Source files vs transcoded files to begin color correcting

The problem is what to start with in DaVinci Resolve when beginning the color correcting process:

Original source footage- Sony FS7 .MXF format
OR
Transcoded MXF to ProRes 422HQ

The main goal is speed of editing and good (but not perfect) quality. If we lose a lot of data in the blacks or introduce a lot of noise by transcoding first, that would be an issue. If first transcoding to ProRes means we can't pull out more colors or contrast, that would also be an issue.

I can't seem to find any forums or straight answers to this, any thoughts?

The problem is what to start with in DaVinci Resolve when beginning the color correcting process:

Original source footage- Sony FS7 .MXF format
OR
Transcoded MXF to ProRes 422HQ

The main goal is speed of editing and good (but not perfect) quality. If we lose a lot of data in the blacks or introduce a lot of noise by transcoding first, that would be an issue. If first transcoding to ProRes means we can't pull out more colors or contrast, that would also be an issue.

I can't seem to find any forums or straight answers to this, any thoughts?

Hey James

What flavor of XAVC do you shoot?
In case its XAVC-i Then I wouldn't bother transcoding, especially not to PR HQ..
XAVC-i is a good, solid recording format and very NLE friendly.
XAVC-L can be bit intensive on processor (Long GOP) and give you less smoothness than PR when scrubbing the timeline or performing other editing tasks.
But even in that case, you can use Optimized Media and you don't really have to transcode

FS7 is on Netflix's approved camera list, which is a good sign of its worth image quality wise.

The only reason you'd want to transcode is if you have a huge project you want to shrink down.
I believe (last time I checked) Resolve still does not support Trimming of XAVC (unlike Avid/Premiere).

In that case (and until we get XAVC trimming) transcoding is the way to go, I'd pick A 4:4:4 variant of Prores/DNX rather than a 4:2:2...

Hope this helped

Thanks for the quick reply Hector. I'm using XAVC-i for recording. Sometimes the projects get very large because we're shooting in 4K. We were transcoding to reduce overall file size. I've read before that shooting 4K on a Fs7 and crunching it to a 1080p signal gives it better overall color and contrast. We're testing that out to see if it's true on a practical level.
We use FCPx for our NLE and used let it transcode into proxy media. Our idea was to do it before we put it into FCPx because then we could control the overall project size to a much more manageable state for pushing to Davinci or between other editor's computers.
Problem is now all the reference videos on the XML are to ProRes vids and not the original MXF's. Suggestions?