Katie Boundary
12th July 2019, 07:03
In the latest episode of "why in God's name would Katie try to do THAT?", I present IVTC without field-matching. Why would this ever be useful? I have no idea. Maybe you're encoding video on a 30-year-old toaster/potato and every CPU cycle counts, or maybe TFM is giving you bad matches because you're dealing with weird content. It doesn't matter. What matters is, it's possible:
mpeg2source("601.d2v").separatefields()
A=selecteven().tdecimate(mode=1)
B=selectodd().tdecimate(mode=1)
interleave(A,B).weave()
I had to add mode=1 to the tdecimate settings due to incorrect decimation around abrupt cuts to black, resulting in incorrect field pairings.
EDIT: there is in fact a slight speed improvement: 29 minutes versus 35 for a more conventional IVTC script. However, there were still incorrect decimations, and therefore incorrect field pairings, during the closing credits because the closing credits consisted of abrupt cuts between still shots held for several frames each.
mpeg2source("601.d2v").separatefields()
A=selecteven().tdecimate(mode=1)
B=selectodd().tdecimate(mode=1)
interleave(A,B).weave()
I had to add mode=1 to the tdecimate settings due to incorrect decimation around abrupt cuts to black, resulting in incorrect field pairings.
EDIT: there is in fact a slight speed improvement: 29 minutes versus 35 for a more conventional IVTC script. However, there were still incorrect decimations, and therefore incorrect field pairings, during the closing credits because the closing credits consisted of abrupt cuts between still shots held for several frames each.