-
scaling, interlace & motion filtering LONG POST
High all,
I wanted to start a discussion about final cuts scaling, handling of interlaced video, de-interlacing and motion filtering quality.
Apple says that the “motion filtering quality”/ “best” setting, in the “sequence settings” window, will produce the highest quality rendered result. Some people have the opposite experience when converting 1080 HDV to SD for DVD. https://www3.telus.net/bonsai/Step-by-Step.html So which setting actually yields the best looking results, and if its fastest (linear), why would Apple mislead people by naming them this way?
My negative experience was a year and a half ago with 1080i footage from a Sony FX1 for a film school project. The project had to be handed in on DVD or DV tape. We exported a self contained QT Movie from the HDV 1080i timeline and compressed it with compressors DVD Best Quality setting. The results were terrifyingly ugly with interlace artifacts and resolution that would make you think the footage was shot on a cell phone camera. We experimented with exporting from a SD timeline, De-interlacing in HD & SD, upper and lower field dominance in compressor, all 3 motion filtering settings in FCP and changing the sequence to almost every codec in FCP 5’s arsenal before export. Everything yielded poor results.
The only conclusion I could come to, was that FCP just plain handled the footage from that camera wrong. Our instructor encoded some 24f HDV footage from a Canon XL H1 using our 1st attempted method and it looked 100% better. We brought him our 1080i HDV QT Movie(which looked great in HD) on a data DVD and he compressed it to DVD on his machine. It looked the same as ours, hideous. Since then we have not shot any HDV with the FX1 for SD delivery and I now own a Canon XH A1 but have not tested any 1080 60i HDV from it. Is FCP retarded when it comes to down-scaling or de-interlacing interlaced footage?
I recently found the site from which I posted the link above. I have not tested the “Bonsai Method”. It seems like a lot of hoops to jump though, when in my opinion, FCP should do what it says it does, and if it cant do it well, it shouldn’t say it can do it at all. I know no software is perfect. They are all a work in progress but we shouldn’t have to learn everything by trial and ERROR.
I recently had the same effect from scaling down some DV footage for a motion menu in DVDSP.
I would like to hear if anyone else has experienced FCPs short-commings in this area. Do Adobe and Avid handle these simple but important tasks better than FCP?
I welcome any and all comments. If I am just stupid please teach me why.
Thanks to all who reply.
Shiloh
P.S. Sorry about the length of this post.