Resolve to FCPX and then Back to Resolve - RAW Workflow

Hi guys,

I'm shooting a music video this saturday. We're shooting with two Blackmagic cameras and we're trying to decide whether to shoot RAW or not.

Our ideal workflow would be as follows :

Shoot RAW -> Open in Resolve -> Transcode everything to Prores 4444 -> Edit the ProRes 4444 footage in FCPX -> export XML to Resolve and do the color grading in Resolve.

My question is : How do I re-connect the edited ProRes 4444 footage via the XML to my RAW footage, so that I will actually be grading the Raw footage and not the ProRes 4444 footage.

Hope this all makes sense, thank you in advance for all the help.

Our ideal workflow would be as follows :

Shoot RAW -> Open in Resolve -> Transcode everything to Prores 4444 -> Edit the ProRes 4444 footage in FCPX -> export XML to Resolve and do the color grading in Resolve.

My question is : How do I re-connect the edited ProRes 4444 footage via the XML to my RAW footage, so that I will actually be grading the Raw footage and not the ProRes 4444 footage.

I believe in Resolve you import the XML from FCPX and in the Conform page point to the original RAW footage. You could even just export ProRes 422 HQ or even ProRes Proxy from Resolve if you want to save some effort and resources in FCPX if that is only being used to make your edits.

Rick Lang
Sent from my iPad using Tapatalk HD

These vids should help:

[list=]http://www.youtube.com/watch?v=Oxv-RWQmI1M

[/list]

thank you very much !

Alexis Van Hurkman has a really great tutorial on DaVinci Resolve available at rippletraining.com. It covers Resolve 9, and I've been slowly working my way through it (11 hours of training broken up into 64 topical lessons). One of the lessons is on RAW workflow.

I would highly recommend it if you are like me and are just getting into Resolve.

Thanks,

I have a question that pertains to this specific workflow and the previous question.

If I initially transcode RAW to 4444 in Resolve, then edit in FCPX, then grade the 4444 in Resolve, what are the consequences of doing so?

Are 4444 files larger respectively? Would I lose any quality or headroom grading 4444 as opposed to the original RAW files? Ultimately if I were to keep only one version of the files as my master both for editing and storage, would I want 4444 or RAW?

Thanks,

Chris

I've tested this with RED footage, if you do a decent first-light in Resolve before transcoding, you visually don't loose anything when staying in ProRes 4444 after that.

But the files will be bigger and you need some speedy disk to play them.

If you have a high shooting ratio, better opt for ProRes 422 offline editing, re-conform in Resolve and finishing there from RAW.

I have a question that pertains to this specific workflow and the previous question.

If I initially transcode RAW to 4444 in Resolve, then edit in FCPX, then grade the 4444 in Resolve, what are the consequences of doing so?

Are 4444 files larger respectively? Would I lose any quality or headroom grading 4444 as opposed to the original RAW files? Ultimately if I were to keep only one version of the files as my master both for editing and storage, would I want 4444 or RAW?

Thanks,

Chris

If one was starting from, for example, a three CCD sensor group with each CCD picking up values for red, green, blue, then I can understand than transcoding to a 4:4:4:4 codec is desired to retain all the colour information captured by the sensors. But for a bayer sensor that has two green photosites and one red and one blue, it would seem to me that using a 4:2:2 codec is the best you can do to match the original colour information captured. I am not an expert on this so I may be wrong, but it seems to me the larger 4:4:4:4 files won't give you an appreciable gain in colour because you can't invent accurate colour values for the 'missing' red and blue values. The debayer can create anything but it may not give better accurate colour. Perhaps there is a benefit if you are doing keying and effects. Can anyone set me straight on this?

Rick Lang
Sent from my iPad using Tapatalk HD

If one was starting from, for example, a three CCD sensor group with each CCD picking up values for red, green, blue, then I can understand than transcoding to a 4:4:4:4 codec is desired to retain all the colour information captured by the sensors. But for a bayer sensor that has two green photosites and one red and one blue, it would seem to me that using a 4:2:2 codec is the best you can do to match the original colour information captured.

Visually, there is no difference between 4444 and 422.

BUT!

I know that when using BMCC camera in post, and pushing the color and contrast of footage around, 4444 supports 12 bit color (which is equivalent to BMCC's RAW bit depth) and 422 only supports 10 bit color. This means 4444 has significantly more headroom and can maintain a higher quality when "adjusted" from it's natural state. You will see a difference between the two when grading them.

The question I have, is will you see any difference/loss in quality in grading potential between 4444 and using the native RAW files?