[mythtv] Perfect sync to interlaced TVs

Paul Gardiner lists at glidos.net
Wed Sep 24 09:06:00 UTC 2008

I have an idea for an alternative form of interlacing to achieve
perfect sync for interlaced content played on interlaced TVs. In
theory the best picture is achieved if the interlaces are displayed
as is, with no processing, but there are problems in doing so
in MythTV.

Here's some notaton:

[i/j] means a frame consisting of field i in odd position, and field
j in even position being presented to the output stage.

i/    means the output stage displaying field i in odd position.

/j    means the output stage displaying field j in even postion.

Here's what you get when all is synced correctly.

   [1/2] 1/ /2 [3/4] 3/ /4 [5/6] 5/ /6

Here's what you get when the frame presentation is synced to the
wrong vsync pulse.

   [1/2] /2 1/ [3/4] /4 3/ [5/6] /6 5/

That looks awful because the fields appear out of order. It
give really bad comb-teeth ghosting.

Here's what you get by using bob deinterlacing, when all is
synced correctly.

   [1/1] 1/ [2/2] /2 [3/3] 3/ [4/4] /4 [5/5] 5/ [6/6] /6

Here's what you get with bob deinterlacing incorrectly synced
(I have to admit I've never seen evidence of this happening, but
it should be the case).

   [1/1] /1 [2/2] 2/ [3/3] /3 [4/4] 4/ [5/5] /5 [6/6] 6/

The fields are displayed in the correct order, but the odd fields
are displayed in even position and the even in odd.

Here's what you would get with the deinterlacing I'm proposing
(call it adjacent pairing maybe). First the correctly synced

   [1/2] 1/ [3/2] /2 [3/4] 3/ [5/4] /4 [5/6] 5/

And here's what you get with adjacent pairing when badly synced.

   [1/2] /2 [3/2] 3/ [3/4] /4 [5/4] 5/ [5/6] /6

Note you still get the correct order and positioning of fields,
just all offset by a 50th of a second.

Is this of any use?


More information about the mythtv-dev mailing list