A question that comes up quite a bit is – how do I get the LUT I have been using in the camera into DaVinci Resolve.
There are two parts to this. The first is how do you get the LUT you are using in the camera, out of the camera. Perhaps you want to export the s709 LUT or perhaps some other LUT.
To export a LUT from the camera you can use the embedded LUT option that is available when using the Cine EI mode.
If you turn on “Embedded LUT” on the camera and record a clip the camera will save the LUT on the SD card under:
FX3/FX30 – private – M4ROOT – GENERAL – LUT folder.
FX6/FX9 – private – XDROOT – GENERAL – LUT folder.
Then to get a LUT into DaVinci Resolve the easy way is to go to the Resolve preferences colour management page, scroll down and there is an “open LUT folder” button that will open the LUT folder. Copy your LUT into this folder. Then click on the “Update Lists” button. Now your LUT will be available to use in Resolve.
If you like what you read here, please also visit and subscribe to my YouTube Channel by clicking here. Thanks.
Semi-related question that I have been meaning to ask, Alister – do the LUTs you have created include a Rec. 709 CST? And should one use Cineon gamma with them?
A LUT for S-Log3 that looks good on a 709 display will include a CST as part of the LUT.
Why would you use Cineon gamma with a LUT for S-Log3? Cineon was a special gamma based on film density designed for film to digital transfers in early DI workflows. It isn’t really designed for modern all digital workflows.
Well Cineon gamma is used in conjunction with a “look” LUT when such LUT is used after the final CST to rec709 and a normal gamma like 2.4. Best example is the “Darstyn node tree” designed by Darren Nostyn and being used by a lot of professional colorists.
But all that is doing is altering the contrast because it creates a gamma error. It’s not something specific or unique to Cineform gamma, it just happens to add a bit of contrast because it introduces a gamma miss-match and whenever you add a gamma miss-match the contrast will change. For some reason this has become considered to be some sort of magic ingredient when the reality is there really isn’t anything special or unique happening, it’s a mistake that results in additional contrast.
it never ceases to amaze me how stuff like this takes on almost mythical properties.
Well you may be right. Like every LUTon its own, it works better on some clips than others. And it works better with LUTs that don’t have built in CSTs. BTW, it is not Cinefom but Cineon. I assume that they are two different things.
S-Log3 and Arri LogC are both based on Cineon gamma. The “C” in LogC stands for Cineon. S-Log3 and LogC both use almost exactly the same curve as Cineon but they are shifted 1 stop darker. The curves were kept the same so in post production there would be very little difference in the grading process, just shift S-Log3 content up 1 stop and apply the same grade as you would for a Cineon film transfer and the end result would be similar. If you apply the Cineon CST to S-Log3 material it will make the footage 1 stop darker and very slightly increase the contrast in the shadow range because it will put and extra stop of the original range into the shadow range. This would be exactly the same as adding a properly corrected -1 stop offset to the S-Log3 footage before the application of the LUT or using a LUT with a -1 stop offset. If people are finding using Cineon in the CST beneficial, then it suggest the footage has been exposed a stop brighter than the target exposure for the base ISO/base EI.
In a colour managed workflow where the EI metadata is correctly applied such as when using MXF material in Resolve the ISO adjustment tool in the raw control panel will also have the same effect.
Again: There is nothing magic about using the Cineon CST, it is introducing a gamma miss-match, but it just so happens that because S-Log3 and Arri LogC are both based on Cineon but with a 1 stop offset, that the downward gamma shift it introduces acts as a -1 stop exposure shift which can be beneficial for dealing with footage that has been exposed brighter than the recommended base levels. Alternatively you can just bring the exposure down by 1 stop prior to adding the LUT or better still use a LUT designed for the exposure that has been used. A LUT with the correct offset will also put the highlight roll of in the the right place allowing the best possible highlight retention. Dealing with under/over exposure has always been an issue in LUT based workflows because LUT’s are extremely sensitive to the exposure used and resulting input values. Using the Cineon CST instead of S-Log3 is a bodge that gives a -1 stop offset, something that could be better achieved with a proper exposure correction or a LUT correctly built for the actual exposure used.
Or better still get rid of LUTs with all their shortfalls and limitations and start saving the actual grade instead.
As always, thanks, Alister. Apparently the Darstyn node tree is really designed for grading using the Davinci Resolve CSTs in the “Effects” of the Color Tab and using the LUTs that come provided by Davinci, which do not have CSTs embedded in them, such as Davinci’s Film Look LUTs. You might want to give a look to this – there are lots of professional colorists on Youtube, in addition, to Mostyn promoting this node tree. I assume that to use LUTs that have embedded rec709 CSTs one would simply turn off the CST to rec. 709 as well as the CST to Cineon gamma before the “Look” node. https://www.youtube.com/watch?v=kdTMRQP_V7E
In the video linked the LUT is added after the DaVinici wide gamut to Rec709 CST node, so everything is now in 709 gamma/colourspace.
The reason why Darren adds a new CST is to go from Rec709 gamma to Cineon gamma because the Resolve film look LUTs are designed for the Cineon input gamma. He then doesn’t need to add another CST after the LUT node because the LUT includes a Cineon gamma to 709 CST within the LUT so the output of the LUT node is back in 709 colour/gamma again.
When you say “get rid of LUTs and save the actual grade instead,” are you suggesting not using CSTs or using the CST “Effect” in Davinci?
LUTs degrade the image, even the best 65x LUTs break the image into 65 correction bands, a 33x LUT 33 steps. These bands\steps can introduce artefacts into the image as the corrections step from one band to the next – although with good interpolation most of these artefacts should not be seen, there are certain compression processes or scaling processes that may occur downstream of the grade that can cause the steps to become visible again. CST’s are useful tools but when you start converting colour spaces you introduce more mathematical errors. These should be very small, but if you start flipping back and forwards between different spaces they start to add up and will also introduce gamut limits. In Darrens example when he adds the film look LUT first he has to come out of that nice big wide gamut colour space into cineon gamma + Rec709 colour so you are limiting the colour space to that of Rec709 for everything beyond that point. Then he adds a LUT that includes a CST from Cineon to rec709 gamma, so now everything is limited to Rec709 dynamic range. Even if you then added another CST to go from Rec709 back to a wider gamut, you will no longer have all of the original information in the captured footage to work with because you added a CST bottleneck. LUTs are a quick fix solution but they are very limiting because 99% of the time they incorporate a rec709 CST.
Alister , Is there any difference between the s709 LUT that you can download from Sony and the s709 LUT that is in the camera that your describing here?
Couple of CST questions:
– I’m curious that when I use a CST in Resolve or set it up for Resolve Color Management, it always creates an overly saturated image to my eyes. How come?
– What do you mean when you say above that “A LUT for S-Log3 that looks good on a 709 display will include a CST as part of the LUT.” Is that different from the “Sony” setting that comes built in wth resolve?
– Also I’ve always been a bit confused because color of the FS7 and the FX9 are different – the FS7 always looks ‘pinker” to me and the standard LUT for FS7 seemed to be 709A as opposed to s709 for the FX9. Actually I found if I used s709 for FS7 footage 7609A for FX9 they matched better . So why is there no separate settings for CSTs with those cameras.?
Thanks – Hope those questions make sense.
Any LUT that goes from S-Log3 to something viewable in 709 contains a CST. There are very, very few LUTs that are not also CSTs. If it didn’t have the CST element, it would be S-Log3 in, S-Log3 out.
CST’s transform between colourspaces. They do not add or change any artistic intent and are not meant to be used for camera matching, they simply go from one colourspace to another, hopefully retaining the “look” of the source material, just now in the new colourspace. The look of Sony’s cameras prior to the FX9 (and Venice) was meant to be different, it was meant to match earlier generations of cameras. A CST should not change this, the look of the earlier cameras and later cameras is designed to be different.
ACES has a slightly different approach which is to include a camera specific transform at the input, based on how the camera responds ro light to the ACES colourspace. This is designed to remove any camera differences. But for the Sony cameras there are very few options, although when using X-OCN there are transforms specifically for each Venice camera.
The in cameras s709 LUT and the downloadable s709 (v2) LUT should be the same.