DSLR workflow for ideal render and export

Question

I recently completed my first 5 minute project in Premiere Pro CS6 with great results.  The source material is a Canon 7D, and I'm working with unconverted native files.  Unfortunately, as the project had color-correction with Magic Bullet looks, filters like sharpening, and layers to "blur" certain elements on-screen, the initial compressed h264 export took nearly 7 hours.
A bit of research led me to realize I was better off exporting a "master" and converting outside of Premiere Pro.  So, what would be the best way to proceed?
Should I render my timeline and then choose export with "match sequence settings" and "use previews"?  Should I not pre-render and instead export only with "match sequence settings"?  If I do pre-render and use previews, will the resulting quality be the same as if I render while exporting?
Or is there another way?  Could I create preview files with ProRes and export to ProRes?
Thanks for helping out a thrilled PP convert.

Answer 1

A little primer on Preview files...(btw I think they are great when used as they are intended)
Preview files are caches of raw frames [with effects applied to them] that are encoded into the Native Editing format (eg DV, HDC, MPEG-Intra).
It's a little known fact that ANY Export format can be made into an editing format. When you make a new Sequence you just have to pop over to the Custom Tab to change to that video format. In general this isn't a good idea though. Why? Because most Export formats are not good Digital Intermediates (DI). By that I mean that repeatedly compressing and decompressing damages/degrades the footage or they are slow to decode and encode.
So, what format to choose? Formats like Avid DNxHD and Apple ProRes are great editing modes. They don't degrade over multiple generations very much and are relatively low on CPU power to decode and encode. There are lossless DI's out there too like UtLossless (which is free). At the other extreme are codecs that don't fair well to being used as a DI - like H.264/AVC. They are simply too CPU hungry for the task.
Don't misunderstand me - editing source AVC footage is fine because decode is waaaay easier than encode. It's just not practical because of the encode time.
So why use Preview files at all?
IF your effects (like Magic Bullet or Motion Stabilizer) takes a LONG time you only have to do the computation once. Work in the highest resolution you'll be using (eg 1920x1080p). After you have the preview files "rendered" Premiere Pro will use then instead in that section and you'll have nice smoooooooth editing.
You then are free to Export your final video to whatever format you want in whatever size you want. And because you chose a DI that isn't hard on the CPU to decode you pick up a TON of time vs having Magic Bullet recalculate the same thing over and over again for each output format/size/datarate. - Just make sure you check the "Use Previews" button otherwise they will be ignored.
Level 200 tips:
IF, and this is a big IF, you happen to be exporting your final video at the same
    FRAME SIZE
    FRAME RATE
    COMPRESSION SETTINGS THAT MATCH THE EDITING MODE
Then, you can get an extra Encode time boost for Exporters that support "Smart Rendering". Smart Rendering is when all the above conditions are true and the compression scheme supports being able to COPY the preview frame instead of de-compressing it and then re-compressing it.
Basically there is little to no chance that's gonna happen unless your final output is a Digital Intermediate format to then go onto some other post-production step.
So... choose a good Editing Mode that doesn't lose quality when encoded (during the preview render) then decoded (during the early phase of Exporting each frame) to make the raw uncompressed frame that is Encoded into your final output format (eg F4V, AVC for Bluray, YouTube, Vimeo etc).
'hope that helps.
Rallymax.