[mythtv-users] IR Blaster Inconsistencies With Recordings Only

[Nemo Inis]

Jeff wrote:
> On Fri, Jan 23, 2009 at 2:25 PM, <scott at tbwifi.ca> wrote:
> 
>>> scott at tbwifi.ca wrote:
>>>>> On Jan 23, 2009, at 10:45 AM, scott at tbwifi.ca wrote:
>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> I am running 0.21 fixes on a Mandriva 2008.1 box with an Asus M2N-
>>>>>> Sli MB.
>>>>>> My satellite receiver is an Expressvu 4100 (Dish).  I am having
>>>>>> trouble
>>>>>> getting my ir blaster arrangement to work consistently.  Oddly if I
>>>>>> request a channel number using the script from the command line
>>>>>> everything
>>>>>> works fine.  Also if I manually key in a channel number and hit enter
>>>>>> using the keyboard while watching live tv in the frontend, it works
>>>>>> too.
>>>>>> The issues seem to always occur with programmed recordings.  Is it
>>>>>> possible for the database to be scrambled somehow causing this?  Does
>>>>>> anyone have any other tips to look for?
>>>>> It's possible that the permissions aren't appropriate for the use that
>>>>> is running mythbackend. Try logging in as that user and executing the
>>>>> channel change script from the command line. At least, without backend
>>>>> log files or more of a description of what "trouble...to work
>>>>> consistently" means that would be where I would start.
>>>>>
>>>> My mention of success running commandline entries was done using the
>>>> mythtv user, and permissions aren't a problem.
>>>>
>>>> The trouble I mention seems to usually be missed digits or what appears
>>>> to
>>>> be that.  If the intended program is on 210, sometimes two digits get
>>>> through and the sat box just assumes an error and flips to a totally
>>>> non-related default.  On one occasion last night it seems to flip to a
>>>> three digit channel correctly, but not the channel that was expected.
>>>> As
>>>> I mentioned, if I key in 210 in the frontend using the keyboard, I get
>>>> tuned to 210.  I'm not sure if this is something people have run into
>>>> before.
>>>>
>>>> Take note that about 50% of the recordings do get recorded correctly.
>>>>
>>>> I cannot guarantee that my commandline successes are a fluke, but they
>>>> do
>>>> seem to show an almost flawless success rate.
>>> You might want to look into the timing of the numerals being sent by the
>>> blaster, but since the script works from the command line this is
>>> probably not the problem.
>>>
>>> In any case, you might try inserting delays and see what happens, or
>>> reduce the delay if you are using them now.
>>>
>>
>> I've played with those timings extensively and have also tried adding
>> variations of requesting the guide first, adding cancels, and selects  You
>> name it, I've done some serious muddling in the script with little or no
>> improvement.
>>
>> One thing I did wonder is how one requests a longer illumination time as
>> opposed to the delays inbetween.
>>
>>
>> I should also note that I am feeding the signal from the satellite box in
>> through the coax input of a PVR-350 if that provides and ideas for anyone.
>>
> Scott,
> 
> If it works 100% of the time from the command-line, then there is no reason
> for it to not work from myth.
> 
> Go into "Live TV" and switch between a bunch of channels - does it work fine
> then?  If it does, then there's no reason for timed recordings to fail.
> 
> So if that works perfectly, I'd start looking into what other things would
> cause interference with the IR signal.  Things like bright sunlight, or a
> fluorescent light close by (or in some cases, even across a room), -
> anything that could interfere with the IR signal or cause erroneous IR
> detections.
> 
> I found that the only times I miss channel changes are if my track light is
> moved and aimed right into my receivers, thus flooding the poor receiver's
> IR detector with light (including IR light).  (My receivers and myth server
> are on shelves up near the ceiling, and if I move the track light to
> illuminate the inside of my myth box, it floods the receivers too, so I have
> to remember to move it away from them when I'm done...)
> 
> Oh yeah, you can also try raising the lirc "min_repeat" value -- if it's "2"
> now, put it to "3" or "4".  That basically simulates you holding down the
> number key on the remote a little longer.
> 
> J-e-f-f-A
> 
> 

I was going to suggest the same thing (lighting issue). I found that my early morning recordings
where prone to missing the channel due to IR error.  After lots of experimentation, I found out that
at that time of the day, sunlight was hitting an IR target in an upstairs bedroom (I had IR
repeaters from my pre-mythtv days) which was then transmitting garbage back to the satellite
receivers downstairs.
Try keeping your IR target away from excessive light, see if that cures the problem.

Also, make sure you do everything Dan Wilga tells you to, and use his scripts (and IR data). The man
knows his stuff... :-)