Re:Vision Synch Problem

[Peter Litwinowicz]

All that FieldsKit Deinterlacer does is convert each frame from 2 fields to 1 frame (in the default setting, as you say). So are you saying that the video is out of sync, or the audio? And you say that some shots that were 3 seconds become 1 second… that’s not something FieldsKit is equipped to do (but I believe you are seing it!).

If you turn off the FieldsKit filter, do you see the same behaviour (that is, I’m wondering if there is some problem going from Avid–>After Effects)?

You can always send us a project and source footage that shows the problem to techsupport@revisionfx.com

Pete Litwinowicz
https://www.revisionfx.com

[Alan Howard]

Thanks Pete,
I know it seems odd and I guess that in the tired state I was that night I may have rendered the wrong sequence with a stray edit in (I really doubt it tho!)… Later, when I put the rendered QT over the original Video layer in Avid to check if pics were drifting (Re:Visioned footage on top video line and original layers below), there are one or two points where the edit points didn’t synchronize, but at the next cut point they did… crazy I know!.. I’ll redo and hopefully it’ll go away …. Pete, while I have you can I ask is it usual to try to output a 60 min piece using the same settings ? I get the impression that the settings should be altered for various situations but my 60 min pieces contain fast moving, slow moving, panning, tripod, hand held shots etc (as do most 60 min pieces I’d suspect). Hence my need for a “Cover all” default setting that will look good for all of the above. Whats the normal work flow? I assume that you cut your piece, export as QT Ref file into AE and off you go.. is that the best route? I generally shoot on 16:9 PAL DVcam(why are tutorials always in NTSC?)and the eventual destination is DVD via Encore…
Thanks
Alan

[Peter Litwinowicz]

[AlanEHoward] “I get the impression that the settings should be altered for various situations but my 60 min pieces contain fast moving, slow moving, panning, tripod, hand held shots etc (as do most 60 min pieces I’d suspect). Hence my need for a “Cover all” default setting that will look good for all of the above.

Unfortunately, I think that having a “cover all” setting is generally not going to give you the best results for all footage (that’s why we have the tweakable settings) However, I understand the need for that in your case. If you need something that will look good in most cases you might want to set the interpolation type to “Best 3 Neighbors” and set the Smooth Vertically to 1.0 and Sharpen set to 0.5.

You might consider animating FieldsKit’s settings so that you have a setting per edited segment rather than once for the entire piece. In particular, you might be able to get much better results if one off the segments has a locked-off camera (using the motion detection settings of FieldsKit). Unfortunately the motion masking settings tend to be content dependent, so a “cover all” setting for that is not really appropriate… for a “cover all” setting I would advise you to leave the Detect Motion Mask setting of FieldsKit to “No Motion Mask”.

[AlanEHoward]Whats the normal work flow? I assume that you cut your piece, export as QT Ref file into AE and off you go.. is that the best route?

Yeah, I think this is a pretty good route.

[AlanEHoward] I generally shoot on 16:9 PAL DVcam(why are tutorials always in NTSC?)and the eventual destination is DVD via Encore… “

Why are the tutorials always in NTSC? Because that’s what we work in. The difference between NTSC and PAL for deinterlacing is minimal, no? In both you still have two different fields per each frame, on alternating lines of the frame. The basic concepts of using FieldsKit will be the same in NTSC or PAL.

In any event, if you have any more questions, please don’t hesitate to ask.