QAM record problem with logs -- HELP please (3 Viewers)

[cewillis]

MediaPortal Version: 1.1.0.0
MediaPortal Skin: Blue3
Windows Version: Xp Sp3
CPU Type: Intel Core I3
HDD: WD 1 TB
Memory: 4 GB
Motherboard: Intel DH57DD
Video Card: none -- on chip
Video Card Driver:
Sound Card: on MB
Sound Card AC3:
Sound Card Driver:
1. TV Card: Hauppauge HVR 2250
1. TV Card Type: Dual Hybrid
1. TV Card Driver:
2. TV Card:
2. TV Card Type:
2. TV Card Driver:
3. TV Card:
3. TV Card Type:
3. TV Card Driver:
4. TV Card:
4. TV Card Type:
4. TV Card Driver:
MPEG2 Video Codec:
MPEG2 Audio Codec:
h.264 Video Codec:
Satelite/CableTV Provider: Western Broadband
HTPC Case: nMedia 1080P
Cooling:
Power Supply: Antec 450 (old spare)
Remote: HAMA clone
TV: Sony 46 LCD
TV - HTPC Connection: HDMI

This is a continuation of https://forum.team-mediaportal.com/general-support-51/mediaportal-htpc-worked-great-until-rescan-96854/, which contains the brief history -- the sum of which is the htpc was working perfectly, until I had to rescan for the QAM channels.

As stated there:
1. all channels can be tuned and viewed without problem
2. any single analog or QAM recording works fine
3. simultaneous or back-to-back analog recordings work fine
4. manually starting ANY TWO recordings (2 QAM, 2 analog, or 1 of each), one right after the other, works fine.
5. BUT -- any attempted QAM recording that starts after any other recording fails:
(note that this 2nd recording SHOULD use the other hybrid tuner - pre-record and post-record are both 2 minutes)
5a. the recording starts, but is heavily pixelated, torn, and/or distorted
5b. within a few seconds - 'Connection to TVserver lost', and TVserver terminates, per task manager.

The recordings for the logs are from 8:28 am to 9:02 am analog, and 8:58 am to 9:32 am QAM.

As seen in the MediaPortal log, there are thousand of pairs of these messages from 8:57 to 10:04. No idea what or why???
2011-05-24 09:04:35.312500 [Debug][MPMain(1)]: d3dapp: RecoverDevice called
2011-05-24 09:04:35.312500 [Debug][MPMain(1)]: d3dapp: DeviceLostException

Thanks for any help or suggestions.

 

[Owlsroost]

At this point I think I'd try doing a clean install of MP including dropping the existing database (which I suspect may have problems).

Tony

 

[mm1352000]

Hi guys

I don't mean to contradict you Tony, but from my perspective there seem to be signal or HDD speed issues to me...

cewillis: Now before you remind me that you're able to do things manually *I have read what you've said in both threads*. The evidence of server-side issues is really obvious though if you take a look at the TsWriter log. The "Discontinuity header bit set!" and "No PayLoad..." messages indicate bad packets are being received from the tuner. Trying to watch a TV stream full of garbage could definitely crash the renderer (which is what the messages that you found in the MP log refer to). I'm not aware of how the HVR2250 aerial connections work but it looks like there are only two connections, so I'm assuming that they're shared somehow. That *can* and *does* sometimes cause problems, even sometimes only with specific combinations or when tuning specific frequencies. In other words, although you may have tested QAM on tuner 1 with analog on tuner 2 that doesn't mean that analog on tuner 1 with QAM on tuner 2 will work. It also doesn't mean that every combination of channels would work. You mentioned that some of your QAM channels changed frequency - perhaps the new frequency or frequencies is problematic for interference and/or signal sharing issues...

PS: you may be desparate to get help, but it is *not* helpful to us to post two threads about the same issue within 3 days...

 

[cewillis]

Now before you remind me that you're able to do things manually

Sorry, but unless there's an explanation for that, the rest doesn't follow. The difference (which is?) between scheduled and manual recording starts might isolate the problem.

I'm not aware of how the HVR2250 aerial connections work but it looks like there are only two connections, so I'm assuming that they're shared somehow.

I don't know what to tell you -- it's a dual hybrid card. To me, that MEANS there are only 2 connections, with a QAM and an analog tuner on each.

In other words, although you may have tested QAM on tuner 1 with analog on tuner 2 that doesn't mean that analog on tuner 1 with QAM on tuner 2 will work. It also doesn't mean that every combination of channels would work.

ALL QAM channels worked before, in every sequence that I used, which is all of them. Now NONE of the QAM channels work in the sequence I specified, neither the ones with changed frequencies or the ones with the same frequencies.
There is no reason to believe this is a problem with the Hauppauge card.

perhaps the new frequency or frequencies is problematic for interference and/or signal sharing issues...

Again, why do the channels with the original frequencies now not work?

but it is *not* helpful to us to post two threads about the same issue

If there was any way to tell that any development member had even seen the first post, I wouldn't have posted the 2nd.

I'd try doing a clean install of MP including dropping the existing database

I'd come to the same conclusion -- BUT I am really uncomfortable about leaving a problem not understood -- it WILL be back someday. I was hoping someone could indicate that I had done something wrong, or at least done something that MP is known not to handle.

MP has a huge amount of capability and flexibility -- way more than I will ever use. But I'll try again -----
Thanks for looking.

 

[mm1352000]

cewillis

I'm trying to help you but it feels like you're annoyed at me for some reason which I can't comprehend. You only gave me a very short section of logs to work with and I'm only reporting what I saw. It stood out like a sore thumb to me. Having seen similar symptoms in logs before, I gave my opinion of what I have seen. If you want to post longer logs then please do. If you'd rather not have my input then that is fine - I can leave you to resolve this with anyone else who is willing to help you.

Now before you remind me that you're able to do things manually

Sorry, but unless there's an explanation for that, the rest doesn't follow. The difference (which is?) between scheduled and manual recording starts might isolate the problem.

Right now I can't explain why that happens - sometimes I just don't have an explanation. If I did then I'd have given it to you. That doesn't mean that the symptom that you're seeing is a reflection of the actual problem (ie. the problem could be something that you don't expect). Like I said: short section of logs... reported what I saw... Thinking further: it could be the sequence in which you're using the tuners. It could be the overlap time - if you tune a tuner with interference present sometimes the demodulator won't correct itself until you take the interference away and force a retune.

I'm not aware of how the HVR2250 aerial connections work but it looks like there are only two connections, so I'm assuming that they're shared somehow.

I don't know what to tell you -- it's a dual hybrid card. To me, that MEANS there are only 2 connections, with a QAM and an analog tuner on each.

I haven't got one of these cards - it was a genuine question. The second connector could have been a loop out for all I know. Many tuners share inputs these days.

ALL QAM channels worked before, in every sequence that I used, which is all of them. Now NONE of the QAM channels work in the sequence I specified, neither the ones with changed frequencies or the ones with the same frequencies.
There is no reason to believe this is a problem with the Hauppauge card.

No, there is probably nothing wrong with the card. I never said there was. What I was trying to point out is that in my opinion the issues you're experiencing stem from the fact that the tuner is producing very "dirty" output. That could be because it is only just picking up the signal because you've picked up the wrong frequencies by scanning with the wrong QAM standard or something like that. In such marginal conditions, having two tuners operating at the same time in such a close environment can be enough to cause the stream corruption that you're seeing.

Again, why do the channels with the original frequencies now not work?

As above. Potentially wrong standard used for scanning. Something signal related anyway.

but it is *not* helpful to us to post two threads about the same issue

If there was any way to tell that any development member had even seen the first post, I wouldn't have posted the 2nd.

Mate, remember that we are doing this in our free time. If your thread doesn't get answered within a time that you are comfortable with that is not a license to post a new one. There are only so many of us and there are a constant stream of threads to answer - it is a wonder we get any development done sometimes. Most of us work full time jobs like normal people. Hardly any of us are in a compatible timezone with you. In short it is easy to miss your thread - please be a little more patient with us next time...

I'd try doing a clean install of MP including dropping the existing database

I'd come to the same conclusion -- BUT I am really uncomfortable about leaving a problem not understood -- it WILL be back someday. I was hoping someone could indicate that I had done something wrong, or at least done something that MP is known not to handle.

My opinion is that will not help, however if you have the time to do that then it can't hurt...

 

[Owlsroost]

Hi guys

I don't mean to contradict you Tony, but from my perspective there seem to be signal or HDD speed issues to me...

No problem - I was hoping a TV Server expert would come along eventually

Tony

 

[cewillis]

I'm trying to help you but it feels like you're annoyed at me for some reason

NO - not at all. Sorry. I AM annoyed at MP when something so simple as a QAM rescan breaks it, when TVs and my cheap DVR handle rescans routinely and easily. That's why I thought I must be doing something wrong.

a very short section of logs

I don't understand this. I set MP to DEBUG logs, stopped MP and TV, deleted all old logs, restarted MP (which restarted TV), let it run through the problem scheduled recording, stopped everything again, and sent all the logs that were produced.
What should I do differently?

The second connector

Whoops -- actually there is only one connector, with an internal divider. My mistake - sorry. Should have said two tuner sections.

What I was trying to point out is that in my opinion the issues you're experiencing stem from the fact that the tuner is producing very "dirty" output. That could be because it is only just picking up the signal because you've picked up the wrong frequencies by scanning with the wrong QAM standard or something like that. In such marginal conditions, having two tuners operating at the same time in such a close environment can be enough to cause the stream corruption that you're seeing.

I certainly agree that a recording produced under the particular problem circumstances is 'dirty', and a tuner is the most likely to be misbehaving. I used the same QAM standard as before -- is there some application that can tell me positively what standard is on the cable? I will try asking the head technician for the cable company.

Something signal related anyway.

Any suggestions? Are particular transmission frequencies known to be 'hard to tune' right?
The fact that manually starting two recordings back to back works is really puzzling. Seems like there has to be some difference in how MP sets up and starts the tuners ---
Compare logs from preset and manual starts?

like normal people

Well, that leaves me out.

In short it is easy to miss your thread - please be a little more patient with us next time...

OK, lesson taken. But question: if a thread is missed, does it somehow get un-missed later?

 

[mm1352000]

Hi again cewillis

I AM annoyed at MP when something so simple as a QAM rescan breaks it, when TVs and my cheap DVR handle rescans routinely and easily. That's why I thought I must be doing something wrong.

Fair enough. I don't think you're doing anything wrong, but if this used to work before then something has changed to cause it to start happening. The things that have changed that I can think of can be counted on one hand:
- tuning details in the DB
- hybrid card assignments in the DB
- channel mappings in the DB
- frequencies of some of your channels

Logically the cause of the issue is one of those things.

a very short section of logs

I don't understand this. I set MP to DEBUG logs, stopped MP and TV, deleted all old logs, restarted MP (which restarted TV), let it run through the problem scheduled recording, stopped everything again, and sent all the logs that were produced.
What should I do differently?

The logs showed the two scheduled recordings and their interactions very well. However you have done much more testing than that including manual recordings and different combinations of tuners. Without having logs to show what happens when things go right then I had to guess the differences between okay (for example, manually starting the recordings) and bad scenarios are...

...actually there is only one connector, with an internal divider...

Hmm, that makes it very easy for the two tuners to affect eachother with interference when they are both operating. They have to amplify the parts of the signal that they are "interested in".

I used the same QAM standard as before -- is there some application that can tell me positively what standard is on the cable?

I don't think so. My understanding is that the differences between standards are not about what the signal contains. More about the gaps between channels or the range of frequencies used. In such a situation it would be really easy to get slightly wrong tuning details. I stumbled across -->this<-- site a month or so ago. I don't think it is complete, and it doesn't seem to list the QAM standard (could be wrong...), but it is the best information that I've found in one place about USA tuning details...

I will try asking the head technician for the cable company.

That is an excellent idea.

Any suggestions? Are particular transmission frequencies known to be 'hard to tune' right?

It is the tuning environment that makes a frequency difficult to tune or not. The factors are things like whether there is noise/interference on a frequency that is close to the frequency that you want to tune, how strong the signal is... From a practical perspective the things with those kinds of effects are weather, aerial/dish alignment, cable quality, cordless phones, electric fences... sometimes really random things.

The fact that manually starting two recordings back to back works is really puzzling. Seems like there has to be some difference in how MP sets up and starts the tuners ---
Compare logs from preset and manual starts?

Yes it is incredibly puzzling. Logs for manual starts would definitely help - in particular, if you can reproduce a problem with only two channels and show with logs that manual tuning works and a scheduled recording with or without overlap does not then that could go a long way to figuring this thing out...

like normal people

Well, that leaves me out.

My turn to apologise - *sorry*... unreservedly. :sorry:
I didn't mean to cause offense and I hope none was taken.

In short it is easy to miss your thread - please be a little more patient with us next time...

OK, lesson taken. But question: if a thread is missed, does it somehow get un-missed later?

No, and that is a fair point. There does come a point when you need to repost, but it is not uncommon for threads to take a week to get answered. I don't know about other people but I'm constantly trawling the areas of the forums where questions come up that I'm often able to help with, and I definitely go back over the last couple of weeks. I also monitor the MP Live section when I'm available and search for unanswered threads. In your case I probably missed the live post and because you added information I missed you in the unanswered threads section too. Hence it is a little unusual for me to have missed your first thread...

 

[cewillis]

Yes it is incredibly puzzling. Logs for manual starts would definitely help - in particular, if you can reproduce a problem with only two channels and show with logs that manual tuning works and a scheduled recording with or without overlap does not then that could go a long way to figuring this thing out...

I'll make new logs for a manual start of the same two channels I used for the last log, using the same procedure: stop all, delete logs, restart, manually start two recordings quickly, wait until recording end, compress and send logs.
OK?

didn't mean to cause offense and I hope none was taken.

No, of course not. I assure you that few people find me 'normal'. (and I'll leave you guessing as to why not)

 

[mm1352000]

Yes it is incredibly puzzling. Logs for manual starts would definitely help - in particular, if you can reproduce a problem with only two channels and show with logs that manual tuning works and a scheduled recording with or without overlap does not then that could go a long way to figuring this thing out...

I'll make new logs for a manual start of the same two channels I used for the last log, using the same procedure: stop all, delete logs, restart, manually start two recordings quickly, wait until recording end, compress and send logs.
OK?

Sounds perfect.