PAL Digibeta masters just arrived, a technical query

[Pal 25fps DigiBeta, Upper field first, ProRes 422]

We've had the Digibeta tapes transferred to ProRes and I'm noticing a sort of single frame print through of the next scene at every scene change. Is this a technical glitch of some sort? Or are the PAL masters poor standards conversions? Is there a way to eradicate this?

1st question: is this a field issue?

open the clip into the viewer, and set it to 100% (you can press and hold the H key to do this)
is there combing of the image on the scene changes, or elsewhere?
then there is a field order issue.

another Q: is this a film telecine done at 24fps?
that'll give you shared fields


i believe i have seen something like you describe that hasn't been a field issue.
i found it curious, wondered if it was some sort of magnetic transference thing (I'm kind of making that term up, by the way)
but it wasn't a show-stoper for us, so we never followed it though.


nick

As Nick explained, looking at any frame that includes movement on the viewer at 100% will reveal interlace. You can open the video in Avid and view it field-by-field to determine the field dominance (upper first vs. lower first). I posted an FCP7 method for seeing fields that reveals the dominance.

Three things can be wrong with someone's "upper field first ProRes".
1. The top line actually shows a later moment than the second line. The ProRes is actually lower field first when you, FCP7, QT think it's upper field first. I'd then edit the QT metadata using QT Edit. (And check it again for the final output video, if interlaced.) To use FCP7's field change filter makes one dark line.
2. The transfer from tape has dropped the first field. Then, if the tape included cuts between scenes, there is one double-exposure-like frame at each cut. There is a tedious method for discarding the first field from your first frame and then shifting fields from one frame to the next, but it's simpler to just cut out the offensive frames at the cuts.
3. The top line shows what should be in the second line, and vice versa. Third line and fourth. Etc. You can judge this by looking at a slanted edge that isn't moving on the viewer at 100%. It will have juts. There is a tedious method for fixing this but better get a new transfer.

Only #2 gives the double-exposure-like frame at each cut, but when #2 occurs, #1 and #3 can occur as well.

Dennis Couzin
Berlin, Germany

It's a telecine of a 35mm film trailer, but I have no way of knowing if it was transferred at 25fps or 24fps, and then converted to PAL. Luckily the accompanying feature film itself does not have this problem.

I may have to just cut these frames out, as you suggested.

if it was telecined as PAL, it would be 25fps, and running a little bit faster.
possibly the audio would be a pitched up a touch.
and most likely every frame would be whole.

if it was telecined as NTSC, then you would get a putdown, where occasional video frames share adjacent film frames.
this can be removed


but did you do the simple check for combing?
that would help a lot.


nick