[mythtv-users] Victims...... I mean Testers wanted.....HD transcode script that should keep AC3 intact if it's in the original stream.

Another Sillyname anothersname at googlemail.com
Thu Aug 12 04:20:31 UTC 2010


On 12 August 2010 04:55, Raymond Wagner <raymond at wagnerrp.com> wrote:
>  On 8/11/2010 22:54, Another Sillyname wrote:
>>
>> 5.  Pulls from the mythDB ONLY the HD transmissions and presents the
>> user with a list of recordings and the station that transmitted them.
>
> As always when reading from the database, and especially when writing to the
> database, make sure to check the database schema to ensure you're working
> against a compatible version.

No writing to the DB happens, this script only reads values. (I'd
never even contemplate writing to the DB given my programming
limitations!!).  I haven't coded any schema version checking yet (want
to make sure the core functionality works in other territories first).

>
>> 6.  Prefers to run the transcode on a different machine from the
>> mythbackend (to reduce loading on the main box)
>
> Does this mean you're running outside the normal MythTV jobqueue? If so, you
> should be managing your own entries in `inuseprograms` to make sure you
> don't conflict with anything.

Yes.  This script is a standalone and entries only show up when
recordings are finished.  The script pulls a copy of the original
recording to the local box to work on so shouldn't conflict with
inuseprograms.
>
>> 7.  Detects and uses the maximum 'sensible' number of threads for the
>> transcode, i.e. an I7 chip would utilise 6 threads.
>
> If you simply omit any manual setting for threads, x264 will automatically
> use the 'sensible' number of encoding threads, number of cores + 1.

Wasn't aware of that, my experience in the past with mencoder and x264
was that it only used 1 thread unless told otherwise.

I'd be happy if someone thinks what I've produced has any value and
incorporates it into nuvexport, if I had the skillset I'd do it
myself.


More information about the mythtv-users mailing list