Best Log to work with?

After finding out about the color space transform OFX node, I have just realized I can match any camera to any other, but since finding this out, I have no idea which Log space is the best to color grade in. I know S-Log3 is really hard to grade, V-Log is relatively easy, and Blackmagic Film Gen 4 is medium in difficulty. I see some recommending LogC (Arri), some recommending DaVinci Wide, and some recommending just jumping straight into conversion to Rec.709 and working with that. I don't use color management since I prefer just doing everything through nodes and work solely in R709 for delivery to YouTube.

Which Log is most friendly and flexible for grading? Arri LogC, DaVinci Wide, straight conversion to R709, or something else that I haven't touched on like Canon Log3 or something.

All responses welcome!

I know S-Log3 is really hard to grade,

It is? never had an issue with it.

I know S-Log3 is really hard to grade,

It is? never had an issue with it.

For me I found it hard to grade personally compared to V-Log and F-Log but that's from personal use. Different types of log seem to be easier to grade for some people than others. I'm glad that it is easy for you but the nice thing about the color transform is that we can effectively convert log right inside of the editor.

It means I could get the hardest to grade Log in the world and switch it to something easier.

I can't really use S-Log3 on the actual camera either since it is one of the older 8-bit Sonys and can't afford an extra $3500 camera to be able to use it, sadly.

After finding out about the color space transform OFX node, I have just realized I can match any camera to any other, but since finding this out, I have no idea which Log space is the best to color grade in. I know S-Log3 is really hard to grade, V-Log is relatively easy, and Blackmagic Film Gen 4 is medium in difficulty. I see some recommending LogC (Arri), some recommending DaVinci Wide, and some recommending just jumping straight into conversion to Rec.709 and working with that. I don't use color management since I prefer just doing everything through nodes and work solely in R709 for delivery to YouTube.

Which Log is most friendly and flexible for grading? Arri LogC, DaVinci Wide, straight conversion to R709, or something else that I haven't touched on like Canon Log3 or something.

All responses welcome!

None of the above. You're going about it the wrong way.

My recommendation is to go with a Color Managed workflow and either use ACES or RCM with DaVinci Wide Gamut / Intermediate. This will allow the footage to be transformed to a higher DR and Gamut, then mapped down to 709. You're basically trying to achieve something similar with a CST.

Of course, S-Log3 is crap out of a camera with only 8 bit recording. From a professional camera it's easier to grade than S-Log2.
That is, if you like Sony's skin tones…

Of course, S-Log3 is crap out of a camera with only 8 bit recording. From a professional camera it's easier to grade than S-Log2.
That is, if you like Sony's skin tones…

I find them sort of natural but greenish in Log which is why I use HLG3 as the acquisition format

After finding out about the color space transform OFX node, I have just realized I can match any camera to any other, but since finding this out, I have no idea which Log space is the best to color grade in. I know S-Log3 is really hard to grade, V-Log is relatively easy, and Blackmagic Film Gen 4 is medium in difficulty. I see some recommending LogC (Arri), some recommending DaVinci Wide, and some recommending just jumping straight into conversion to Rec.709 and working with that. I don't use color management since I prefer just doing everything through nodes and work solely in R709 for delivery to YouTube.

Which Log is most friendly and flexible for grading? Arri LogC, DaVinci Wide, straight conversion to R709, or something else that I haven't touched on like Canon Log3 or something.

All responses welcome!

None of the above. You're going about it the wrong way.

My recommendation is to go with a Color Managed workflow and either use ACES or RCM with DaVinci Wide Gamut / Intermediate. This will allow the footage to be transformed to a higher DR and Gamut, then mapped down to 709. You're basically trying to achieve something similar with a CST.

What if I have footage from multiple cameras then? Usually the input color grade can only be one thing unless nodes state otherwise using color transform

The new RCM can handle that. I can highly recommend the tutorial by Rippletraining, done by Alexis van Hurkman himself.
https://www.rippletraining.com/products ... esolve-17/

The new RCM can handle that. I can highly recommend the tutorial by Rippletraining, done by Alexis van Hurkman himself.
https://www.rippletraining.com/products ... esolve-17/

So I suppose that each clip has its own input thing (HLG BT.2020) and then the timeline is say, Davinci Wide, then the output is R709 2.4?

For example, yes.

What if I have footage from multiple cameras then? Usually the input color grade can only be one thing unless nodes state otherwise using color transform

I recommend using an ACES workflow, group all the clips for each camera log, and have an IDT and ODT nodes for the pre-group and post-group clips respectively to do the transforms.
Then do the grade. There are good tutorials on ACES like this 3 part intro...

So I suppose that each clip has its own input thing (HLG BT.2020) and then the timeline is say, Davinci Wide, then the output is R709 2.4?

This is how color management works, yes. You set appropriate input transforms based on what your different footage is (if they differ from default interpreration) and CM moves them all to intermediate colorspace where grading happens. Project and metadata interpretation defaults are just ”use this unless I override it”, not hard fixed ”it is what it is”, sometimes it seems to trip people over. Set your IDTs as appropriate for every clip, changing it manually if needed. In the end, output transforms move data to whatever colorspace you need for display or export.

What if I have footage from multiple cameras then? Usually the input color grade can only be one thing unless nodes state otherwise using color transform

I recommend using an ACES workflow, group all the clips for each camera log, and have an IDT and ODT nodes for the pre-group and post-group clips respectively to do the transforms.
Then do the grade. There are good tutorials on ACES like this 3 part intro...

Thanks, I'll go take a look.

So I suppose that each clip has its own input thing (HLG BT.2020) and then the timeline is say, Davinci Wide, then the output is R709 2.4?

This is how color management works, yes. You set appropriate input transforms based on what your different footage is (if they differ from default interpreration) and CM moves them all to intermediate colorspace where grading happens. Project and metadata interpretation defaults are just ”use this unless I override it”, not hard fixed ”it is what it is”, sometimes it seems to trip people over. Set your IDTs as appropriate for every clip, changing it manually if needed. In the end, output transforms move data to whatever colorspace you need for display or export.

I have found when working using color management, the lift, gamma, and gain wheels act way differently. Have you experienced this?

It depends on working space where color manipulation happens. Different working spaces express same colors with different RGB values and thus same applied math produces different effect.

What ”colorspace aware tools” means (came to R in v17) is that there is an additional implicit working space where color ops are applied, and data is moved to/from that to produce predictable response independent from timeline space.

So I suppose that each clip has its own input thing (HLG BT.2020) and then the timeline is say, Davinci Wide, then the output is R709 2.4?

This is how color management works, yes. You set appropriate input transforms based on what your different footage is (if they differ from default interpreration) and CM moves them all to intermediate colorspace where grading happens. Project and metadata interpretation defaults are just ”use this unless I override it”, not hard fixed ”it is what it is”, sometimes it seems to trip people over. Set your IDTs as appropriate for every clip, changing it manually if needed. In the end, output transforms move data to whatever colorspace you need for display or export.

I have found when working using color management, the lift, gamma, and gain wheels act way differently. Have you experienced this?

Yes this is well known. The workaround in ACES is to create an in and out conversion ‘sandwich’ node from ACES to REC709 and back again so all your tonal adjustments using LGG happen in REC709 which is where they are designed to operate. In RCM LGG are unintuitive at best and unusable at worst. BM really needs to fix this as shadow mid and highlight work over very Narrow tonal ranges and are not designed for large tonal shifts as they create ugly artefacts at their tonal limits. You may have more luck with the HLG controls but I don’t like the loss of ‘perceptual’ saturation when contrast is added.

This is how color management works, yes. You set appropriate input transforms based on what your different footage is (if they differ from default interpreration) and CM moves them all to intermediate colorspace where grading happens. Project and metadata interpretation defaults are just ”use this unless I override it”, not hard fixed ”it is what it is”, sometimes it seems to trip people over. Set your IDTs as appropriate for every clip, changing it manually if needed. In the end, output transforms move data to whatever colorspace you need for display or export.

I have found when working using color management, the lift, gamma, and gain wheels act way differently. Have you experienced this?[/quote]
Yes this is well known. The workaround in ACES is to create an in and out conversion ‘sandwich’ node from ACES to REC709 and back again so all your tonal adjustments using LGG happen in REC709 which is where they are designed to operate. In RCM LGG are unintuitive at best and unusable at worst. BM really needs to fix this as shadow mid and highlight work over very Narrow tonal ranges and are not designed for large tonal shifts as they create ugly artefacts at their tonal limits. You may have more luck with the HLG controls but I don’t like the loss of ‘perceptual’ saturation when contrast is added.[/quote]
I'll take a look at the sandwiching, thanks. Would NR go after the sandwich, before, or somewhere in between the grade? I know sharpening would absolutely go last.

Trying to sort out the conflicting information you're getting here may not be the best approach. Alexis van Hurkman, who I believe is involved in writing Davinci manuals, has online courses for color management, including the modifications introduced in 17. That may be one place to start. And he's a working colorist. The same may not be true of everyone who offers advice here.

"I heard it on the internet" or "I read it on the BMD forum" may not be the best way to find your way around this complicated subject.

I do NR, basic color balance and exposure normalisation before the REC709 sandwich in my ACES workflow (which learnt from a 'working colorist'.......)

Trying to sort out the conflicting information you're getting here may not be the best approach. Alexis van Hurkman, who I believe is involved in writing Davinci manuals, has online courses for color management, including the modifications introduced in 17. That may be one place to start. And he's a working colorist. The same may not be true of everyone who offers advice here.

"I heard it on the internet" or "I read it on the BMD forum" may not be the best way to find your way around this complicated subject.

Will take a look. Thanks!

Trying to sort out the conflicting information you're getting here may not be the best approach. Alexis van Hurkman, who I believe is involved in writing Davinci manuals, has online courses for color management, including the modifications introduced in 17. That may be one place to start. And he's a working colorist. The same may not be true of everyone who offers advice here.

"I heard it on the internet" or "I read it on the BMD forum" may not be the best way to find your way around this complicated subject.

Alexis van Hurkman literally wrote the book on color correction, and was employed by Blackmagic Design as the lead writer on Davinci Resolve's manual for the last 10 years up until he moved to Frame.io almost 2 months ago. When Resolve 17 came out last year (and when that ripple training was released), at the time he was quite literally the source of truth from inside Blackmagic Design to explain to the world how the software works.