-
DV to Flash FCP workflow
So I’m working with a solid print company that has me freelancing on a number of videos. A year ago any project was being optioned out to the studios, mainly for DVD. But web video has finally arrived. With web video came a refurbished desktop mac, canon gl2, and a sennheisser mic set. Being brought up in the news arena, everything was DV NTSC. Shoot, edit, and export DV NTSC. The web people did all the flash processing. All was well. Then internal changes led me to do all the exporting myself. In looking at my first few flv files i found the interlacing lines to be ugly and the video nearly unwatchable. The easy fix was to have Compressor and Flash do the deinterlacing. But there was a very noticeable loss of quality.
With lots of experiments and research it dawned on me that I should do the deinterlacing through FCP and work in a progressive timeline. This was reinforced by the Cow’s DVD on Video compression. I created a DV NTSC capture preset with proressive scan mode. Within the sequence settings you can select “progressive” scan mode. I also thought it would be weird to have any field dominance and selected “none”. This seemed to work for awhile. And I had no clue what pixel aspect really did so I just left it with CCIR 601.
Then as we expanded resources we added to the mix a Panasonic HVX-200 which shoots 480p, started adding Animation QTs from After Effects, layered Photoshop and Illustraor files, Livetype and Motion project files to our FCP sequences. The video played back progressive but the the Animation QTs, livetype files, and motion projects started to be very chunky around the edges. No sort of de-interlacing helped. I messed around with project settings in all the other applications for better exports. Nothing seemed to work. The full-time guy found that selecting lower field dominance made all the graphics and After Effects videos look perfect but I found the all video started having interlacing. It was more prevalent with footage from the GL2 but all the video had interlaced features. So I had nice progressive graphics but interlaced footage in a progressive but lower field dominated sequence. I’m going crazy comprehending how this even works.
It’s now obvious that even with a progressive scan mode capture preset, FCP will sense what type of footage is coming in and override such settings. And let’s be honest why am using DV NTSC when I don’t have to. And from what I know, internet video uses square pixels.
So with all this said, I’m still trying to answer these questions:
WIth combining footage from a GL2 (interlaced) and a HVX-200 (progressive) (and both shooting on mini-DV tapes), what would be a good capture setting? Would it be smart to import with something like ProRes422 even though its a higher sampling rate?In setting up a sequence, what pixel aspect and compressor selection make sense for Flash/Web (assuming answer from #1)?
How should the deinterlacing of GL2 footage be dealt with, simply render or filter selection or IDK?
Based on sequence settings, what type of project settings or video exports should i consider (if at all) for After Effects, Photoshop, Illustrator, Livetype, and Motion?
-
1 Reply
-
Your post is pretty long and meandering…
quick reax
Shoot progressive
Work 24P
Deinterlace using higher quality methods (Compressor is fine for this)
Richard M. Harrington, PMP
Author: Photoshop for Video, Understanding Adobe Photoshop, and ATS:iWork
Reply to this Discussion! Login or Sign Up