Rantanen Teemu wrote:
> Rantanen Teemu wrote:
>
>>Nope. Three attempts, still 66-67 seconds each. If I initiate an EPG
>>scan as soon as it's possible from the menu, then the time gets down to
>>26-27 seconds.
>
>
> Hmm, zeroing both 'lastScan' and 'lastActivity' on cEITScanner
> constructor dropped the Conax detection time to 9-10 seconds. Apparently
> tuning into the encrypted frequency does speed up the detection
> process...
>
> I tried also to tune the card into the channel (before VDR main loop)
> using GetDevice() / SwitchChannel() but that didn't help...
>
> I don't know if there are issues if EPG scan is triggered right from the
> VDR start. If not, could this change be included into the next VDR
> version?
I really don't see why tuning to a particular channel should speed up
the initial setup of the CAM communication.
I have three different CAMs here (none of them Conax, though), and
the are all recognized within less than 10 seconds after the tuner
thread started on the individual device. It would have to be a
very Conax specific quirk if they were required to be tuned to
a special channel in order to set up the communication protocol.
The only thing I could imagine that might speed up things for you
is causing the tuner thread to loop more frequently, as stated in
my message from 2005-08-07, 17:22:
You could test the following: change the line
newSet.Wait((ciHandler && (time(NULL) - startTime < 20)) ? 100 : 1000);
to
newSet.Wait((ciHandler && (time(NULL) - startTime < 80)) ? 100 : 1000);
(make the 20 an 80). Does this change anything?
But since you replied to that message that it didn't make a difference,
I'm out of ideas.
Klaus
