Nikon Z-Log L or S-Off - CST Input Color Space & Gamma?

[MarkAm]

Disclaimer: Complete Noob here learning how to grade my own videos.

I'm starting off with clips from a D5300 shot with the Nikon Z-log - L color profile that I want to grade in Davinci Resolve.  I know these clips are in the sRGB color space so if I am adding a node with color space transforms is my best approach to select sRGB for input color space and gamma?  I was thinking of outputting to  Davinci Wide Gamut & Intermediate and adding a few more nodes to grade before adding another CST node to go from DWG/I to Rec 709.  Am I better off with some other input color space and gamma as a starting point?

If I were to shoot with the Cinema Z-Log S-Off, same question - would I select srGB for the input color space and gamma or would I be better off with some other input color space and gamma?

[kye]

Welcome to the forums!

After almost a decade of trying to wrap my head around gammas, colour spaces, colour management, and Resolve, I think I am finally coming out the "other side" and have a semi-clear understanding on things.

As you are starting out, I'd suggest the following:

• The overriding principle is "if it looks good, it is good".  If you're getting good results then (to a reasonable extent) it doesn't matter how you did it.  If you just want good results and whatever you did is working, then keep doing that and move onto other things and don't feel bad.
• Once I finally wrapped by head around Colour Management I realised that much / most of the online content about it is outright wrong, or potentially misleading.  It was only once I had a professional colourist explain it that I realised that even the way that people talk about it is wrong.  To this end, I recommend the video below - it is a step-by-step explanation of what colour management is, what it's for, and then how to go about it.  It takes a bit of effort to sit through and learn things properly, but I find that it's the best strategy long-term if you want to grow your skills:

• In order to build your skills in colour grading (not just colour management) I recommend ignoring all colour grading content online, except for Cullen Kelly (the presenter in the above video) and Darryn Mostyn (link to his channel), who are not only both working professional colourists, but are also very good at outlining the concepts as well as the tools.  If you're willing to pay for content then that's also a good way of getting high quality information.  Once you're feeling really comfortable and have watched a lot of their content then you can start watching content from other colourists, because you'll see that what they're telling you might be only half the picture, might be bad advice, might be flat out wrong, or might not even make sense.  
  When I started learning about colour grading and colour management, I read and watched a ton and I think it actively hurt my journey and just filled my head with a bunch of half-truths and confusing tips that I then had to un-learn once I realised how bad the information is.

Good luck!

[MarkAm]

On 1/4/2024 at 4:14 AM, kye said:

Welcome to the forums!

After almost a decade of trying to wrap my head around gammas, colour spaces, colour management, and Resolve, I think I am finally coming out the "other side" and have a semi-clear understanding on things.

As you are starting out, I'd suggest the following:

• The overriding principle is "if it looks good, it is good".  If you're getting good results then (to a reasonable extent) it doesn't matter how you did it.  If you just want good results and whatever you did is working, then keep doing that and move onto other things and don't feel bad.
• Once I finally wrapped by head around Colour Management I realised that much / most of the online content about it is outright wrong, or potentially misleading.  It was only once I had a professional colourist explain it that I realised that even the way that people talk about it is wrong.  To this end, I recommend the video below - it is a step-by-step explanation of what colour management is, what it's for, and then how to go about it.  It takes a bit of effort to sit through and learn things properly, but I find that it's the best strategy long-term if you want to grow your skills:

• In order to build your skills in colour grading (not just colour management) I recommend ignoring all colour grading content online, except for Cullen Kelly (the presenter in the above video) and Darryn Mostyn (link to his channel), who are not only both working professional colourists, but are also very good at outlining the concepts as well as the tools.  If you're willing to pay for content then that's also a good way of getting high quality information.  Once you're feeling really comfortable and have watched a lot of their content then you can start watching content from other colourists, because you'll see that what they're telling you might be only half the picture, might be bad advice, might be flat out wrong, or might not even make sense.  
  When I started learning about colour grading and colour management, I read and watched a ton and I think it actively hurt my journey and just filled my head with a bunch of half-truths and confusing tips that I then had to un-learn once I realised how bad the information is.

Good luck!

Thanks  so much for the great advice!  I've started watching Darryn Mostyn'stutorials and will also check out Cullen Kelley's and the video's you included here.  Thanks so much again!

[MarkAm]

Just following up - after watching a few Mostyn tutorials and Cullen Kelley's excellent beginner's tutorial, I'm back to my initial question.  Each uses a CST node to take log footage to a reasonable starting point to begin corrections and grading in Davinci Wide Gammut/Intermediate  (or to go directly to a Rec 709 output).  So if I'm shooting video on my D5300 with the Nikon Z-Log L picture profile, does that picture profile map to any of the Davinci input color spaces and gamma settings in CST, understanding that the color space transform just gets me to a starting point?

[kye]

A bit of googling found this - I would imagine this answers your questions?

 

[D Verco]

Are you talking about the eoshd log thing.

 

It says "Gamma curve is similar to V-LOG / S-LOG / F-LOG style."

 

So you'll have to try a few 

[MarkAm]

11 hours ago, D Verco said:

Are you talking about the eoshd log thing.

 

It says "Gamma curve is similar to V-LOG / S-LOG / F-LOG style."

 

So you'll have to try a few 

Thanks, yeah it seems no matter what picture profile I choose for shooting in camera I'm going to have to do a bit of experimenting to use a CST node.  The N-Log gamma in the input settings for DR are probably directly mapped to the N-Log picture profile in the Z9's and the earlier cameras outputting to an external recorder.  I might try it but I'm guessing the mapping will result in some color shifts somewhere.

I think my only other option to start off with a "what the camera saw" first-step approach that Mostyn and to some degree Kelley use is to use my X-rite color checker in lieu of a CST node and use DR's color chart matching feature.  I'm kind of surprised someone out there hasn't come up with a picture profile that maps directly with one of the default gamma settings in DR to make this process much quicker for folks shooting with pre-Z9 nikons.  (Heck, even finding tutorials videos with an example of a CST node used on Nikon non-N Log footage is hard to find.)  

Anyway, I'm a noob and the math/technical aspects of creating a pp that maps directly with a gamma in DR is way above my head so perhaps it's a bit harder and there may be some proprietary issues as well.  

[kye]

1 hour ago, MarkAm said:

Thanks, yeah it seems no matter what picture profile I choose for shooting in camera I'm going to have to do a bit of experimenting to use a CST node.  The N-Log gamma in the input settings for DR are probably directly mapped to the N-Log picture profile in the Z9's and the earlier cameras outputting to an external recorder.  I might try it but I'm guessing the mapping will result in some color shifts somewhere.

I think my only other option to start off with a "what the camera saw" first-step approach that Mostyn and to some degree Kelley use is to use my X-rite color checker in lieu of a CST node and use DR's color chart matching feature.  I'm kind of surprised someone out there hasn't come up with a picture profile that maps directly with one of the default gamma settings in DR to make this process much quicker for folks shooting with pre-Z9 nikons.  (Heck, even finding tutorials videos with an example of a CST node used on Nikon non-N Log footage is hard to find.)  

Anyway, I'm a noob and the math/technical aspects of creating a pp that maps directly with a gamma in DR is way above my head so perhaps it's a bit harder and there may be some proprietary issues as well.  

If you shoot a test shot of the X-Rite then you should be able to just pull it into Resolve and have a bunch of guesses on CST settings and just watch the vector scope and waveform to see which are closest.  Just remember to shoot the chart with high CRI lights and expose it well.  I don't know the exact theoretical way to shoot them, but I shoot my chart by just exposing normally and it comes out fine in the charts with the right CST.  

Even if you only find a combination that is close, it's like I said before "if it looks good, it is good" and unless you're doing scientific work or commercial shoots then the colours don't have to be exact, and if you're doing narrative work then the colours are best when being true to the emotion of the scene, rather than the specific objects that were filmed in the scene.

[MarkAm]

Just a follow up - I've been playing with D5300 footage shot in the Neutral and the Z-Log L profiles.  For Z-Log-L, an input color space of Rec 709 or srg with the input gamma for Fuji F seemed to get me to a quicker good starting point when using a CST starting node on a non-color managed workflow.   

But with Nikon's neutral profile, I found it odd that input settings of Rec 709/Rec 709 gave me a slightly more contrasty starting point than when selecting srgb/srgb for the input color space and gamma settings.  I assumed the results would be identical.  Selecting Rec 709 or srgb for input color space made no difference - the shift was caused by the different srgb or Rec 709 gamma setting.  What's more, leaving it to Davinci Resolve to do the conversion in an automatic color managed workflow led to a third different starting point that was in between the Rec 709 and srgb gamma input settings in terms of contrast.  And taking a closer look at the gamma settings Davinci chose for the managed workflow, all Davinci would tell me was that it was an "other" gamma setting.

So poking around the interwebs I came across this handout on Nikon's Color Management System: https://cdn-10.nikon-cdn.com/pdf/manuals/Scan4/NikonScan-4_CMS_en.pdf.  While it is focused on color settings for Nikon scanners (and their previews) it does discuss Nikon's color management generally.   Long story short - the handout points to an srgb color space and gamma 2.2 for images that are previewed on window's based computers.  And sure enough, in Davinci Resolve when I selected srgb and gamma 2.2 for the input color space and gamma on a starting CST node, the results were identical to the results that came up in the Davinci Resolve automatic color managed workflow.  So somehow Davinci Resolve seemed to pick gamma 2.2 for its input setting when doing the transform on its own.  

Granted the gamma shifts are not that dramatic from one setting to the other and we're just "transforming" from Rec 709 to Rec 709 or srgb to srgb.  But still, for what it's worth, a Rec 709/srgb with gamma 2.2 might be as good a place to start when input color space and gamma settings are required.