h: 0x0
maximum_bitrate_l: 0x0
components
type @ elementary_PID
0x1b (ISO/IEC 13818-1 reserved) @ 0x203
stream_identifier_descriptor: 0x52 len 1
component_tag: 0x1
video_stream_descriptor: 0x02 len 3
multiple_frame_rate_flag: 0x0
frame_rate_code: 0x6
MPEG_1_only_flag: 0x0
constrained_parameter_flag: 0x1
still_picture_flag: 0x0
profile_and_level_indication: 0x44
chroma_format: 0x1
frame_rate_extension_flag: 0x0
data_stream_alignment_descriptor: 0x06 len 1
alignment_type: 0x2
AVC_video_descriptor: 0x28 len 4
AVC_profile: 0x64
set0_constraint_flag: 0x0
set1_constraint_flag: 0x0
set2_constraint_flag: 0x0
AVC_compatible_flag: 0x0
AVC_level: 0x0
AVC_still_picture: 0x0
AVC_24_hours_video_flag: 0x0
maximum_bitrate_descriptor: 0x0e len 3
maximum_bitrate_h: 0x0
maximum_bitrate_l: 0x0
AVC_timing_and_HRD_descriptor: 0x2a len 2
HRD_management_valid_flag: 0x0
picture_and_timing_info_present: 0x0
N: 0x300971f
K: 0xef08de2
fixed_flame_late_flag: 0x0
time_POC_flag: 0x1
video_display_conversion_flag: 0x0
reserved: descriptor 0x97 len 0
0x03 (ISO/IEC 11172 Audio) @ 0x28d
stream_identifier_descriptor: 0x52 len 1
component_tag: 0x2
maximum_bitrate_descriptor: 0x0e len 3
maximum_bitrate_h: 0x0
maximum_bitrate_l: 0x0
ISO_639_language_descriptor: 0x0a len 4
ISO_639_language_code: gne
audio_type: 0
0x11 (ISO/IEC 14496-3 Audio with the LATM transport syntax) @ 0x2f1
stream_identifier_descriptor: 0x52 len 1
component_tag: 0x4
ISO_639_language_descriptor: 0x0a len 4
ISO_639_language_code: sua
audio_type: 3
extension_descriptor: 0x7f len 5
descriptor_tag_extension: 0x6
selector_byte: len 4
00000000: 87 65 6E 67 | | | | | | | | | | | | | .eng
maximum_bitrate_descriptor: 0x0e len 3
maximum_bitrate_h: 0x0
maximum_bitrate_l: 0x0
AAC_descriptor: 0x7c len 2
profile_and_level: 0x58
AAC_type_flag: 0x0
SAOC_DE_flag: 0x0
AAC_type: 0x0
0x06 (ISO/IEC 13818-1 Private PES data packets) @ 0x243
stream_identifier_descriptor: 0x52 len 1
component_tag: 0x3
teletext_descriptor: 0x56 len 5
ISO_639_language_code: gne
teletext_type: 16
teletext_magazine_number: 16
teletext_page_number: 1
maximum_bitrate_descriptor: 0x0e len 3
maximum_bitrate_h: 0x0
maximum_bitrate_l: 0x0
0x05 (ISO/IEC 13818-1 Private Section) @ 0x157c
stream_identifier_descriptor: 0x52 len 1
component_tag: 0x10
maximum_bitrate_descriptor: 0x0e len 3
maximum_bitrate_h: 0x0
maximum_bitrate_l: 0x0
application_signalling_descriptor: 0x6f len 3
signalling: 0x0 0x10 0xe1

active PMT
program_number: 0 => pmt pid 0x108
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 0 => pmt pid 0x109
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 0 => pmt pid 0x10a
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 0 => pmt pid 0x10b
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 0 => pmt pid 0x10c
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 0 => pmt pid 0x10d
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 0 => pmt pid 0x10e
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

active PMT
program_number: 0 => pmt pid 0x10f
version_number: 0
PCR_PID: 0x0 (0)
components
type @ elementary_PID

Teletext Sync Issue - PID 0x243 Analysis

I analyzed a recorded transport stream using tsanalyze and found that the Teletext stream is embedded in PID 0x243. However, there seems to be a synchronization issue. Below is an analysis of possible causes and solutions.

Teletext PID and Descriptor

• PID: 0x243
• Teletext Descriptor:
  - ISO 639 Language Code: gne (possibly incorrect, expected something like eng)
  - Teletext Type: 16
  - Magazine Number: 16
  - Page Number: 1

Possible Causes of Sync Issues

1. Missing or Incorrect PCR-PID
- The PMT entry for PID 0x243 does not specify a PCR_PID.
- Without a valid PCR-PID, video, audio, and Teletext may not be synchronized properly.

2. Incorrect or Corrupted Timestamps
- If PTS (Presentation Time Stamps) do not match between Teletext and video, desync can occur.
- The current analysis does not provide PTS values for Teletext—further investigation is needed.

3. Teletext Data Stream Issues
- The language code is detected as "gne", which seems incorrect. It should usually be eng, ger, fre, etc.
- If this is a metadata issue, it might cause improper Teletext handling.

4. Buffering Issues in Receiver
- If the receiver's buffer size for Teletext packets is too large or too small, it can cause Teletext to lag.
- Some receivers require a consistent Teletext bitrate—fluctuations may lead to desync.

[IanSav]

Hi Teddles,

Ian, my V2 units with stock f/w are already on the second most recent driver version on @Captain's list - 20201218.
The newest version listed - 20250218, was trialed with experimental image version 7.6-alpha which I run some days back without success. So only the 20201218 version to try? I've got a vanilla version of image 6.4 installed; OK to use that for the test or would latest 7.5.1 be preferable?

Yes, given you had issues with the latest 20250218 drivers I would go with the next most recent drivers that match the drivers on the Beyonwiz firmware. As they have been working for Beyonwiz we are hoping that they will also make OpenATV work better for you. If these drivers work then @Captain can work on the best solution with which to move forward.

Given that 7.6 is still an Alpha build I think it would be best if you start your testing with a clean version of the production 7.5.1 firmware. While it may potentially have other issues, those issues will be shared by everyone and should be able to be verified by others. Please do your testing without unneeded skins, plugins and settings. Please let us know how you go.

I will ask the other tester to record the show on the service you recently proposed on his V2 and provide samples so that you can test his recordings as @Captain suggested.

You also now have a new tool to analyse your recordings so that we can try and identify any issues in the recording. If they point to issues in the ABC broadcasts then I will present them to the ABC for their consideration and hopefully a resolution.

@Captain, thank you for your assistance and expertise in investigating this issue.

@Teddles, over to you.

Regards,
Ian.

[Captain]

better tool

tsduck_3.21+gitNone0+0fbbc7e064-r0_cortexa15hf-neon-vfpv4.ipk

(4.59 MiB) 7-mal heruntergeladen

https://tsduck.io/

[Teddles]

If they point to issues in the ABC broadcasts then I will present them to the ABC for their consideration and hopefully a resolution.

Just to clarify, Ian, the issue is present with HD broadcast on all of the services. Have just nominated ABCTV HD/'Planet America' as a reference point as per your earlier suggestion.

Teddles

[IanSav]

Hi Teddles,

Thanks for the clarification. I was thinking it was ABC only.

Hopefully the driver tests will shed more light on the issue.

Regards,
Ian.

[Teddles]

Hopefully the driver tests will shed more light on the issue.

Ian, using @Captain's script, have installed the test driver on a plain vanilla version of image 7.5.1:




However, after a reboot the Info panel on the V2 still shows the pre-existing driver:



Have I missed something?

Teddles

[IanSav]

Hi Teddles,

I don't have a V2 so I can't investigate. I will ask @Captain for advice.

Regards,
Ian.

[IanSav]

Hi Teddles,

Did you test some recordings to see what happened?

Did you reboot the box to allow the drivers to load?

Regards,
Ian.

[Teddles]

Did you test some recordings to see what happened?
Did you reboot the box to allow the drivers to load?

Yes & yes - no luck with either

Teddles

[IanSav]

Hi Teddles (and others following this topic),

There has been a lot of work and testing going on behind the scenes in ongoing attempts to find the cause of these subtitle issues.

We currently feel that the issue is not in the tuner areas of the code. We also think that the recording files seem to be checking out okay. We are now investigating the playback code to see if the issue is lurking therein.

There are some code changes we are investigating and testing at the moment. If we see improvements we will make the new code available for further testing.

Regards,
Ian.

[Teddles]

There has been a lot of work and testing going on behind the scenes in ongoing attempts to find the cause of these subtitle issues.

Thanks for the update, Ian ... fingers crossed!

I take it that no further action needs to be taken for now on @Captain's driver installation script?

Teddles

[IanSav]

Hi Teddles,

I think you should stick with the latest, 2025, version of the drivers for now. The changes being tested at the moment are contained within the Enigma2 binary.

Regards,
Ian.

[IanSav]

Hi Teddles,

I hope you have some testing time this weekend. Progress is being made on the subtitle timing issues. I expect that we will have an update you can test within about 24 hours.

Regards,
Ian.

[Teddles]

I hope you have some testing time this weekend.

Hi Ian, I'll see what I can do. Have a busy schedule over the coming days but should at least be able to get the ball rolling when a test image drops!

Teddles

[IanSav]

Hi Teddles,

Sorry for the delays, internal testing did not go well. We seemed to fix Australian subtitles and then seem to have broken them for everyone else.

We are testing some new changes and I will update the thread when I have more news.

Regards,
Ian.

[Teddles]

Sorry for the delays

No probs Ian, happy to have the weekend off

We seemed to fix Australian subtitles

Sounds promising

Teddles

[IanSav]

Hi,

Just as we prepare to submit some code for user testing and some more clean up and fixes are applied. The latest build is being tested as I type this message.

I can tell you that the new test code will be made to the 7.6 Alpha builds. You should also ensure that you are using the latest (2025) version of the Beyonwiz drivers. If you prepare a 7.6 slot the fixes should come in a regular update to that image.

I will post more as it comes to hand.

Regards,
Ian.

[Teddles]

I can tell you that the new test code will be made to the 7.6 Alpha builds.

Ian, the most recent build of 7.6 Alpha (2025-04-02) fails to boot on my V2. Goes into a repeated boot cycle and eventually becomes unresponsive. Had to turn the box off (back switch) and access a working image via a usb stick.
The previous 7.6 Alpha build (2025-03-31) works ok.

Teddles