[TV] Automatic Detection of Hybrid Cards (4 Viewers)

[mattjcurry]

Hey guys,

This is a carry over from this post (thanks for the suggestion @mm1352000):
https://forum.team-mediaportal.com/...stallation-process.119532/page-2#post-1010439

My suggestion is that we try to tune channels at the same time to see if we can come up with some automatic mechanism to detect hybrid cards rather than relying on the user to figure it out.

My thought was that we could tune the 1st tuner and then go through each one and automatically figure out what the duplicates are. Even better yet would be if we could do it automagically during channel scan. We might be able to speed up channel scanning and do a test for hybrid tuners simultaneously, which would drastically improve the overall user experience.

I am hoping that we could use this thread to tease out the idea.

Can we tell if two tuners are attached to the same slot? If so then we know that only tuners that are in the same slot are eligible to be hybrids of eachother.

Are there any other keys that we could use? For example can you have a hybrid that is QAM/QAM or is it only a hybrid if the tuners are of different types (QAM/Analog)? This would help us narrow down the permutations and spped up detection.

Matt

 

[FreakyJ]

are hybrid cards that often that it is worth the effort?
Anyway: Try and error doesn't seem to be a good option. I hat a short google session, but couldn't find anything. Has someone a hybrid card and can give us the device pathes? Maybe that would give us a first hint

 

[FreakyJ]

[offtopic]
By the way:
maybe we should collect data about which cards are used from our user base. That way we would know which cards are used the most.
[/offtopic]

 

[mattjcurry]

This is mostly about improving MediaPortal usability. If we want to improve usability for US users, then a lot of them will have Hybrid cards. However most people do not even know if they have a hybrid card or not. I know when I setup mediaportal, I had no idea if I had a hybrid card. It took me quite a while to figure it out.

Here are the device paths from my system:

I have another tuner on the same card and the device path is the same all the way up to the last digit for tuners of the same type.

I am not sure how you would know which of the two tuners were on the same port and which two were not.

Also, I was not suggesting that we throw an error that is visible to the user, but that we just trap exceptions during channel scanning and see if we can use that information to determine if there are any hybrid cards. It would only happen once during setup, and then we could remember the settings.

 

[kiwijunglist]

FYI - Most new zealand users who use DVB-T have hybrid cards. My card is 2x[DVB-T + Analog].

However very few people need to setup hybrid mode because no one tunes analog tv and analog tv is being phased out in a few years. The only people using analog and DVB-T together might be people who use either FM radio or an external set top box.

I am not sure what the deal is with satellite cards and if they have hybrid setups

 

[mm1352000]

Thanks for starting the thread Matt

Maybe we could clarify what types of hybrid tuners are we referring to here?

• DVB-C + DVB-T/T2 (eg. DD Cine C/T)
• DVB-C + DVB-T/T2 + analog (eg. Blackgold BGT 36**)
• DVB-T/T2 + analog (eg. Hauppauge HVR22**)
• ATSC/QAM + analog

From a TV Server code perspective the characteristics of a "hybrid tuner" are:

• tuner drivers present two source filter interfaces to the operating system for a single tuner + demodulator pair
• only one of the interfaces can be used at any given time
• the tuner graph must be rebuilt in order to switch tuning modes

I think this comment from kiwijunglist is probably quite true:

However very few people need to setup hybrid mode because no one tunes analog tv and analog tv is being phased out in a few years. The only people using analog and DVB-T together might be people who use either FM radio or an external set top box. (EDIT: NEW ZEALAND)

Matt, your setup with [I guess] analog cable is rarer and rarer. That isn't to say that we should use that as an excuse to do nothing. Just that there are maybe other things we should give higher priority.

With device paths...

We know that we can identify tuners that are physically on the same device from the first part of the device path. For example, all of the tuners on Matt's HVR-2250 will have device path starting @devicenp:\\?\pci#ven_1131&dev_7164&subsys_88800070&rev_81#4&15001d53&0&0008#. This identifies the chipset vendor (1131 = NXP/Philips/Trident), chipset (7164 = SAA7164), product (8880 = HVR 2250), reseller (0070 = Hauppauge), revision (81), and product instance number.

We know that vendors typically use a part of the tuner instance GUID to specify tuner type and instance number. For example, with Matt's HVR-2250 the instance GUIDs look like {e7624e1b-0f75-4048-b65d-f51fb313150X} or {06dad091-e58d-4e4d-9479-334aff77340X}. The X is the tuner instance number. The first analog and ATSC/QAM tuners have X = 0; the second analog and ATSC/QAM tuners have X = 1.

Problem one is that we can't tell that a tuner is hybrid or not hybrid. This means that if a card has three tuners - analog, ATSC/QAM and satellite - we can't tell that the analog and ATSC/QAM tuners are hybrid and the satellite tuner is not hybrid. They will all have tuner instance ID 0 (because they are the first tuner of each kind on the card... but that is not enough.

Problem two is that we can't tell which other tuner(s) a hybrid tuner is related to. As per the above example, we can't tell that the ATSC/QAM and analog tuners are a hybrid pair and that the satellite tuner is independent.

I was thinking automatic tuning tests might help us to identify hybrid pairs. Then I realised we probably won't have any accurate/working tuning details when it matters. In other words, although I think we could tell that tuners were linked if tuning with one "half" of a hybrid pair interrupts the stream from the other half, we don't have the necessary tuning details to perform the test when the tuners are first detected...

 

[kiwijunglist]

Can we ask the manufacturers? Maybe there is a special clue / trick?

 

[FreakyJ]

An idea would be to have a kind of a "automatic mapping button".
after you did all the channel scans you can click on that button and tve will perform its test.
not ideal but doable?!
what do you think?

 

[mm1352000]

Can we ask the manufacturers?

Sure, but...

Maybe there is a special clue / trick?

If there is such a thing then it is likely different for each manufacturer. Even the tuner instance parameter in the GUID is different from vendor to vendor. For Hauppauge it is the last digit. For Digital Devices it is the last digit of the first part.

I'll ask Hauppauge for a start and see if we can get somewhere...

 

[kiwijunglist]

An idea would be to have a kind of a "automatic mapping button".
after you did all the channel scans you can click on that button and tve will perform its test.
not ideal but doable?!
what do you think?

I disagree we don't need a button, All scanned channels should be mapped to all valid tuners at the time they are scanned. What are the situations where you don't want to map a channel onto a valid tuner?