Re: How to UNDO incorrect Advanced Pulldown Removal

[Bill Russell] "The removal of coursed messed up the resulting 23.976 clip and it is now unusable with random mixed-field frames"

Working through the math here...

What they started with was normal 2:3 field cadence according to what I am able to determine from the above. Typically that would be a SMPTE ABCD cadence. If there was an Advanced Pulldown removal applied (how did they do this?) which removes the third frame of every five-frame 29.97 or 59.94 group, what should result is 4 progressive frames. AP is 2:3:3:2, or SMPTE ABDA, so what gets removed is the third field of the B frame and the first field of the D frame, which are redundant and what you theoretically yield is an AAAA sequence. There is no C-frame in Advanced Pulldown, by definition.

What has been removed from the ABCD sequence though would have been the third field of the B frame and the first field /half frame of the C frame, leaving a mixed CD frame as the third progressive frame. This shouldn't be random at all unless there is something else going on. As you step through the 4-frame sequence you should notice that it is always the third frame that is mixed and will interlace/dither.

I fixed something like this once with one of the Nattress smart re-interlace filters which pulled out the C-portion on the upper dominant, and removed the D-portion, which really calmed down the motion artifacts.

If this pulldown removal was performed as a batch process on an edited/baked sequence, there is really no hope of regaining a coherent segmented field sequence.

I can see the big question marks over the heads of everybody at Apple... Huh??? What's a "field"????

jPo

"I always pass on free advice -- its never of any use to me" Oscar Wilde.