Help deinterlacing Ren & Stimpy DVDs

[CompuKid101]

Here is a sample with constant panning (good for seeing whether a frame is progressive or interlaced). I got the below sequence. It looks to be PPPIIPI repeating but then near the end it gets crazy.

IIPPPIIPIPPPIIPPPIIPIPPPIIPPPIIPIPPPIIPPPIIPPPPPIIPPPIIPPPPI

heres the sample (37.6 MB): http://www.jevinskie.com/ren_and_stimpy.m2v

Any help? Thanks!

[CompuKid101]

Help, please? I'm pretty lost with this one. I've tried quite a few deinterlacing methods and none have given very good results.

[cc979]

using megui, it reports hybrid film/interlaced but works best with source: interlaced - so maybe something wrong interlace flags

[Guest]

Where did that come from? You can recover progressive frames with Telecide(), but it's jerky on the pan. Knowing where it came from or how it was made may shed some light on that.

[ChiDragon]

Unforunately the source contains "orphaned fields" (film frames represented by only one video field), which occur in both top and bottom fields. If they only happened in either top or bottom you could use TFM(PP=7) but since they occur in both you'll have to use something like TDeint(mode=1,tryweave=true,MI=20) unless you can find a pattern. Then figure out cycleR and cycle values for TDecimate...

EDIT: I found a pattern that works for the whole panning section but it breaks during the conversation after the pan...
EDIT 2: Should've mentioned, the base frame rate seems to be around 24.9 fps. Maybe the PAL DVDs would be better.

[R3Z]

Yeah this is a strange one, when using a film deinterlacer such as Decomb IVTC it looks pretty good until about the 4th or 5th frame after the first P frame at which point it looks semi interlaced.

This is beyond me though its interesting to read why it happens (thanks ChiDragon).

[CompuKid101]

Where did that come from? You can recover progressive frames with Telecide(), but it's jerky on the pan. Knowing where it came from or how it was made may shed some light on that.

http://en.wikipedia.org/wiki/Prehistoric_Stimpy/Farm_Hands
Seasion 4, Episode 4a, from the Ren and Stimpy DVD collection. The DVDs show fields every once and a while even while viewing on a TV.

[CompuKid101]

Unforunately the source contains "orphaned fields" (film frames represented by only one video field), which occur in both top and bottom fields. If they only happened in either top or bottom you could use TFM(PP=7) but since they occur in both you'll have to use something like TDeint(mode=1,tryweave=true,MI=20) unless you can find a pattern. Then figure out cycleR and cycle values for TDecimate...

EDIT: I found a pattern that works for the whole panning section but it breaks during the conversation after the pan...
EDIT 2: Should've mentioned, the base frame rate seems to be around 24.9 fps. Maybe the PAL DVDs would be better.

Very odd... :sly:
I'll tinker with it some more tonight. Anyone ever seen anything like this before?

[Mug Funky]

these aren't natively progressive... they look to have been edited from video tape rather than film (at least in parts).

and the PAL discs are converted from the NTSC originals, so that's no help.

this series wasn't created with backup in mind :)