I shoot with a variety of cameras, from Sony XDCAM to Canon DSLRs and GoPros. Whichever footage I start with, FCP automatically sets the timeline so that it won't require rendering. But when I add video from another codec, I have to render each clip.
I know the general answer: transcode all footage to a common codec. But I often operate under extremely tight deadlines, so lengthy and laborious rendering sessions would be problematic.
What would be the most efficient workflow which would allow footage from all of these sources to co-exist on a FCP 7 timeline, without rendering? Currently, I'm just accepting the performance penalty of rendering on the timeline, but that is getting old.
Canon DSLRs and GoPros record in a version of H264, so you'd have to convert them to something else anyway if you want to avoid the lengthy and laborious rendering that you're putting up with now. FCP7 works acceptably with XDCAM in my experience (EX1 and EX3). So I reckon the choice is either to transcode the DSLR & GoPro footage to XDCAM and edit in XDCAM, or convert them to ProRes and use the Sony plug in to log and transfer only the sections of XDCAM that you want to use into ProRes (which would save time over transcoding everything).
TBH, for a big job I'd be transcoding everything to ProRes, but for fast turnaround I'd say do a couple of tests and if your system handles XDCAM ok (XDCAM is a GOP format so it does have processing overheads but it's not as severe as MPEG types like H264) then go with that.