ATSC Video, Sync - Solved

[sambat]

Recording OTA using Windows Media Centre sometimes there is a noticeable transmission glitch, yet the recorded video keeps sync.
It's still in sync after conversion from .wtv to .mpg, but converting to AVI (after the glitch point), there is noticeable out of sync (sometimes constant - mostly progressive).
What is the mechanism in the .wtv or .mpg that ensures synced playback - and how is it routinely lost when compressed to XviD (using Staxrip, VirtualDub, AutoGk) or when converting the .mpg file to DVD format using AvsToDVD?

[usually_quiet]

This is my best guess...

If you are recording OTA, you are capturing an already multiplexed signal. If the right programs are used to change containers, the audio and video remain in sync because the data is not written to separate files during the conversion process.

If you are converting to a different codec, the audio and video are demultiplexed and written to separate files during processing. When they are re-multiplexed after the conversion process is finished, the multiplexer either doesn't have a frame of reference to compensate for missing data, or was never designed to compensate for missing data.

[sambat]

The stations I record are all in the same general location on the same line of sight.
I can't figure out why, at the same time last year, I could record and convert without problems.
Example, Ch 43 and 44 are problematical and 41 and 45 are ok.
Maybe when the leaves come off, it'll resolve itself.

[edDV]

The stations I record are all in the same general location on the same line of sight.
I can't figure out why, at the same time last year, I could record and convert without problems.
Example, Ch 43 and 44 are problematical and 41 and 45 are ok.
Maybe when the leaves come off, it'll resolve itself.

ATSC transport streams are designed to be robust at the expense of bit rate. In urban areas with tall buildings, multipath is a problem that can be compounded by the user's antenna and tuner selection.
http://en.wikipedia.org/wiki/MPEG_transport_stream

Cable also uses MPeg2 transport streams but the nature of the closed cable environment allows almost twice the bit rate from a 6 MHz RF channel (36 Mbps for QAM vs 19 Mbps for ATSC).

As for leaves dropping, I get nearly twice the ATSC channels here during Nov-March when the local oak trees shed their leaves. Even in Oct I'm now pulling in TV transmissions 130 miles distant for VHF and 75 miles UHF but they aren't yet stable enough for one hour recording.

[edDV]

What is the mechanism in the .wtv or .mpg that ensures synced playback - and how is it routinely lost when compressed to XviD (using Staxrip, VirtualDub, AutoGk) or when converting the .mpg file to DVD format using AvsToDVD?

WTV is post transport stream conversion. The quality of the signal (e.g. audio sync) is mostly an issue with your tuner and the way it deals with lost packets prior to conversion to an mpg container. WTV is a Microsoft container that holds MPeg video and AC3 audio streams but adds DRM and Media Center metadata.
http://en.wikipedia.org/wiki/Windows_Media_Center


PS: I just tested reception and am pulling intermittent UHF stations out to 90 miles but note that I'm at ~3000ft elevation and these transmitters are at 1,000-2500ft elevation hence earth curvature and ground weather effects aren't an issue. I'm using a Wineguard HD7697+CM7777 Pre-amp. I'm locking in VHF CH7 at 126 miles (2 Edge) most of the day. "2 Edge" means two mountain ridges in the path.


[Attachment 9146 - Click to enlarge]
http://www.tvfool.com/

[edDV]

The stations I record are all in the same general location on the same line of sight.
I can't figure out why, at the same time last year, I could record and convert without problems.
Example, Ch 43 and 44 are problematical and 41 and 45 are ok.
Maybe when the leaves come off, it'll resolve itself.

How distant are those stations?
Are the transmitters in clear line of sight?
What antenna are you using?

[sambat]

~20 miles, rooftop rotor.
The line of site takes it through a tall maple about 100' away.
The odd thing is that the ABC affiliate on ch41 is ok, while NBC 44 and PBS 43 are giving problems.
All the transmitters are in the same area - NW of my location.

[edDV]

~20 miles, rooftop rotor.
The line of site takes it through a tall maple about 100' away.
The odd thing is that the ABC affiliate on ch41 is ok, while NBC 44 and PBS 43 are giving problems.
All the transmitters are in the same area - NW of my location.

At 20 miles all should work but the tree may be causing issues. It is possible your problem is multipath or a too strong signal. Try a splitter in the antenna coax to knock the level down on 44/43 as an experiment. Multipath is solved with directional antennas.

[edDV]

You may not know about this tool. You can add the TVfool database to Google Earth (separate overlays for digital TV, analog TV, AM radio and FM radio). Then you can tip back into 3D terrain mode and observe obstacles. This doesn't work as well with flat Detroit.

[edDV]

Also a signal strength plot for each transmitter. This is WDIV-DT (NBC).

[edDV]

As an example, Google Earth can show the hills blocking access to the towers. Walnut Creek, CA is a notoriously poor reception area.

[sambat]

Thank's to usually_quiet and edDV.
I resolved the issue by going over the connections; finally on the roof I noticed the balun had one lead disconnected;the other lead disintegrated when I touched it.
Replaced the balun and it's working the way it did last year (I only record during the new season, so never knew there was a problem during the summer months).