Automatically Tune to Strongest Duplicate Channel (3 Viewers)

framug

Super Moderator
  • Team MediaPortal
  • January 31, 2005
    5,884
    1,956
    South of France
    Home Country
    France France
    I do not use the TV part, because I do not have a satellite, etc., there is only IPTV, but the Media Portal does not support it. :)
    Well ajs, not having a DVB-T, DVB-S, DVB-C or ATSC card should not be a definitve answer for vapourEyes.
    There is no signal for IPTV, then....
    Maybe others MP devs could at least, answer about it ?
     

    vapourEyes

    Portal Pro
    July 31, 2013
    144
    43
    Home Country
    United Kingdom United Kingdom
    Right - on we march into the void....

    I've just tested with weak-channel filtering via a threshold and then strongest based on normalised Signal Strength and Quality.

    I've grabbed the most channels to date, but some weaker ones crept in due to the threshold here of 35 being too low. 40 Seems better. Channel movement detection is off coz I just want to see what it will bring in.

    This thread is a good read regarding tuning.
    Thanks for that !!! (y)
     
    Last edited:

    vapourEyes

    Portal Pro
    July 31, 2013
    144
    43
    Home Country
    United Kingdom United Kingdom
    Best guess so far.

    SetupTv.exe.

    Is anyone able to test ?
     

    Attachments

    • SetupTv.zip
      533 KB

    vapourEyes

    Portal Pro
    July 31, 2013
    144
    43
    Home Country
    United Kingdom United Kingdom
    Signal Strength
    For the signal calculation, I changed a bit my first proposal, since the "3" "-97" reported.
    What you have to know :
    • mDB (1/1000 DB) is always converted as DB then, the -100000 to +100000 become -100 to +100 rank.
    • Positive signal is reported as it in all case.
    • Negative signal is reported by two ways, now :
    • You don't check Nova T-500 like for the cards ==> the -97 value will became +97.
    • You check Nova T-500 like for the cards ==> the -97 value will became +3.

    I only check for:
    - Positive signal is reported as it in all case.
    * You don't check Nova T-500 like for the cards ==> the -97 value will became +97.

    I DO NOT check for:
    * You check Nova T-500 like for the cards ==> the -97 value will became +3.

    *** this could be catered for if there is some form of detection.

    WIth my 2 cards from the 2 different manufacturers the crude current approximation works well.

    ALL 130+ channels are at >60% signal and quality strength. It can drop to 35 or so over the course of the day.
     

    CyberSimian

    Test Group
  • Team MediaPortal
  • June 10, 2013
    2,849
    1,771
    Southampton
    Home Country
    United Kingdom United Kingdom
    I think that I have lost track of what is being proposed here. It seems to me that this thread is confusing two different capabilities:

    Retain strongest
    During "scan for channels", when a duplicate MUX is encountered (can occur when the user is midway between two different transmitters), "TV Server" should retain the stronger of the two MUXes, and discard the weaker.

    For this capability, there is no need for a signal-level threshhold. During scanning, TV Server remembers the signal level of each MUX. When the next MUX is found, TV Server compares the various ids with the previous MUXes that it found during the scan, to determine if the new MUX duplicates a previous MUX. If it does, TV Server retains the stronger of the two, and discards the weaker. Some understanding of how TV Server stores information during the scan is required so that the information for the weaker MUX can be discarded.

    This capability needs to be selectable by the user prior to scanning (i.e. a new setting on the "Scan for Channels" panel in TV Server).

    Omit weak MUXes
    Prior to the scan, the user specifies a signal-level threshhold that TV Server uses to accept or reject MUXes during the scan. MUXes which are weaker than this threshhold are ignored. MUXes which equal or exceed this threshhold are retained. Note that this does not discard duplicate MUXes (retention of duplicate MUXes may be the user's desired outcome, due to different local content on the two transmitters).

    It is important that this threshhold be user specifiable, as what one user considers to be a "weak" signal, another user may consider to be a "normal" signal. Example: there are nine MUXes that serve my location in the UK; all MUXes are broadcast from the same transmitter:

    (1) Three MUXes are broadcast with a power of 200 kW.
    (2) Three MUXes are broadcast with a power of 50 kW.
    (3) Three MUXes are broadcast with a power of circa 20 kW.

    So there are very-significant differences between the signal levels that I receive. If I had a roof aerial, I would want to use a low value for the signal-level threshhold, as I would be able to receive all nine MUXes. But with the loft aerial that I actually use, I would use a higher value for the signal-level threshhold, as the three low-power MUXes can be received, but not at acceptable quality for viewing (frequent pixelation of the picture).

    The user should be able to specify the signal-level threshhold on the TV Server "Scan for Channels" panel. The value should be in the range 0 to 100, with the default being 0. The value 0 is the value that gives identical behaviour to previous releases of MP.

    Conclusions
    Both of the capabilities described above are useful, but they are independent of each other. Either or both could be implemented, and would (I think) improve the product. But as far as I can see, the current change does neither, and so (reluctantly) I must vote against it being included in the next release of MP.

    I may have misunderstood what the new code does; if so, apologies. And of course it is possible for the new code to be reworked to provide either or both of the capabilities described above, at which point inclusion in MP should be reconsidered.

    -- from CyberSimian in the UK
     

    vapourEyes

    Portal Pro
    July 31, 2013
    144
    43
    Home Country
    United Kingdom United Kingdom
    I think that I have lost track of what is being proposed here. It seems to me that this thread is confusing two different capabilities:

    Retain strongest
    During "scan for channels", when a duplicate MUX is encountered (can occur when the user is midway between two different transmitters), "TV Server" should retain the stronger of the two MUXes, and discard the weaker.
    So far I have a threshold, OVER which a channel can overwrite another. It would ideally need to be surfaced to the user.
    Its a subjective threshold due to transmitter strength, weather and maybe other factors.

    Omit weak MUXes
    Prior to the scan, the user specifies a signal-level threshhold that TV Server uses to accept or reject MUXes during the scan. MUXes which are weaker than this threshhold are ignored. MUXes which equal or exceed this threshhold are retained. Note that this does not discard duplicate MUXes (retention of duplicate MUXes may be the user's desired outcome, due to different local content on the two transmitters).
    So far I have a threshold, UNDER which a channel is ignored. Due to the likely hood of it not rendering well.
    Its a subjective threshold again, due to transmitter strength, weather and maybe other factors.

    Conclusions
    Both of the capabilities described above are useful, but they are independent of each other. Either or both could be implemented, and would (I think) improve the product. But as far as I can see, the current change does neither, and so (reluctantly) I must vote against it being included in the next release of MP.

    I may have misunderstood what the new code does; if so, apologies. And of course it is possible for the new code to be reworked to provide either or both of the capabilities described above, at which point inclusion in MP should be reconsidered.

    -- from CyberSimian in the UK

    Indeed noted.

    There is a little bit of UI work to do, but the behaviour outlined above is running very well here. I'm trying it different times of the day and adjusting thresholds too.

    I can genuinely say I have now got the best tuning experience to date.
    *** for my geo-location, current weather etc...

    It would be great for others to try the SetupTv.exe attached in thread and report back good, bad or ugly ?

    Future...

    In addition for the UI it would need 2 threshold settings and a switch to enable 'Smart Scan' or something ?
    Also, now we in theory have stronger signals coming into the system, we could enable auto register channel in EPG grab list.
    Admittedly another UI switch.

    I think there is only a need to maybe maintain the highest seen signal on a card. This could be used to 'normalise' strengths seen on other cards at other times. I would then use card strength to 'order' which card is tuned first. The strongest card should always be tuned last to ensure the strongest channels do indeed over write the weakest ones.

    Code...
    I can post this if needs be.
     

    framug

    Super Moderator
  • Team MediaPortal
  • January 31, 2005
    5,884
    1,956
    South of France
    Home Country
    France France
    I DO NOT check for:
    * You check Nova T-500 like for the cards ==> the -97 value will became +3.
    Then, if you don't take care about it, (or -11 => +89) you could possibly have a strongest signal wich will not be detected.

    *** this could be catered for if there is some form of detection.
    Unfortunately, there is not.
    Rest assured that if there was a way to determine it, it would be done since age.
    It's manufacturer/cards dependant.
    We can imagine that an engineer decided that signal value is -11 because, it's 11 less than max (100%).

    WIth my 2 cards from the 2 different manufacturers the crude current approximation works well.
    Yes but, it's with YOUR cards, you have to think "worldwide"

    ALL 130+ channels are at >60% signal and quality strength. It can drop to 35 or so over the course of the day.
    For you, yes.
    As CyberSimian explained (if any), a switch should (at minimum) be needed.
    For example, I tested your modified setuptv.
    At least, there is no crash but, where I am today, I don't have an antenna which receives different transmitters then finally, what is the gain, for me (and maybe plenty of users) ?
     

    vapourEyes

    Portal Pro
    July 31, 2013
    144
    43
    Home Country
    United Kingdom United Kingdom
    Hi @framug,

    Thank you for testing.

    From the tuning / signal-strength thread, cards fall into 3 categories:
    Low Strength -> High Strength
    0 -> -100
    0 -> 100 \_____ completely opposite (database ?)
    -100 -> 0 /

    Normalising can be applied for 2:
    0 -> -100 \_____ normalise with Maths.Abs()
    0 -> 100 /

    This alone means that 2 cards can give completely opposite results when tuning.

    In the tuning thread. They suggest a database?

    My solution so far is clearly a rough approximation that works very well in my situation, which may not be replicated world-wide.
    Thank you for making me fully aware of the need to reach a global base.

    I am in a logical straight-jacket of the card manufacturers making of sorts.

    There are 3 ways to represent values of strength with 2 completely opposite - making this logically impossible, without some switch for a user to identify the card type they have ?
     
    Last edited:

    vapourEyes

    Portal Pro
    July 31, 2013
    144
    43
    Home Country
    United Kingdom United Kingdom
    Hi @framug, @CyberSimian, @ajs,

    One solution approximation is available in this thread at least.

    Is it sensible to drop this idea and mark it experimental in case anyone does want try it ?
     

    Users who are viewing this thread

    Top Bottom