the unnecessary deinterlace bug

It's exasperating when you drop a clip into an empty FCP7 sequence that you know fits it, maybe the clip was exported from that very sequence, and yet there's the pop-up: "This clip does not match this sequence's settings..." inviting you to let FCP7 change the settings appropriately. You say "yes" and the 100% canvas shows a nice image as if FCP7 knew something you didn't. You then examine the sequence settings and find, for example, that they're set to interlaced when you were sure the clip was progressive. The clip was made progressive. QT Edit said the clip was progressive. Etc. So you change the sequence settings to progressive. Then there are two possibilities. For most images you don't notice a change on the canvas. All is well. But some images go to hell on the canvas. Download example here. I investigated further.

For some codecs, including these

Animation codec
H.264
None
Uncompressed 8-bit 4:2:2
Uncompressed 10-bit 4:2:2

when the frame rate is 25, 29.97 or 30, FCP7 wants them to be interlaced, period. If you set the sequence to progressive, FCP7 does a crude, line-doubling deinterlace on the 100% canvas. Other frame rates have no such problem.

Other codecs, including these

AIC
ProRes family

are immune to the bug.

Mercifully, the crude unnecessary deinterlace is limited to the canvas. On export, the whole progressive image is there (so far as I've checked).

How then to set the sequence in the affected cases? You have two bad choices: leave the setting at interlace according to FCP7's whim; or change the setting to progressive and sacrifice critical viewing in the canvas. I fear the first choice could cause some filters to operate incorrectly. How are other FCP7 users coping with the bug?

Dennis Couzin
Berlin, Germany

Use a proper reference monitoring system, not the computer monitor, set the sequence to progressive. QC on the reference monitor. Told you you'd run into a lot of hurdles trying to get that FCP canvas to work as a reference monitor.



www.strypesinpost.com

Touché. Maybe.
Are you certain that "a proper reference monitoring system" will be immune to FCP7's bug? The reference monitoring system is not monitoring a file, but rather the contents of the FCP7 timeline. If for certain codecs FCP7 sends poop from the timeline to the 100% canvas (but not to the export), then how do we know what FCP7 sends to the reference monitoring system?

When an editing system has bugs, the very idea of "a proper reference monitoring system" of the timeline is questionable. For example, suppose this FCP7 bug were a little different so it sent the poop not only to the 100% canvas but also to the export. Then what should "a proper reference monitoring system" show as being in that timeline?

Dennis Couzin
Berlin, Germany

With an IO costing under $200 these days, it's a no brainer vs trying lots of workarounds.



www.strypesinpost.com