[mythtv-users] interlace issue on TV?

[Tom Dexter]

On Sat, May 2, 2009 at 10:26 AM, Udo van den Heuvel <udovdh at xs4all.nl> wrote:
> Udo van den Heuvel wrote:
>>>
>>> 2009-05-02 15:55:11.720 Failed to approve 'fieldorderdoubleprocessdeint'
>>> deinterlacer
>>> 2009-05-02 15:55:11.720 Couldn't load deinterlace filter
>>>
>>> Hmm. :-(
>>>
>>> Why oh why?
>>> Anybody?
>>
>> Hmm..:
>> http://mythtv.org/pipermail/mythtv-dev/2007-September/057314.html
>>
>> And maybe related:
>> http://www.mythtv.org/pipermail/mythtv-users/2008-May/222097.html
>>
>> How can I verify the true reason for not approving the deinterlacer?
>
> xvidtune shows me a 50.00 Hz refresh.
> I am using the "720x576Over" mode with the openchrome driver on a VT1625
>  TV-out. A Y/C cable goes to the Philips CRT TV.
> The mpegs that MythTV recorded from Digitenne are obviously PAL formatted.
>
> What else do I need to check to find the reason for not approving the
> deinterlacer?
>
> Udo
>

I don't think you're going to get it to approve the deinterlacer with
the code as it is right now.  I'm fairly certain something needs to be
changed to allow it.  That error is coming from this block in
libs/libmythtv/videooutbase.cpp:

            if (!ApproveDeintFilter(m_deintfiltername))
            {
                VERBOSE(VB_IMPORTANT,
                        QString("Failed to approve '%1' deinterlacer")
                        .arg(m_deintfiltername));
                m_deintfiltername = QString::null;
            }

...and that method is doing this:

/**
 * \fn VideoOutput::ApproveDeintFilter(const QString& filtername) const
 * \brief Approves all deinterlace filters, except ones which
 *        must be supported by a specific video output class.
 */
bool VideoOutput::ApproveDeintFilter(const QString& filtername) const
{
    // Default to not supporting bob deinterlace
    return (!filtername.contains("bobdeint") &&
            !filtername.contains("doublerate") &&
            !filtername.contains("opengl"));
}

Someone who understands what's going on there better than I may need
to look at this.  What confuses me is that this looks as though it
would reject bob as well, but obviously it doesn't...which makes me
wonder if it ever gets to this point at all when using bob.

Tom