I don't know if you mean bad signal by "randomly corrupted data" but I'll assume so as I'm not aware that there's any other reason MP would be getting corrupted data.
Stream corruption / randomly corrupted data could be caused many different things - most of the time it is bad signal. For example:
- Long DPC latency caused by some driver (BDA driver has no change to deliver data and some of it needs to be discarded)
- Buggy BDA driver
I'd still think that if the signal was so bad that it couldn't decode the audio, the video would also be affected. In the cases where the audio drops-out for about a second, even if the scene is fairly static if there's someone talking I imagine I'd notice if the video skipped or froze.
Well, I have seen corrupted recordings on my own HTPC where some dropouts were affecting only the audio. Remember that it is completely random where the corruption usually hits. Transport Stream packets are 188 bytes long, and one packet carries only audio or video (or subtitle etc.). So with an extra bad luck you could have 1 hour long period of corrupted audio data and not a single hickup in video stream. Quite unlikely event, but losing 1 second of audio is much more likely (it depends much on the encoding and decoders how much a single lost audio packet will drop the audio stream in the decoding phase).
OK, thanks for explaining. I didn't realise the audio/video was carried in separate packets received at different times. I guess it's possible when the audio drops-out, that the bad signal was actually less than one second but it causes the audio renderer/audio device to lose signal and reset, making the drop-out more obvious.
How do you determine whether it's bad signal or one of the other possible causes though, without spending loads on replacing (possibly unnecessarily) the aerial?
I've just switched from LAV to MS DTV audio decoder and had a very obvious audio problem at around 20:21. I didn't see if it affected the picture as I was at my PC at the time. I notice that TsWriter shows Continuity errors at 20:20:56 - 20:20:57 and evr.log shows Dropping samples at the same time, whilst TsReader shows Continuity errors at 20:20:59 - 20:21:00, which I don't understand as I would have thought any display problems reflected in evr.log would be in sync with TsReader, not TsWriter.
You'll also notice I'm still getting the "PMT Pid wasn't found on the PAT" errors, despite doing a re-scan (on both tuners) and setting the debugoption.
United Kingdom