Originally Posted by gorthocar
I still have a question related to the OP's question:
I'm assuming that standard definition dvds are not originally encoded at 24fps, but at 30fps (60i?).
They can be encoded at 60i with every field encoded, or 24p within a 60i container, with the duplicate field flagged rather than separately encoded. (MPEG2 allows for this 3:2-free 60i compression, with the 3:2 re-introduced after decompression)
Please correct me if that is wrong.
How would a player correctly know which frames to keep or drop to playback standard definition dvds at 24fps?
The same way that a TV would - by detecting the 3:2 pull-down sequence (whether originally recorded or reconstructed by the player), ditching the redundant field (1 in every 5) to deliver a 48i signal and then de-interlacing this to 24p. Of course it would be defeated if the material was edited in the 60i domain with no consideration to 3:2 cadence consistency.
SOME 24p-sourced DVDs are mastered with only the 24p/48i unique fields recorded and the duplicate fields are flagged in these circumstances - however basing a de-interlacing strategy to 24p based on this would not be a good move - as it would then depend upon how the DVD were mastered (and wouldn't always work - particularly on off-air recordings of film material made on domestic DVD recorders?)