I am a big fan of this forum. I spend a lot of time reading it but this will be my first post. I seem to have encountered a problem that I just cannot figure out. I am shooting a series of one minute pieces on boutique hotels in NYC. Due to the nature of this project, I use a lot of pans. The pans are the issue here, but let me back up and briefly describe my workflow. I am shooting on a Canon T2i at 30fps. I am digitizing through iPhoto and then transcoding through mpeg streamclip. My settings for the initial transcoding are as follow: Apple ProRes 422, quality 100%, 1920 x 1080 (unscaled), interlaced scaling is unchecked, and deinterlace video is left unchecked. I then import footage into FCP. I am using version 6.0.6. Once in FCP, my pans seem relatively smooth, meaning as smooth as I shot them. Then, once I export my sequence from FCP as a quicktime movie, and watch the piece again, I see that my pans are suddenly choppy or jumpy. The beginning of a pan may be smooth, but then as it unfolds, suddenly it will begin to get choppy, as if it is dropping frames. The bizarre part about this is that not all of the pans do this. And not all of them do it in similar places along the pan. I believe, but not completely sure, that that the choppiness seems to manifest less in slower pans. When I transcode the footage (again though mpeg streamclip) back into H264, I use the same settings as before but usually limit the data by 10000 kbs, or around that. The choppiness is still there, but seems to manifest in yet different places within the pans. Does anybody have any ideas as to why this might be happening, and furthermore how to get around it?
One thing to check (and I am not even sure this is what is going on for you in particular) but make sure that the Quicktime player has fully buffered your video and then play it again...this can make a large difference in the playback quality - sort of like watching a streaming or progressive download version of your video coming off some Web site...that can be choppy too if the video is not fully or largely buffered
Just an idea but something tells me you have a different situation going on.