Interpreting the effects of 'separatefields'

[pooksahib]

Hi. Most of the DVDs I've worked on have followed the same route - watch it frame-by-frame, no miceteeth=fine, miceteeth=run separatefields(). Motion every frame=use QTGMC(), motion every second frame=use IVTC/TFM.

Today's DVD is an odd one in that it has miceteeth but separatefields, when I click forward a frame at a time, is giving me an image which I can best describe as 'two steps forward, one step back'. Also, every second frame is showing a little bit of blackbar along the top.

I know some of you like to have a clip but, from the above description, does anyone recognise what the issue is and what deinterlacing method should be used? Thank you.

[jagabo]

Today's DVD is an odd one in that it has miceteeth but separatefields, when I click forward a frame at a time, is giving me an image which I can best describe as 'two steps forward, one step back'.

Wrong field order. Use AssumeTFF() or AssumeBFF() before SeparateFields().

Also, every second frame is showing a little bit of blackbar along the top.

One field has a black line at the top. That's not unusual.

I know some of you like to have a clip but, from the above description, does anyone recognise what the issue is and what deinterlacing method should be used?

Probably QTGMC() or Yadif(mode=1).

[pooksahib]

AssumeTFF() does the trick. Thank you once again, jagabo.

[jagabo]

Note that the default, if the source filter doesn't indicate the field order, is BFF.

[pooksahib]

Duly noted...