[mythtv-users] Mythcommflag in Italy or Europe

Brian Wood beww at beww.org
Wed Feb 15 16:54:40 UTC 2006


[Top-Posted intentionally, as it seems appropriate for this]

Thoughts on commercials from somebody who used to air them :-)

Most people are not aware of the different "types" of commercials  
that are encountered in TV programming, and some are a lot harder to  
"detect" than others.

The hardest to detect are the "self-integrated" spots on syndicated  
programs, or network-sponsor commercials in network programming.  
These spots were "integrated" into the program by the syndicator, and  
are usually very smooth in their transitions to and from the program.  
These are spots that the TV station is contractually obligated to run  
as a condition of purchasing the program, and, in a case where the  
station pre-empts the program (say for coverage of a local news  
item), they must then edit out the spot and run it separately,  
sometimes in a specific time frame. These spots include ones  
integrated into the network feed by the network, sometimes even by  
the program creator. These spots are not necessarily 30 or 60 second  
increments, and can be VERY difficult to detect sometimes.

Then there are locally run spots. A local station breaks away from  
the network feed or the syndicator's tape, usually with a synchronous  
switch, but with no real consideration of program content. You will  
usually get some black frames and often radical scene change. These  
spots are predictably 30 or 60 seconds in duration and much easier to  
detect that the above ones.

Probably the easiest to detect are the local spots inserted by your  
local Cable or Satellite provider on a Cable Network feed. These are  
inserted in "local availabilities" which usually occur at predictable  
times. The programmer usually provides some sort of signal to  
indicate the avail (in the old days this was a series of DTMF tones  
several seconds before the break to allow for the "pre-roll" time  
requirement of a VTR). The best systems at least switch  
synchronously, but many do not even do that, and the switch can be  
pretty crude, often with large changes in sound level and video  
quality.  Almost always these are 30 or 60 second increments, VERY  
easy to detect.

On Local stations you will also get "bumpers" or "pod-breakers" of 5  
or 10 seconds to break up a long commercial interval, and "Legal  
IDs", which must include the call letters and the City of the  
Station. These can make the total break not an even-30 or 60 second  
interval but if you can identify them they can be subtracted from the  
total and the result will usually be an even number of 30-second  
periods.

The Network habit of putting up a "bug" or "logo" upon return to  
programming can also help a lot in identifying commercials.

Anyway, just in case anyone was interested.

On Feb 15, 2006, at 9:16 AM, Marco Ridoni wrote:

> Steven Adeff wrote:
>> On 2/15/06, Paul Andreassen <paulx at andreassen.com.au> wrote:
>>
>>> On Wed, 15 Feb 2006 08:00 pm, Chris Hembrow wrote:
>>>
>>>> On 2/15/06, Pshem Kowalczyk <pshem.k at gmail.com> wrote:
>>>>
>>>>> On 15/02/06, Marco Ridoni <m.ridoni at cicero.it> wrote:
>>>>>
>>>>>> Is someone successfully using mythcommflag to remove  
>>>>>> commercials from
>>>>>> recording in Italy or in other European countries? I had to  
>>>>>> disable it
>>>>>> since it apparently couldn't flag a single commercial even with
>>>>>> "aggressive" settings.
>>>>>
>>>>> Not sure about Europe, but here in NZ I have similiar problem -  
>>>>> some
>>>>> of the adds are not getting marked properly (begining or end of  
>>>>> the ad
>>>>> block is not recogniesed on some channels).
>>>>>
>>>>> regards
>>>>> pshemko
>>>>
>>>> I have the same problem in the UK. Some ads aren't matched, or  
>>>> it'll
>>>> match the start of the ads, then match the end of the channel
>>>> "trailers" as the end of the adverts, so I still have to skip the
>>>> adverts.
>>>>
>>>
>>> Could you please have a look at the output of a run of  
>>> mythcommflag, it may
>>> give some ideas of the problem.
>>>
>>> export DEBUGCOMMFLAG=1
>>> mythcommflag -c 1002 -s 20050925223500 --verbose commflag --force  >
>>> commflagout.txt 2>&1
>>>
>>
>>
>> In the US and I'd say its 95% accurate on about 20% of the channels,
>> 70% accurate on about 30% of the channels and 30% accurate on the
>> other 50% of the channels.
>>
>> All of this is due to how the different channels handle switching to
>> and from commercials from the actual show. Some make mythcommflag's
>> life easy some make it quite difficult.
>>
>
> Yes, by reading the lists I got the feeling the situation was more  
> or less this, but nevertheless I wanted to know,
> before making more attempts, if it was worth trying.
>
> I'm going to do a few tests with the parameters Paul suggested and  
> then report back to the list
>


More information about the mythtv-users mailing list