[mythtv-users] Commercial flagging on a sample basis

[Mike Perkins]

Michael T. Dean wrote:
> On 03/25/2008 02:34 PM, Mike Perkins wrote:
>> Yes. *None of the above works satisfactorily in the UK*.
> 
> Yeah, TTBOMK, none of the above (with the exception of logo detection,
> but in a bit more robust implementation) are actually implemented--so I
> guess they don't work satisfactorily anywhere.  ;)  I was just
> experiencing some deja vu and thought others may want some, too.
> 
> It just seems like every time the issue of commflagging precision comes
> up, we get tons of suggestions and no code, so I figured I'd get the
> suggestions out of the way and let everyone else help me with the no
> code part--though I would love for some one to decide not to help with
> the no code part as some of these may actually be very good for.  :)
> 
Eh, sorry, Mike, I guess my sarcasm meter was unplugged. I merely wished to 
point out that we keep hearing from people that commflagging works ok, when in 
reality it only works properly in US/Canada.

I suspect this is because of the networked nature of the TV market over your 
side. Each station knows exactly how many seconds break they have between 
program segments, etc so divining the rules for commflagging is (relatively) 
easy. I say relatively since I know they are always trying to defeat any 
attempts by users to bypass the commercials.

Over here each channel can basically do what it likes. ITV1 has a sort of 
regional splitup so there are occasions where the local ads are different, but 
in the main the whole country sees the same thing. That means that (1) the ad 
breaks can be any length of time, provided the total is not greater than the 
legally-allowed maximum per hour, (2) ads themselves can be any number of 
seconds long and (3) each channel can do different things to try and defeat the 
likes of us.

Hmm. It is possible that a number of us in the UK could perhaps define the 
problem, even the specifics of how each commercial is made up, but I don't know 
if it would be possible to code against what we find. I might be wrong: the 
problem may be too difficult.

As for other countries - [goes for a lie down in a dark corner].

Mike Perkins