good refresh. many times we get too much info (especially when learning new stuff) because we are all in our own journey. we hear the same thing 6 months a part, it will hit differently. Good stuff.
"many times we get too much info" This really resonates with me. Not to mention the information overlap from the same channel. This isn't to say that it's a bad thing necessarily but very few channels like Cullen's, know how to handle that deftly.
Great information. The color management that Resolve does is excellent and used even on our biggest features. The auto color management is good for many projects. It takes some knowledge to set up the CSC nodes to match the RCM full automatic mode. The node based approach gives you flexibility when cameras or footage is not properly supported by the built in transforms. As well it gives flexibility on using output luts. This is typically how we utilize the resolve tools as it provides the consistent built in help with color management, but can be customized if something weird comes along.
As you said it takes knowledge to imitate the RCM with CST. Up until watching this video I have not succeeded achieving exactly the same results. Great video! At least now I know it was user error
Suspected this to be the case but always good to see it in action. The group/timeline tuck trick for the bread of the sandwich.....awesome technique. Huge help. Thank you.
Just a note: when using Davinci YRGB, in the project settings, the "timeline color space" setting will affect the CST nodes if you choose the "Use timeline" option for gamut/gamma for the node. I would never do this but it's helpful to know that they're linked that way.
I was just wondering about this. For output gamut/gamma couldn’t you just put “Use timeline” in the CST in node? As he already selected the same setting as timeline gamut/gamma in the project settings.
@InLightVFX: Can you please elaborate? So are you saying that we should never leave it as “timeline color space” ? I thought he said it didn’t matter what you set that to?? I’m a bit confused now. Thx in advance.
@@JSalas-fh9xb The color space that you choose here at 3:02 for "timeline color space" (in this case set to Davinci WG/Intermediate) is the color space that will be used in the CST node when "use timeline" is selected (which is set by default as you can see at 3:14). If you're still confused, just basically never select "use timeline" in the CST node - and then it doesn't matter what you put for the overall "timeline color space".
Great point and great discussion here. My preference is to always avoid 'Use timeline' and take the extra few seconds to explicitly state my intention for the CST...just a personal best practice
Thank you Cullen - Not new news. BUT always useful to hear you working methodology and the reasons behind your working concepts. Your group application of CSTs confirmed my own methods for multi-cam grading.
Thanks to your videos, I'm familiar with it, but I'll teach beginners RCM. I taught them node-based before and some got confused. RCM may be a set-it-and-forget-it way for beginners to get good results.
Oh wow - I can't thank you enough for this explanation - I was assuming they did something similar but wasn't 100% sure and I was also trying to work out the logic as to why one would go with a CST in the node graph, when color management can be setup to do the same thing under Project Settings. Now I can see the logic behind a comprehensive node-based workflow. Brilliant job!
Thanks for the refresh and confirmation they’re both the same. A related aspect I’m not as clear on when using Group CST is where to place a LUT for a particular look. Especially when the LUT is made for Rec709 footage.
If the LUT is made for 709, you’d place it after the CST that outputs to Rec 709. For example, my first node is always a CST that converts my log footage to DWG (Davinci Wide Gamut and Davinci Intermediate) then I have 6-7 nodes after for color grading, and then finally my last CST that’s converts it from DWG to Rec 709. So you’d add another node after the CST converting DWG to 709 and add the LUT on.
@@nowhereland_2 thank you. That makes sense, and that’s what I’ve done at the clip level. I wasn’t sure if this would be same at the Group level as an Out CST as well.
@@inbloom9083 If the LUT is made to convert Slog to Rec709, you’d place it right at the end of the node tree, so that all your grading is being done in the wider s-log color space and you’re able to see how it changes the final Rec709 image (or whatever the output color space is for the project). For example, a simple node tree could be: Exposure > WB > Contrast > Slog to Rec709 The last node is where you’d right click on the node and add the Slog LUT.
Very helpful clarification. Newbie learning to use Resolve, but finding it incredibly useful compared to using adjustment layers in Premiere. Huge thanks to you and Darren Mostyn for making picking all of this up very easy for a beginner.
As a noob on da vinci resolve I've been learning it quite intensely over the past 6 weeks and refreshers like these are gems when found. I think little updates on the under the hood parts of grading are important for the newcomers and regulars.
I always use the node-based workflow because I often work with graphics and lower 3rds, and the workarounds to bypass RCM or reverse the color management applied to them is more of a tedious workaround than it needs to be, especially with transitions and effects. With transitions, I end up having to use nodes anyways because I have to bypass both clips on either side of the transition because there's no way to color manage the transition itself.
Great video! thanks for sharing. I have some questions to ask you. What happens if the timeline has graphic files or different cameras, wouldn't it be better to handle it at the clip level? And secondly, can you make a CST group (as you did) and put together other groups by scenes? can a clip be in two groups? thank you so much!
A little more detail to finish. Round trip past the RD3 clip to prove the point on all clips. Also, perhaps compare with using camera manufacturer LUTs for the previously poorly informed - who might mistakenly double up in colour management. In another forum that I lurk in, colour management double up problem is surprisingly common. And an easy trap to fall into. Cheers. Was looking forward to your next drop of wisdom.
Thanks for clearing this up, I was going crazy trying to figure it out myself. But now I'm curious as to what the difference is between the ACES Transform node and the Color Space Transform node.
Great question! Yep, in this scenario you'd want to place your output transform somewhere else. Here's a video where I break down several options: ua-cam.com/video/ReHA9bhT168/v-deo.htmlsi=tG_3XlN6_fIl_iAU. Hope this helps!
Yep, Brandon nailed it. Main reason is that the images I'm grading for YT demos are primarily going to be viewed on gamma ~2.2 displays, rather than gamma 2.4. For client-facing work I still primarily use gamma 2.4.
Wow man, thanks for showing how to use grouping to add the CST nodes to multiple clips without having to do it for all clips individually, huge timesaver for real!! You earned a new sub! Thanks for sharing.
I started using the 'group' function recently and have found it to be helpful and efficient. I am also a fan of node-based color grading, I tried the Resolve color management system for a while but the node-based structure works better for me. Thank you for posting this video.
Wonderful information Kelly. I am beginning to use DaVinci Resolve and learning some important stuff is really helpful especially to color grade multiple clips at the same time. I initially thought you would like DaVinci Color Management better as you have an array of DaVinci Keyboards and other color management tools. Thanks for sharing.
this is new to me did not know they were both the same,,, good to know that i will have to try both ways to see what i like the best... but i do have one question what is the different between Rec.709 Gamma 2.2 and 2.4
Funny coincidence: Yesterday I was discussing with some other filmmakers which is better. So thank you for the short and very clear clarification! I prefer color management because of its simplicity (fewer knots, fewer sources of error when you are short on time). In normal cases (shooting in log, fairly simple color grading) I don't see any advantage in using CSTs.
Generally no -- my take is that if there's something in your working space or CST from camera to working space that limits you, something is wrong with the choice of working space or the CST.
Could you explain the reasoning behind having the output as 709 gamma 2.2 instead of 2.4? Is your monitor setup for 2.2? I understand that internet delivery likes 2.2 over 2.4. I usually do a timeline CST of 2.4>2.2...
Sure! Gamma is set to 2.2 for YT content because it matches the gamma of most of the web devices the videos will be watched on. I have a 2.2 preset on my reference monitor that I flip to when recording demos, but for client-facing work I'm still generally mastering in 2.4.
This is head clearing and useful! However, the question is, how come you can set the input gamma using CSTs, but in the settings you only have the option to choose the input colour space?
Hey !! i really really love your video. would you please make a video on how to analyze reference for grade like exposure, contrast, saturation, Balance, skin tone, look and whatever its.
I'm always messing with some footage which is not on the list at "In Put Color Space"...for example Samsung Mobile footage and Dji drone shot in D Log-M mode, Those are not quite perfect when I put that footage in Rec709 input space
What do you set your General settings to. I see you're using a Mac, but with a non-Apple display. Does this mean you're selecting the "Use Mac display color profiles for viewers" option? Would you or would you not be selecting "Automatically tag Rec709 Scene clips as Rec709-A"? Please explain, thank you.
Hello folks! How do you use the group post clip level of your node graph vs the timeline level? Do you use the timeline level only for the output transform or what else do you use it for? Thankful for answers.
I wonder why scopes then show slight differences though when you wipe between the two images - is it a scope rendering issue or there is still some minute discrepancy between the two?
Great catch! Difference is in the Gamut Mapping settings I'm using in my CST vs the ones used by Resolve Color Management. Good example of the extra control gained when color managing with CSTs
I’m finding myself using cst more than project. Even though for ARRI raw I like the project based. I’m refreshing my Reel and playing with my original footage and the CST and group ling has been very practical
nice content !! I need some help with a big render problem. How do I fix the color wash out issue on output render. I use a Windows 11/ Go Pro 11 Footage (Natural_10Bit_4K 50 Frames). I have already applied Preferences as: Use Windows Color Management... and Color Management: Davinci YRGB Color managed. Rec-709 A / Davinci Wide Intermediate/ SDR100 / Rec 709-Gamma 2.2. Could you please tell me where I am going wrong that my output footage turns washed out? thanks in advance...
When using auto color management, how would you handle an edge case that required correction "pre-transform?" For example i sometimes will need to adjust exposure before the input transform to avoid a clipped image in the dwg working space.
I know you can get identical results, but I am curious about which settings need to be tweaked so that the waveforms don't move. You can see as you toggle on and off at 4:10, in the green and blue channel, a couple of small shifts. While probably not visually noticeable, I am curious what causes the different waveforms. Definitely agree that the results can be identical and they are doing the same operations in the same way, but it seems that there is a tiny settings difference in one of them. If i ignore the image, i can tell when you are switching between by watching the waveform change.
Love your eye for detail! You're actually just seeing the discrepancy between the saved still and the exact same image being live in the timeline. Try grading a shot, saving a still, then wiping to it without changing anything...you'll notice the same slight shifts.
It would have been very interesting to see how this works and looks with Automatic color management. Would it also be the same? Is there anything that speaks against Automatic color management if you find the other ways too confusing and don't need the same control over the process but still want something solid? Can you achieve that with Automactic color management and a simplified workflow?
Thanks for the great explanation. One detail, when you set Out DRT to Luminance Mapping, why is that? Because it has a hue shift compare to Davinci (the default option) So I wonder why that decision. Thank you.
I have a question about color management when working under a LUT that isn't DWG>DWG! My cinematographer is insisting I work under a LUT they used in-camera during production. This LUT takes S-Log to Rec.709. Where would I apply CSTs in this situation if I wanted to work in DWG? SLOG>DWG>SLOG>LUT?
BTW I knew all you talked about because of your great previous videos. Can you do a video how you might use the new grouping 4 dots that are available??
I watched a video two days ago about this, so it's not new (anymore) for me, but it was good to get a second voice about this topic. :) Just one question: Could you tell me, why you are using Gamma 2.2? I've learned from Darren Mostyn, that he uses Gamma 2.4 and many other UA-camrs use this. What's the difference?
Hello! Maybe you know way how can I use this "sandwich" on a MacBook? Because I am Mac user and I have to select Rec.709-A in the settings... If I don't do this, preview in Davinci will be different from the render in QuickTime...
One thing I have found different between the two approaches is related to qualifying. When doing so using RCM, I have no problem since it seems to qualify based on the Rec709 image (despite still working within DWG), however when using CST's, it's more difficult since it's using the, I'll say, washed out version from the DWG colorspace. I notice this when highlighting. I prefer using CST's because it just makes it easier for me when using effects or LUTs that I prefer using in Rec709
@CullenKelly Not sure if you’re aware, but the only link that is working is the RED raw film link, all others give an error. I noticed it initially in my attempt to download your E-Book…
You don't need an out converter, just select the timeline in WG/intermediate color space and gamma, then set up the right CST for the clip so it appears correctly on your screen. You can start grading. The Out node mathematically is not necessary. You just select the desired color space TAG on the Delivery page.
Thanks for the video! Only question is - why not use the "CST Out" node in the "Group Post-Clip" tab vs. the "Timeline" tab as shown in the video? Since using the "CST Out" on the entire timeline would then affect the whole project vs. affecting only the assigned group clips we need, right?
You got it! And great question: reason is because I actually want the 'Out' CST to affect everything, not just the one group. Everything will be mapped into DWG from its original camera color space, and from there everything will receive that same transformation to get out to display space. Hope this helps!
I've seen all colourist doing this. or a clip based CST. But for most of us who have to do everything from editing to color, graphics and subtitles working in groups is way more efficient.
@@daniellrodriguezs Great point! I actually did a full video looking at different places you can put your output transform depending on your needs/preference: ua-cam.com/video/wma5NvhbNUk/v-deo.html
@@CullenKelly found it right after I left this 😂 thanks for leaving the link though! I like the shared nodes as well for things that cannot be done in a post group clip like for example Luma mates. I found some cool effects with it but unfortunately cannot be applied to pre or post clips adjustments so I found shared nodes works just fine.
At the 2 minute mark, after you set up RCM with input color space ARRIlogc3, what is the reason that you had to again specify the ARRIlogC3 input color space for clips 2 through 6?
Great question! This is actually another con of color managing in project settings...Resolve will fairly frequently make its own guess about the source color space of your shot, and it's usually wrong. That's what happened with shots 2-6, so I had to correct that wrong guess. Hope this helps!
I’m grading Apple log footage following the steps of the video to grade it in DWG before converting it to Rec 709 Gamma 2.2, and I made sure to specify the correct tags in the color management settings. However, the image still looks flat on my iPhone and MacBook, except if I use 400 for Max Output (nits) instead of 100. Is there a specific reason why 400 works better than 100 in my specific use case, and if so why?
Hello Cullen, I believe (unless I missed something), that is not true anymore that in Davinci YRGB, the timeline colorspace and output colorspace are just tags and never influence the grading. Actually, the HDR wheels are colorspace sensitive even in Davinci YRGB, so the timeline colospace has an impact in DVYRGB for that tool.
I know it can be a pretty dumb question, but is killing me and I cant find any clear answers on it... in a CST workflow, why setting up the timeline color space as DaVinci Wide Gamut in the project setting? like, if I'm controlling in what space the operations are happening via the CSTs in my node tree, what exactly changes when I change that setting just in the project settings? I know there's something changing but cant point out what it is and cant decide what's the appropriate setting. I assumed a long time ago that it affected nothing so I just leave both options at my output space (Rec709 gamma 2.4/2.2), making all the required transforms via CST, all went smoothly, all my tools behaved in a way that worked and I was satisfied with the results, but one day I change the timeline space to DWG in the project settings and suddenly various tools changed the way they operate (halation, color compressor, color wheels), and I really don't like it, specially when qualifying or using HSL curves. The thing is that now I'm fearing that maybe there's information being clipped or lost if I leave the timeline space in Rec709 😰
Hi Cullen - I like to set my timeline luminance level to 10000 nits using RCM but notice if I'm using graphics or Rec709 material the highlights get dropped way down to almost halfway on the waveform dulling the whites especially in graphics. My workaround was to add a CST node to the clip and set the custom max input to 100. Please tell me if this is correct or if there's a better way - thank you!
What if you do CST, the sandwich, in the entire timeline? I have experienced that the results are different if we use it in the timeline or clip by clip
You could do this, but then you'd effectively be grading in your camera space rather than your working space, which essentially removes the benefit of using the working space at all. Hope that helps!
I have graded my film with outputs Rec.709 Gamma 2.4 for Amazon, Tubi etc. However, I am having to do a version for theatres, and they want P3-D65 Gamme 2.6. What is the proper way to take my Rec.709 Gamma 2.4 grade to P3-D65 Gamme 2.6? Would right-clicking and creating versions work and just switch the output gamma and color space when I do the P3-D65 Gamme 2.6 versions?
Yep it would! In my case I don't often have text or other non-camera elements in my timeline, but if I did I might choose to place my output CST elsewhere. Actually did a full video on this topic if you're interested: ua-cam.com/video/ReHA9bhT168/v-deo.htmlsi=AgvTC_FF4tHKHlV2
Hai Cullen, appreciate all your videos. But maybe this question little bit noob question. What happen if the CST In and Out placed on Timeline not group? is it the same method? Are the CST In always in Group level, or we can do in Timeline level? Thanks Cullen.
Hey Rudolf! You can totally place your 'Out' CST in Group Post-Clip instead of at the Timeline level...only catch is that you need to manually place it in that spot for each of your groups. Here's a video where I do a full breakdown of the different places you can put your output/macro-level transform: ua-cam.com/video/ReHA9bhT168/v-deo.htmlsi=mWBtZypcPEupPQ-C Hope this help!
@@CullenKelly thank you so much Cullen. I trully thankfull with all your content. I was start color grading from 0 and learn a lot from your channel. Im not profesional colorist (yet) but i'm learning so much from you. Thank you.
Hello, thank you for this video ! Why don't you put tone mapping in the "in" corrector ? Is it better to apply the Pro Mist diffusion filter effect with the diffused glow after the "out" corrector ?
Sure thing! I don't tone map at the input stage because I don't to grade an image that's already had its highlights compressed...allows me more control and speed in the grade. Regarding diffusion effects, best to place these to the left of your output transform for a number of reasons, but the most simple one is that you'll get better-looking results. Hope this helps!
Hi! i'm a nooby on this, and i have a question that nobody talks about, probably couse is for stills... I'm using a Lumix S5 ii X, and y tried to make luts for "real time lut" while i was taking photos. But the lut doesnt look equal on the Lumix than in Davinci, working with sRGB>DWG>sRGB. First, i convert the RAW2 to DNG, and then work on Davinci with that Color Maanagement
This one's too complex to fully address in a comment, but I'd start with 1) confirming that your image is in sRGB when applying the LUT in-camera and 2) ensuring you're decoding the DNG into sRGB in Project Settings -> Camera RAW. You may also want to look at capturing in a log space rather than sRGB.
Does it affect how the curves or the gamma or the HDR knobs affect the picture? or is it the same values both ways. I get all confused on this color stuff.
if using the RCM instead of using CSTs, where can I know what color space I am grading? For instance: If there are two luts: 2383DWG to DWG, and 2383 Rec709, what's the correct position(node) I should place?
@@CullenKelly If I did the Color space&transform on Project setting ( instead of using the node tree to transform the color space), how did I know the picky eater (the LUT) was placed in a correct position (color space) ; i.e. the color processing mode is set up as DWG Intermediate, Out put color space is Rec.709 gamma 2.4, the Lut should apply in the DWG color space ( or 709 color space whatever), where should I put? what's the color space of Group Pre Clip level , Group Post Clip Level, and Timeline Level?
@ I was watching another colorist saying that cst are more “destructive” that lowering the saturation and finding the contrast point with a managed timeline said there’s more control im not sure if that was bs lol so just asking around
I always thought CST is better than Color Management, it all makes sense now. Tell me, does using the CST route slow down the performance of the computer or laptop? My lecture said by color managing with CSTs, it'd make my color grading performance very slow in terms of performance. True or False?
Hm, I've never experienced any difference in performance between color managing in project settings vs nodes, and I can't think of a reason why this would be the case...
So I’m still confused lol I thought it was gamma 2.4 not 2.2 I usually color grade my own stuff and I use s-log3.cine I was wondering if I now need to convert it to 2.2
Depends on the display you're mastering with and where you're delivering. For UA-cam I set my reference monitor to 2.2 since that's generally how the content will be viewed. For other scenarios I will set monitor to 2.4 and deliver 2.4
Cullen, just wondering why you placed the output CST for the Log-C group in the timeline instead of the one for end of the group? With different media types that are not using Log-C would that not adversely affect them?
Great question! The cool thing about the output CST when using a working space like DWG is that it works for all cameras, not just LogC3, since everything will be mapped from camera -> DWG on the front end.
Hi Cullen. Thanks again for all your input, you taught me everything I know today. Your photometric approach works well for me as a retoucher and digital tech (stills). One question. I find that my camera groups disappears when creating a new project. Any way of exporting or baking them into my settings? Appreciate. Charl 🥊🍺
Great question! You can maintain your camera groups by creating a template project that you save a copy of whenever you're starting a new project. To include groups, import a few clips into that template project, set up your groups, then delete the clips so that you're starting from a blank slate.
Thanks Cullen. This was indeed a useful review. I would however like to know more about your RCM custom settings, e.g. why is something your "preference". And how are your custom settings replicated in the CST version to produce the same result?
Awsome video, thanks for the explanation. I was testing node base vs project based color management, but when applying the Kodak 2383 LUT from Davinci resolve, I've got different results. Seems that the project based has the skin tones more neutral also the colors is cooler than the node based color management. Do you think is something I've done wrong or this is normal? Thanks in advanced.
I'm thinking if every shot in your timeline is of the same camera and format, RCM may be the better choice? ..as CSTs would be the better choice if there were multiple camera formats etc on the timeline? ..saves time .. 🤔
Could any of you DaVinciResolve experts (interested in solving difficult challenges) take about 5minutes to help me with the following problem that I am currently being confronted with? (Thank you so much in advance! ❤) In the Shorts video „How do I Get JVKE to see this?“ by „The_Magic_Piano“, you can see MIDI animations/ notes dropping down from top to bottom ending on the upper edge of the piano. I desperately want to recreate this idea for my own projects. I tried doing so in DaVinciResolve. After having imported the piano performance video and the MIDI animation video and dragging them both into the timeline I tried editing the MIDI animations by changing x and y positions, making it less wide, stretching it lengthwise, and changing pitch, yaw and rotation settings. No matter what I try, the MIDI Animation video always looks either distorted, the sizes of the animations are different from left to right or from top to bottom or the MIDI animations are Not aligned with the upper edge of the piano keys. Does anyone know how to solve this Problem? Maybe with a different method within DVresolve or a different software? The animations in his Shorts-video Are all uniform in size and width and match perfectly with the upper edge of his piano keys. Also the Level of the animations seems like it‘s accurately parallel to the wall. Lastly, the shadows on the wall in his video are nice but i guess that would be less of a challenge if the main problem is solved. I‘m so desperate for help and I don‘t know who to ask. Even ChatGPTs suggestions aren‘t successfull for me 😂😢 Thank you again and have a Great Rest of your day!❤
good refresh. many times we get too much info (especially when learning new stuff) because we are all in our own journey. we hear the same thing 6 months a part, it will hit differently. Good stuff.
"many times we get too much info"
This really resonates with me. Not to mention the information overlap from the same channel. This isn't to say that it's a bad thing necessarily but very few channels like Cullen's, know how to handle that deftly.
I had assumed that they are the same, but I appreciate the proof laid out clearly.
Great information. The color management that Resolve does is excellent and used even on our biggest features. The auto color management is good for many projects. It takes some knowledge to set up the CSC nodes to match the RCM full automatic mode.
The node based approach gives you flexibility when cameras or footage is not properly supported by the built in transforms. As well it gives flexibility on using output luts. This is typically how we utilize the resolve tools as it provides the consistent built in help with color management, but can be customized if something weird comes along.
Great info! Thanks Joseph.
Thanks Joseph!
As you said it takes knowledge to imitate the RCM with CST. Up until watching this video I have not succeeded achieving exactly the same results. Great video! At least now I know it was user error
We enjoy your videos. Thanks for covering this topic, it helps to clarify a point that a lot of people get confused about. Thanks Cullen.
Thank you!
Thanks for going over this as well as showing how you group your clips in Resolve. You just saved me a lot of effort with multi-cam grading
Suspected this to be the case but always good to see it in action. The group/timeline tuck trick for the bread of the sandwich.....awesome technique. Huge help. Thank you.
That was well explained, and now I understand that both approaches produce identical results. You, the man. Thanks for your skillset and knowledge
Just a note: when using Davinci YRGB, in the project settings, the "timeline color space" setting will affect the CST nodes if you choose the "Use timeline" option for gamut/gamma for the node. I would never do this but it's helpful to know that they're linked that way.
I was just wondering about this. For output gamut/gamma couldn’t you just put “Use timeline” in the CST in node? As he already selected the same setting as timeline gamut/gamma in the project settings.
@InLightVFX: Can you please elaborate? So are you saying that we should never leave it as “timeline color space” ? I thought he said it didn’t matter what you set that to?? I’m a bit confused now. Thx in advance.
@@JSalas-fh9xb The color space that you choose here at 3:02 for "timeline color space" (in this case set to Davinci WG/Intermediate) is the color space that will be used in the CST node when "use timeline" is selected (which is set by default as you can see at 3:14). If you're still confused, just basically never select "use timeline" in the CST node - and then it doesn't matter what you put for the overall "timeline color space".
Great point and great discussion here. My preference is to always avoid 'Use timeline' and take the extra few seconds to explicitly state my intention for the CST...just a personal best practice
@@CullenKelly Yes, sounds like a good practice to keep things clear!
Showing these basics again, I clearly understand now how I can re-boot my workflow using CSTs or RCM. Thanks a lot Cullen!
Thank you Cullen - Not new news. BUT always useful to hear you working methodology and the reasons behind your working concepts. Your group application of CSTs confirmed my own methods for multi-cam grading.
Thanks to your videos, I'm familiar with it, but I'll teach beginners RCM. I taught them node-based before and some got confused. RCM may be a set-it-and-forget-it way for beginners to get good results.
Oh wow - I can't thank you enough for this explanation - I was assuming they did something similar but wasn't 100% sure and I was also trying to work out the logic as to why one would go with a CST in the node graph, when color management can be setup to do the same thing under Project Settings. Now I can see the logic behind a comprehensive node-based workflow. Brilliant job!
Thanks for the refresh and confirmation they’re both the same. A related aspect I’m not as clear on when using Group CST is where to place a LUT for a particular look. Especially when the LUT is made for Rec709 footage.
If the LUT is made for 709, you’d place it after the CST that outputs to Rec 709.
For example, my first node is always a CST that converts my log footage to DWG (Davinci Wide Gamut and Davinci Intermediate) then I have 6-7 nodes after for color grading, and then finally my last CST that’s converts it from DWG to Rec 709.
So you’d add another node after the CST converting DWG to 709 and add the LUT on.
@@nowhereland_2 thank you. That makes sense, and that’s what I’ve done at the clip level. I wasn’t sure if this would be same at the Group level as an Out CST as well.
@@nowhereland_2very helpful any help where it may be placed if the lut was designed for say slog3?
@@inbloom9083 If the LUT is made to convert Slog to Rec709, you’d place it right at the end of the node tree, so that all your grading is being done in the wider s-log color space and you’re able to see how it changes the final Rec709 image (or whatever the output color space is for the project).
For example, a simple node tree could be:
Exposure > WB > Contrast > Slog to Rec709
The last node is where you’d right click on the node and add the Slog LUT.
@@nowhereland_2 that’s brilliant thanks so much for letting me know.
This is new to me, I'm super grateful for taking your time to explain this
Hi Cullen, you're one of the few colorists who outputs to Gamma 2.2. If that is so, do you also calibrate your monitor to Gamma 2.2?
Yes
Yep, I have a 2.2 preset on my monitor that I flip to when recording YT videos. For client-facing work, I still generally master in 2.4.
Very helpful clarification. Newbie learning to use Resolve, but finding it incredibly useful compared to using adjustment layers in Premiere. Huge thanks to you and Darren Mostyn for making picking all of this up very easy for a beginner.
As a noob on da vinci resolve I've been learning it quite intensely over the past 6 weeks and refreshers like these are gems when found. I think little updates on the under the hood parts of grading are important for the newcomers and regulars.
Thanks so much as a newbie to Resolve this has been super helpful, not sure which way Ill be working going forward but definitely want to try both
Thanks so much. Definitely helped me with the confusion regarding the colorgrading. cheers
Great demo to show the differences. Personally like using node based management but love having both options available.
I always use the node-based workflow because I often work with graphics and lower 3rds, and the workarounds to bypass RCM or reverse the color management applied to them is more of a tedious workaround than it needs to be, especially with transitions and effects. With transitions, I end up having to use nodes anyways because I have to bypass both clips on either side of the transition because there's no way to color manage the transition itself.
try using an adjustment layer over the transition and media layers to affect them both!
Great video! thanks for sharing. I have some questions to ask you.
What happens if the timeline has graphic files or different cameras, wouldn't it be better to handle it at the clip level? And secondly, can you make a CST group (as you did) and put together other groups by scenes? can a clip be in two groups? thank you so much!
A little more detail to finish. Round trip past the RD3 clip to prove the point on all clips.
Also, perhaps compare with using camera manufacturer LUTs for the previously poorly informed - who might mistakenly double up in colour management. In another forum that I lurk in, colour management double up problem is surprisingly common. And an easy trap to fall into.
Cheers. Was looking forward to your next drop of wisdom.
i think the point has been proved.
Great idea to discuss inadvertent "double" color management -- you're right that this happens a lot
Thanks for clearing this up, I was going crazy trying to figure it out myself. But now I'm curious as to what the difference is between the ACES Transform node and the Color Space Transform node.
Why did you choose Gamma 2.2 for the output color space instead of Gamma 2.4?
I use 2.2 when creating YT content, as that's the spec for most of the displays it will be viewed on (phones, laptops etc)
@@CullenKelly so your monitor gamma is still 2.4？
This is great, thank you for explaining that so clearly and succinctly
How does this work if you have images and motion graphics in your timeline? Won't the input colour space effect the way those things look too?
Great question! Yep, in this scenario you'd want to place your output transform somewhere else. Here's a video where I break down several options: ua-cam.com/video/ReHA9bhT168/v-deo.htmlsi=tG_3XlN6_fIl_iAU. Hope this helps!
Cullen, why do you use Gamma 2.2 now when you used to use Gamma 2.4 before?
It's just a preference of him. In other videos he used to use 2.2 too for devices like iphones or ipads because these have a gama of 2.2
Yep, Brandon nailed it. Main reason is that the images I'm grading for YT demos are primarily going to be viewed on gamma ~2.2 displays, rather than gamma 2.4. For client-facing work I still primarily use gamma 2.4.
Wow man, thanks for showing how to use grouping to add the CST nodes to multiple clips without having to do it for all clips individually, huge timesaver for real!! You earned a new sub! Thanks for sharing.
Welcome to the channel David!
I started using the 'group' function recently and have found it to be helpful and efficient. I am also a fan of node-based color grading, I tried the Resolve color management system for a while but the node-based structure works better for me. Thank you for posting this video.
You got it!
Wonderful information Kelly. I am beginning to use DaVinci Resolve and learning some important stuff is really helpful especially to color grade multiple clips at the same time. I initially thought you would like DaVinci Color Management better as you have an array of DaVinci Keyboards and other color management tools. Thanks for sharing.
Will the use of "out" node at the timeline level affect graphics, additional texts, animations, etc.?
Very helpful Mr. Kelley. I’ve always wondered if there was a difference. Thank you
"Thank you so much. I have learned a lot from you. I can't say thank you enough. It was a good refresher for me."😊
You got it!
Dude, this is the video I've been wondering about for a long time. Thank you!
Sure thing!
this is new to me did not know they were both the same,,, good to know that i will have to try both ways to see what i like the best...
but i do have one question what is the different between Rec.709 Gamma 2.2 and 2.4
Funny coincidence: Yesterday I was discussing with some other filmmakers which is better. So thank you for the short and very clear clarification! I prefer color management because of its simplicity (fewer knots, fewer sources of error when you are short on time). In normal cases (shooting in log, fairly simple color grading) I don't see any advantage in using CSTs.
Glad this was timely!
@06:23 is there ever any reason to make adjustments (for a individual clip) before the CST? Noise reduction? Levels?
Generally no -- my take is that if there's something in your working space or CST from camera to working space that limits you, something is wrong with the choice of working space or the CST.
@@CullenKelly Thank you. Is the same true if a CST is not available and I need to use a transform LUT?
Could you explain the reasoning behind having the output as 709 gamma 2.2 instead of 2.4? Is your monitor setup for 2.2? I understand that internet delivery likes 2.2 over 2.4. I usually do a timeline CST of 2.4>2.2...
Sure! Gamma is set to 2.2 for YT content because it matches the gamma of most of the web devices the videos will be watched on. I have a 2.2 preset on my reference monitor that I flip to when recording demos, but for client-facing work I'm still generally mastering in 2.4.
what about applying a lut?
This is head clearing and useful! However, the question is, how come you can set the input gamma using CSTs, but in the settings you only have the option to choose the input colour space?
I have been using the RCM color management you taught me, but I can't restore the color when I enter FUSION! What about teacher?
Hey !! i really really love your video. would you please make a video on how to analyze reference for grade like exposure, contrast, saturation, Balance, skin tone, look and whatever its.
I’ve been using the group CST method BUT also using YRGB Color Managed - is this a problem?
I'm always messing with some footage which is not on the list at "In Put Color Space"...for example Samsung Mobile footage and Dji drone shot in D Log-M mode, Those are not quite perfect when I put that footage in Rec709 input space
Yep, color management definitely gets a bit tougher when your footage has an unsupported or "709-ish" color space
This is very helpful, thanks for making videos for people like me who are still at the very beginning of the journey.
You got it Miguel!
What do you set your General settings to. I see you're using a Mac, but with a non-Apple display. Does this mean you're selecting the "Use Mac display color profiles for viewers" option? Would you or would you not be selecting "Automatically tag Rec709 Scene clips as Rec709-A"? Please explain, thank you.
Hello folks! How do you use the group post clip level of your node graph vs the timeline level? Do you use the timeline level only for the output transform or what else do you use it for? Thankful for answers.
I wonder why scopes then show slight differences though when you wipe between the two images - is it a scope rendering issue or there is still some minute discrepancy between the two?
Great catch! Difference is in the Gamut Mapping settings I'm using in my CST vs the ones used by Resolve Color Management. Good example of the extra control gained when color managing with CSTs
I’m finding myself using cst more than project. Even though for ARRI raw I like the project based. I’m refreshing my Reel and playing with my original footage and the CST and group ling has been very practical
and the color wheels behave the same way? regardless of color managed or cst?
nice content !! I need some help with a big render problem. How do I fix the color wash out issue on output render. I use a Windows 11/ Go Pro 11 Footage (Natural_10Bit_4K 50 Frames). I have already applied Preferences as: Use Windows Color Management... and Color Management: Davinci YRGB Color managed. Rec-709 A / Davinci Wide Intermediate/ SDR100 / Rec 709-Gamma 2.2. Could you please tell me where I am going wrong that my output footage turns washed out? thanks in advance...
Last step is new for me. thank you
You got it!
Awesome explanation
Cullen.
When using auto color management, how would you handle an edge case that required correction "pre-transform?" For example i sometimes will need to adjust exposure before the input transform to avoid a clipped image in the dwg working space.
I know you can get identical results, but I am curious about which settings need to be tweaked so that the waveforms don't move.
You can see as you toggle on and off at 4:10, in the green and blue channel, a couple of small shifts. While probably not visually noticeable, I am curious what causes the different waveforms.
Definitely agree that the results can be identical and they are doing the same operations in the same way, but it seems that there is a tiny settings difference in one of them. If i ignore the image, i can tell when you are switching between by watching the waveform change.
Love your eye for detail! You're actually just seeing the discrepancy between the saved still and the exact same image being live in the timeline. Try grading a shot, saving a still, then wiping to it without changing anything...you'll notice the same slight shifts.
Thanks for clarifying!
It would have been very interesting to see how this works and looks with Automatic color management. Would it also be the same? Is there anything that speaks against Automatic color management if you find the other ways too confusing and don't need the same control over the process but still want something solid? Can you achieve that with Automactic color management and a simplified workflow?
Thanks for the great explanation. One detail, when you set Out DRT to Luminance Mapping, why is that? Because it has a hue shift compare to Davinci (the default option) So I wonder why that decision. Thank you.
Would you be able to comment on how to use node based color management but still benefit from the new CSTs in the Fusion page? Thx Cullen.
I'll have to read up on this!
I have a question about color management when working under a LUT that isn't DWG>DWG! My cinematographer is insisting I work under a LUT they used in-camera during production. This LUT takes S-Log to Rec.709. Where would I apply CSTs in this situation if I wanted to work in DWG? SLOG>DWG>SLOG>LUT?
Great question! And yep, SLOG->DWG->SLOG->LUT would work well!
@@CullenKelly thanks for your insight!
Thanks for all
BTW I knew all you talked about because of your great previous videos.
Can you do a video how you might use the new grouping 4 dots that are available??
I watched a video two days ago about this, so it's not new (anymore) for me, but it was good to get a second voice about this topic. :) Just one question: Could you tell me, why you are using Gamma 2.2? I've learned from Darren Mostyn, that he uses Gamma 2.4 and many other UA-camrs use this. What's the difference?
Hello! Maybe you know way how can I use this "sandwich" on a MacBook? Because I am Mac user and I have to select Rec.709-A in the settings... If I don't do this, preview in Davinci will be different from the render in QuickTime...
One thing I have found different between the two approaches is related to qualifying. When doing so using RCM, I have no problem since it seems to qualify based on the Rec709 image (despite still working within DWG), however when using CST's, it's more difficult since it's using the, I'll say, washed out version from the DWG colorspace. I notice this when highlighting.
I prefer using CST's because it just makes it easier for me when using effects or LUTs that I prefer using in Rec709
Great points all around!
@CullenKelly Not sure if you’re aware, but the only link that is working is the RED raw film link, all others give an error. I noticed it initially in my attempt to download your E-Book…
You don't need an out converter, just select the timeline in WG/intermediate color space and gamma, then set up the right CST for the clip so it appears correctly on your screen. You can start grading. The Out node mathematically is not necessary. You just select the desired color space TAG on the Delivery page.
Hmm, not sure I follow here, but I can confirm there’s nothing redundant or unnecessary in the steps I’m showing here!
You're simply the best teacher
Thanks for the love!
Thanks for the video!
Only question is - why not use the "CST Out" node in the "Group Post-Clip" tab vs. the "Timeline" tab as shown in the video? Since using the "CST Out" on the entire timeline would then affect the whole project vs. affecting only the assigned group clips we need, right?
You got it! And great question: reason is because I actually want the 'Out' CST to affect everything, not just the one group. Everything will be mapped into DWG from its original camera color space, and from there everything will receive that same transformation to get out to display space. Hope this helps!
I've seen all colourist doing this. or a clip based CST. But for most of us who have to do everything from editing to color, graphics and subtitles working in groups is way more efficient.
@@daniellrodriguezs Great point! I actually did a full video looking at different places you can put your output transform depending on your needs/preference: ua-cam.com/video/wma5NvhbNUk/v-deo.html
@@CullenKelly found it right after I left this 😂 thanks for leaving the link though! I like the shared nodes as well for things that cannot be done in a post group clip like for example Luma mates. I found some cool effects with it but unfortunately cannot be applied to pre or post clips adjustments so I found shared nodes works just fine.
At the 2 minute mark, after you set up RCM with input color space ARRIlogc3, what is the reason that you had to again specify the ARRIlogC3 input color space for clips 2 through 6?
Great question! This is actually another con of color managing in project settings...Resolve will fairly frequently make its own guess about the source color space of your shot, and it's usually wrong. That's what happened with shots 2-6, so I had to correct that wrong guess. Hope this helps!
I’m grading Apple log footage following the steps of the video to grade it in DWG before converting it to Rec 709 Gamma 2.2, and I made sure to specify the correct tags in the color management settings.
However, the image still looks flat on my iPhone and MacBook, except if I use 400 for Max Output (nits) instead of 100. Is there a specific reason why 400 works better than 100 in my specific use case, and if so why?
Hmm, I haven't seen that issue so not sure what's going on there...
Hello Cullen, I believe (unless I missed something), that is not true anymore that in Davinci YRGB, the timeline colorspace and output colorspace are just tags and never influence the grading. Actually, the HDR wheels are colorspace sensitive even in Davinci YRGB, so the timeline colospace has an impact in DVYRGB for that tool.
You're 100% right that those tags change behavior of the HDR wheels...I was just pointing out that they don't change your color management.
I know it can be a pretty dumb question, but is killing me and I cant find any clear answers on it... in a CST workflow, why setting up the timeline color space as DaVinci Wide Gamut in the project setting? like, if I'm controlling in what space the operations are happening via the CSTs in my node tree, what exactly changes when I change that setting just in the project settings? I know there's something changing but cant point out what it is and cant decide what's the appropriate setting.
I assumed a long time ago that it affected nothing so I just leave both options at my output space (Rec709 gamma 2.4/2.2), making all the required transforms via CST, all went smoothly, all my tools behaved in a way that worked and I was satisfied with the results, but one day I change the timeline space to DWG in the project settings and suddenly various tools changed the way they operate (halation, color compressor, color wheels), and I really don't like it, specially when qualifying or using HSL curves. The thing is that now I'm fearing that maybe there's information being clipped or lost if I leave the timeline space in Rec709 😰
Hi Cullen - I like to set my timeline luminance level to 10000 nits using RCM but notice if I'm using graphics or Rec709 material the highlights get dropped way down to almost halfway on the waveform dulling the whites especially in graphics. My workaround was to add a CST node to the clip and set the custom max input to 100. Please tell me if this is correct or if there's a better way - thank you!
What if you do CST, the sandwich, in the entire timeline? I have experienced that the results are different if we use it in the timeline or clip by clip
You could do this, but then you'd effectively be grading in your camera space rather than your working space, which essentially removes the benefit of using the working space at all. Hope that helps!
I love love love the new setup ;)
I have graded my film with outputs Rec.709 Gamma 2.4 for Amazon, Tubi etc. However, I am having to do a version for theatres, and they want P3-D65 Gamme 2.6. What is the proper way to take my Rec.709 Gamma 2.4 grade to P3-D65 Gamme 2.6? Would right-clicking and creating versions work and just switch the output gamma and color space when I do the P3-D65 Gamme 2.6 versions?
interesting that you put your out cst in timeline and not post clip, wouldn't this effect anything you add above, text, etc?
Yep it would! In my case I don't often have text or other non-camera elements in my timeline, but if I did I might choose to place my output CST elsewhere. Actually did a full video on this topic if you're interested: ua-cam.com/video/ReHA9bhT168/v-deo.htmlsi=AgvTC_FF4tHKHlV2
Hai Cullen, appreciate all your videos. But maybe this question little bit noob question. What happen if the CST In and Out placed on Timeline not group? is it the same method? Are the CST In always in Group level, or we can do in Timeline level? Thanks Cullen.
Hey Rudolf! You can totally place your 'Out' CST in Group Post-Clip instead of at the Timeline level...only catch is that you need to manually place it in that spot for each of your groups. Here's a video where I do a full breakdown of the different places you can put your output/macro-level transform: ua-cam.com/video/ReHA9bhT168/v-deo.htmlsi=mWBtZypcPEupPQ-C
Hope this help!
@@CullenKelly thank you so much Cullen. I trully thankfull with all your content. I was start color grading from 0 and learn a lot from your channel. Im not profesional colorist (yet) but i'm learning so much from you. Thank you.
Hello, thank you for this video ! Why don't you put tone mapping in the "in" corrector ? Is it better to apply the Pro Mist diffusion filter effect with the diffused glow after the "out" corrector ?
Sure thing! I don't tone map at the input stage because I don't to grade an image that's already had its highlights compressed...allows me more control and speed in the grade. Regarding diffusion effects, best to place these to the left of your output transform for a number of reasons, but the most simple one is that you'll get better-looking results. Hope this helps!
Hi! i'm a nooby on this, and i have a question that nobody talks about, probably couse is for stills... I'm using a Lumix S5 ii X, and y tried to make luts for "real time lut" while i was taking photos. But the lut doesnt look equal on the Lumix than in Davinci, working with sRGB>DWG>sRGB. First, i convert the RAW2 to DNG, and then work on Davinci with that Color Maanagement
This one's too complex to fully address in a comment, but I'd start with 1) confirming that your image is in sRGB when applying the LUT in-camera and 2) ensuring you're decoding the DNG into sRGB in Project Settings -> Camera RAW. You may also want to look at capturing in a log space rather than sRGB.
Does it affect how the curves or the gamma or the HDR knobs affect the picture? or is it the same values both ways. I get all confused on this color stuff.
Sorry, not sure I follow your question?
@@CullenKelly oh nevermind i think im just making things complicated for myself, i think its just the same.
if using the RCM instead of using CSTs, where can I know what color space I am grading? For instance: If there are two luts: 2383DWG to DWG, and 2383 Rec709, what's the correct position(node) I should place?
Sorry, not sure I follow your question!
@@CullenKelly If I did the Color space&transform on Project setting ( instead of using the node tree to transform the color space), how did I know the picky eater (the LUT) was placed in a correct position (color space) ; i.e. the color processing mode is set up as DWG Intermediate, Out put color space is Rec.709 gamma 2.4, the Lut should apply in the DWG color space ( or 709 color space whatever), where should I put? what's the color space of Group Pre Clip level , Group Post Clip Level, and Timeline Level?
ok so im back with a question ive seen that cst are more destructive then input any say to that
Sorry, I don't follow your question!
@ I was watching another colorist saying that cst are more “destructive” that lowering the saturation and finding the contrast point with a managed timeline said there’s more control im not sure if that was bs lol so just asking around
I always thought CST is better than Color Management, it all makes sense now. Tell me, does using the CST route slow down the performance of the computer or laptop? My lecture said by color managing with CSTs, it'd make my color grading performance very slow in terms of performance. True or False?
Hm, I've never experienced any difference in performance between color managing in project settings vs nodes, and I can't think of a reason why this would be the case...
Good refresher!
Cullen, thanks for this
So I’m still confused lol I thought it was gamma 2.4 not 2.2 I usually color grade my own stuff and I use s-log3.cine I was wondering if I now need to convert it to 2.2
Depends on the display you're mastering with and where you're delivering. For UA-cam I set my reference monitor to 2.2 since that's generally how the content will be viewed. For other scenarios I will set monitor to 2.4 and deliver 2.4
@@CullenKelly thank you so much.
Thank you !
why did you always set your output colorspace to Rec.709 Gamma 2.2? and not 2.4?
I use 2.2 for my YT videos because that's the gamma of most computer and phone screens.
@@CullenKelly I did not know that
Cullen, just wondering why you placed the output CST for the Log-C group in the timeline instead of the one for end of the group? With different media types that are not using Log-C would that not adversely affect them?
Great question! The cool thing about the output CST when using a working space like DWG is that it works for all cameras, not just LogC3, since everything will be mapped from camera -> DWG on the front end.
Hi Cullen. Thanks again for all your input, you taught me everything I know today. Your photometric approach works well for me as a retoucher and digital tech (stills).
One question. I find that my camera groups disappears when creating a new project. Any way of exporting or baking them into my settings? Appreciate. Charl 🥊🍺
Great question! You can maintain your camera groups by creating a template project that you save a copy of whenever you're starting a new project. To include groups, import a few clips into that template project, set up your groups, then delete the clips so that you're starting from a blank slate.
@@CullenKelly Awesome thanks, will set this up as a template, appreciate !
Thanks Cullen. This was indeed a useful review. I would however like to know more about your RCM custom settings, e.g. why is something your "preference". And how are your custom settings replicated in the CST version to produce the same result?
One big advantage when using Resolve Color Management is that you have the correct colors also in Fusion
Awsome video, thanks for the explanation. I was testing node base vs project based color management, but when applying the Kodak 2383 LUT from Davinci resolve, I've got different results. Seems that the project based has the skin tones more neutral also the colors is cooler than the node based color management. Do you think is something I've done wrong or this is normal?
Thanks in advanced.
What pen and tablet thing are using for your computer?
I'm thinking if every shot in your timeline is of the same camera and format, RCM may be the better choice? ..as CSTs would be the better choice if there were multiple camera formats etc on the timeline? ..saves time .. 🤔
I like this logic!
Could any of you DaVinciResolve experts (interested in solving difficult challenges) take about 5minutes to help me with the following problem that I am currently being confronted with? (Thank you so much in advance! ❤)
In the Shorts video „How do I Get JVKE to see this?“ by „The_Magic_Piano“, you can see MIDI animations/ notes dropping down from top to bottom ending on the upper edge of the piano. I desperately want to recreate this idea for my own projects.
I tried doing so in DaVinciResolve. After having imported the piano performance video and the MIDI animation video and dragging them both into the timeline I tried editing the MIDI animations by changing x and y positions, making it less wide, stretching it lengthwise, and changing pitch, yaw and rotation settings.
No matter what I try, the MIDI Animation video always looks either distorted, the sizes of the animations are different from left to right or from top to bottom or the MIDI animations are Not aligned with the upper edge of the piano keys.
Does anyone know how to solve this Problem? Maybe with a different method within DVresolve or a different software?
The animations in his Shorts-video Are all uniform in size and width and match perfectly with the upper edge of his piano keys. Also the Level of the animations seems like it‘s accurately parallel to the wall. Lastly, the shadows on the wall in his video are nice but i guess that would be less of a challenge if the main problem is solved.
I‘m so desperate for help and I don‘t know who to ask.
Even ChatGPTs suggestions aren‘t successfull for me 😂😢
Thank you again and have a Great Rest of your day!❤
Perfect explanation