The difficulty in IVTC...

[omf]

Just a general question for those of you more knowledgeable about these things: why is inverse telecining so difficult and seemingly imprecise?

The encoders generating DVDs and other MPEG-format videos obviously intend for decoders out in the world to show these videos properly. And yet, programs like MeGUI seem to have to *guess* at what de-interlacing or de-telecining method will work best to produce the intended picture. Not only that, but a lot of player hardware out there seems to have to guess, too, as I've noticed players having difficulty with particular scenes or entire movies.

Presumably there's a standard for encoding these things that was written with the need to eventually produce accurate decodes in mind. So... why is it so hard, and why does it involve some amount of guesswork?

Just curious,
omf

[Terranigma]

Should'nt this maybe be in the General Discussion or the newbies forums?

[Guest]

IVTC is difficult when soft flags are not used to implement the pulldown (i.e., not MPEG2 with clean RFF flagging). For example, you have an AVI that is hard telecined. Many capture cards capture to AVI. And almost all that capture MPEG2 just treat everything as interlaced and never produce a pulled-down stream. To complicate things further you can have hybrids, where hard telecining is combined with pure video. There is nothing in the stream to tell you which segments are which -- you have to guess. And then to further further complicate things, you can have blended fields, or you can have video overlayed onto film. It goes on and on. The world is complicated; it's not always black and white.

If your display device is interlaced, most of this stuff just looks OK. It's when you want to render it on a progressive display that all the issues arise. DVDs are designed for interlaced devices so all these pathologies are not seen when the output is displayed on an interlaced TV.