CST best practice with BRAW+MOV+MP4 in same project

I work a documentary with multiple format file BRAW+MOV+MP4. BRAW is from Ursa Mini Pro G2, mov are from DJI Phantom 4 v2.0 and graphics animation, and mp4 from outher sources. My project Color management are Davinci YRGB-Davinci WDIntermediate-Rec 709 Gammma 2.4-Trilinear and Camera Raw settinngs BRAW-Full res-Camera metadata. My workflow start with CST IN Blackmagic Gen 4/5-Blackmagic 4.6K-Davinci WideGammut-Davinci Intemediate them adjustable nods from Primaries Colory Wheels or Curves. For BRAW from Ursa Mini Pro G2 is the Blackmagic 4.6K the corect gamma settings?
After the adjustable nods must to do two conversion nods CST OUT Davinci WideGammut-Davinci Intemediate-Rec 709- Gamma 2.4 and then another CST CONVERSION Rec 709- Gamma 2.4-Rec 709-Cineon Film Log before Kodak 2383 Film Look LUT? Can do it directly a single CST OUT nod Davinci WideGammut-Davinci Intemediate-Rec 709 - Cineon Film Log?

I like simple, so I would go a different way.

Rec 709 RCM.png (23.29 KiB) Viewed 416 times

I like simple, so I would go a different way.

Any way could be good. For the easy projects i use another way. But for this documentary i want the best results, i saw that many colorist use CST way.

which line?

work way2.png (850.47 KiB) Viewed 390 times

for this documentary i want the best results

The same settings will produce the same results.

One is easy, one is complicated.

Can do it directly a single CST OUT nod Davinci WideGammut-Davinci Intemediate-Rec 709 - Cineon Film Log?

Your WAY 2 example is a more optimized method because it would yield the same result with less nodes. However, a 'trick' you could do with WAY 1 is to put CST 2 OUT + LUT in a compound node so you can adjust the opacity of that node to blend how much "FPE" you want if you'd be interested in that.

Another optimization step could be to set your BRAW decode settings in your project to DWG/I. That way you do not need a camera to working space conversion step for any BRAW media you bring in.

You also mention you've set your LUT interpolation to Trilinear. This should be Tetrahedral for much better accuracy.

Your WAY 2 example is a more optimized method because it would yield the same result with less nodes. However, a 'trick' you could do with WAY 1 is to put CST 2 OUT + LUT in a compound node so you can adjust the opacity of that node to blend how much "FPE" you want if you'd be interested in that.

Another optimization step could be to set your BRAW decode settings in your project to DWG/I. That way you do not need a camera to working space conversion step for any BRAW media you bring in.

You also mention you've set your LUT interpolation to Trilinear. This should be Tetrahedral for much better accuracy.

I use compound node usualy but for a easy way representation i decomposed, If i do the Way with a directly single CST OUT nod Davinci WideGammut-Davinci Intemediate-Rec 709 - Cineon Film Log the drone DJI mov give me some artefacts (clipping in the sky for example) If i do the Way with two nods Davinci WideGammut-Davinci Intemediate-Rec 709 - Gamma 2.4 + Rec 709 - Gamma 2.4- Rec 709-Cineon Film Log i dont see artifacts, Why?

If i do the Way with a directly single CST OUT nod Davinci WideGammut-Davinci Intemediate-Rec 709 - Cineon Film Log the drone DJI mov give me some artefacts (clipping in the sky for example) If i do the Way with two nods Davinci WideGammut-Davinci Intemediate-Rec 709 - Gamma 2.4 + Rec 709 - Gamma 2.4- Rec 709-Cineon Film Log i dont see artifacts, Why?

It could be that a transform to Rec.709/2.4 clamps negative values that may cause issues bypassed and converting to Rec.709/Cineon directly. Another thing could be a difference in mapping of the DWG/I luminance. You could try changing the max input nits on the CST that converts DWG/I to 709/Cineon.

Further more, the built in Resolve Film look is originally intended as sort of a print reference for real film scans rather than a solid LUT for creative grading of wide gamut digital camera captures. A cineon scan is not really the same as Rec.709 so converting digitally captured data to those primaries could yield values that a cineon scan would never have which causes problems for the LUT.

If you want more robust solutions it can be beneficial to look for film print emulation LUTs or power grades that are designed for log to log for scene referred workflows for digital cameras.