[Brandon Kraemer]

Did you do a clean install of your system or just the upgrade install?

thanks

[Brandon Kraemer]

Thanks Michael for the link. I think it will help.

bk

[Brandon Kraemer]

This isn’t a debate about hardware, I work with what I have.

[Brandon Kraemer]

Were not going back to MC, finishing in FCP. Need some kind of auto conform workflow to re-capture these clips in FCP. I don’t think it’s possible but there should be a way to convert the EDL to remove pulldown.

[Brandon Kraemer]

I’ll respond to my own post because the nice guys at AJA tech support already fielded this one for me. The answer is yes it can… BUT… you have to begin your capture on an “A” frame. Tedious if you have lots of clips.

bk

[Brandon Kraemer]

We are seeing interlaced pull-down frames that are causing strange artifacts in our workflow. According to the FCP manual, anytime your working with 3:2 telecine pulldown, your supposed to reverse telecine these clips. However, this is the first time I’ve seen this issue myself. I have successfully been able to reverse the telecine, converting the clips to 23.98, but now I have a sequence that needs to be manually re-conformed.

I know in the past I have done 29/97 telecine footage with no issue in FCP.

Any thoughts?

bk

[Brandon Kraemer]

Let me clarify my post.

In FCP we can frame advance through the footage, but we never see two repeated frames like we would expect to see with 3:2 pulldown footage. What we do see is field tearing or mismatched fields on two sequential frames that have unique images to them, but we do not see the zipper effect of mismatched fields on the other frames. De-interlacing removes this unwanted effect on these frames, but still does not reveal two repeated frames the way we would expect 3:2 pulldown footage to have.

Also, if we de-interlace all the footage on this same roll, on the shots that do not seem to have the zipper effect, de-interlacing distorts the images, like half a zipper, so to speak.

Sending these clips to shake doesn’t change this, but if we interperate the field order of the clips in shake the bad interlacing disappears. Sending renders back to FCP look normal. If sent on from FCP to Color, interlaced frames re-appear!!! Color renders back to FCP pick up a ghost image where motion is concerned on these two frames, but there is no zipper effect. Color is apparently frame blending these fields into the frames in a different cadence. Color’s renders are not really the problem, just symptomatic of the over all issue which starts with our captured clips.

We digitize this footage to FCP via a Kona 3 and a Phillips Router from a Sony Digital Betacam deck. All footage is 29.97 fps, lower field dominant. Is this a bad cadence issue, a timing issue in our video path? Anything we file export of this keeps this bad interlacing look on playback. Playback of this on the FCP timeline looks normal, but stepping through frames shows the interlace issue.

[Brandon Kraemer]

That is helpful, we often do export DPX sequences (Glue Tools QT Conversion) and shots that require alpha so it became our standard export method. While that proved to be the culprit, what if anything can be done to prevent color shift when exporting Animation with Alpha?

Thanks for your help.

bk

[Brandon Kraemer]

I have tried it a few ways…

QT Conversion using…

Native codec (DVCPRO HD)
Uncompressed 10-bit (our usual standard)
Animation

Each way yields color shift. Animation is by far the most noticeable. These color shifts are evident when I reimport the exported clips back into FCP and compare with my source footage.

[Brandon Kraemer]

Jeremy,

AE CS3 8.0.2.27. No working space assigned with in AE. I quit both FCP and AE before removing the codec. The real issue is even things I export out of FCP and re-import with out going to AE show the same color shift.

FCP timeline is set to DVCPRO HD Codec. We use a lot of P2 HD stuff, Varicam too. When I can we work Uncompressed HD 10-bit, not so with this footage.