I've got some ideas, but first: what audio format are you using, AC3 or LPCM? Or, do you not even need audio? Do you care about preserving timecode?
I don't care about timecode, but I would like to preserve LPCM audio.
OK, cool. Another question: what is your shooting format--frame size and frame rate?
1080/24p most;y. Occasionally I will overcrank to 30p.
OK, but you're shooting with 24p as your timebase, right? Not shooting at 30p to interpret as 24p in post?
(I ask because it makes a difference )
The framerate for the timeline is 24p. If/when I use footage shot at 30p, I will interpret it at 24p to get a slight slow-motion look. Audio on these clips will likely be ignored.
Got it. Let me work on this a bit; I've been experimenting with some workflows using the same kind of material. At least in my present case, I'm not concerned with preserving audio, and that makes things a little simpler. I'll get back to you...
By the way: I'd suggest shooting at 24p and using the frame rate settings to overcrank in-camera, instead of shooting at 30p and conforming to 24p in post. I use the latter workflow with DVCPROHD (which doesn't overcrank at 1080p24) and with HDSLRs like the Canon 7D. The AF100 will make some deliciously, buttery-smooth in-camera overcrank if you set your recording base as 24p and dial the frame rate/shooting base up to 30p. The downside of this is that you get no audio, whereas shooting 1080p30 will preserve the audio (though you probably won't use it since the video will be slowed down).
But hey--whatever works
I woud suggest that you create a workflow that offlines and even deletes anything ( clips) that are "unuseable".
Trimming clips to what may or may not be useable is already going to be hugely time consuming.
Transcoding is actually also going to be a good option and you should not be concerened about doing so.
I would create DNxHD intermediates if it were me. If on a Mac.. I would consider a workflow using Clip wrap as well.
Beefing up your hardware is also something to think about.
I understand that there are some technical issues when the amount of source material that CS5 has to load gets large.
How large? Personally, I haven't had any issues with this, even using hundreds of clips. That makes me wonder if you're trying to solve a problem that may not actually exist.
I've been through the ringer so many times in the past with technical issues, I've developed a defensive mindset. Better to create a workflow that avoids as many potential problems as possible, than to suffer the consequeces later. My system is fairly beefy, so hopefully this will not be an issue, but I'll do what I can to avoid it anyway. And offlining to remove excess footage is always a good idea, as shooternz points out.
This may be a technical issue that was worked out in CS5. I read several posts on other forums that talked about this, but only one that specifically mentioned CS5, and they did not specify what format their source material was in. And it could be that they had hundreds of gigabytes of unused data in their timeline. Other issues that concern me (from other forums) are sluggishness when combining multiple file formats in the timeline (particularly with AVCHD and h.264, the two native formats I'll be shooting with), hence my resistance to transcoding.
The best suggestions I've had so far (on another board) are to use CS5 to export the trimmed clips into h.264 format, or to use a program like iSkysoft or Clipwrap. I've tried Clipwrap, and the resulting files - non-trancoded - don't play well at all in CS5 (horrible playback). Transcoded files playback fine, so that might be the eventual workaround. I'm going to try iSkysoft and see if it works any better.
Colin - I'm just curious, why do you think it is better to shoot with VFR than to interpret 30p footage (aside from having more control over frame rate)?