Kayaker
17th May 2007, 14:50
I've got an interlaced 4:3 video that after SeparateFields() still got some blending/ghosting. But not every frame.
I don't think there's a pattern, anyway I don't study it much. The quality isn't that great anyway.
So I was thinking in just encoding it the fast way.
SeparateFields.SelectEven or SelectOdd. ReduceHorizontalBy2()
So as to get same fps than original. Not double. Crappy size, I don't care.
But I was wondering if there was a way to not select always Even or Odd but just select from the 2 the frames which has less blending.
Of course this decision should be dinamically made from the 2 frames at hand not just at the beggining and stay fixed.
And of course some antibobbing filter.
BTW, what antibobbing filter do you recommend ?
Maybe masktools can give me a metric of "how much blending" ?
But I'm kinda lost. Any suggestions ?
Maybe the movement ends being unnatural because I'm choosing frames that are have not been taken temporally equal in distance.
But I'll give it a shot. At least I'll learn something in the way :)
Thanks.
I don't think there's a pattern, anyway I don't study it much. The quality isn't that great anyway.
So I was thinking in just encoding it the fast way.
SeparateFields.SelectEven or SelectOdd. ReduceHorizontalBy2()
So as to get same fps than original. Not double. Crappy size, I don't care.
But I was wondering if there was a way to not select always Even or Odd but just select from the 2 the frames which has less blending.
Of course this decision should be dinamically made from the 2 frames at hand not just at the beggining and stay fixed.
And of course some antibobbing filter.
BTW, what antibobbing filter do you recommend ?
Maybe masktools can give me a metric of "how much blending" ?
But I'm kinda lost. Any suggestions ?
Maybe the movement ends being unnatural because I'm choosing frames that are have not been taken temporally equal in distance.
But I'll give it a shot. At least I'll learn something in the way :)
Thanks.