@@RustyShackleford9000 Thanks for watching and commenting. I can understand why you'd think that, but green actually is a primary colour. Primary colours 'are sets of colors that can be combined to make a useful range of colors'. Different colour systems use different primary colours. For example in print Cyan, Magenta and Yellow are the primary colours used. In additive colour Red, Green and Blue are used as the primary colours. You might find this article helpful: simple.wikipedia.org/wiki/Primary_color Thanks again for watching and commenting.
THIS is how Color Grading Tutorials should be done! Thorough explanation of how Things Work and why you do what and compare different methods and explain the effects of the techniques on the footage. Great channel
Hello, a big thanks to everyone who has pointed out that subtractive saturation can also be created with a single node set to hsv and then changing the gain of just the green (saturation) channel. We're calling that method hSv. It's a great tip and our apologies for not including that in the video. It doesn't change the fact that HueShift is a great tool that adds value to the process and speeds up your work! Thanks once again to everyone who has watched and commented. We appreciate you all!
I'm two films in with Kaur, Jason, and Stefan Ringelschwandtner's DCTLs and I'm never going back to the "old ways" that involved complex node trees. The bang for your buck is a no-brainer.
That's great to hear. The convenience they bring to your workflow is well worth it.
Рік тому+5
One of the greatest tutorial channels for DaVinci! Remarkable job! One little correction i must mention because I am shooting on film a lot. 3:26, The density on a film is not coming from the saturation or the color richness on a negative film. The density is coming from the exposure. More overexposed the negative, the dense (thick) it will be. Underexposed means less dense (thin). Again. GREAT GREAT job guys. thanks for this wonderful video!
Рік тому
Correction. It is somewhat relative with the color richness also, but mainly the density is exposure.
Heya. Thanks so much for watching the video and leaving such a kind comment. Some thoughts. Yes, density is a consequence of exposure, but exposure is also a consequence of saturation. Highly saturated colours are also typically highly exposed. Hence, the reason why density is being associated with saturation. Saturation occurs when there is a difference in the exposure of the colour records. The greater the disparity the higher the saturation. So in the instance of a highly saturated blue... the saturation is caused by a dense magenta and cyan record. Would love to hear your thoughts.
@@team2films "Highly saturated colours are also typically highly exposed". If you think about light reflectance, it should be other way round. Objects with highly saturated colors reflect less light, because they absorb more wavelengths that they reflect in comparison with pale and white colors. So, they should be less exposed on film.
@@Fedor_Dokuchaev_Color thanks for commenting and joining this thread. Yes, you are right that saturation does technically decrease exposure, many highly perceptually saturated colours in an image have above average exposure. Please allow me to explain what I mean. It all depends on your frame of reference. Yes, a highly saturated colour has less exposure than pure white therefore it has a lower exposure. But in the example provided above of a strong blue it has strongly exposed magenta and cyan records. That’s 2 out of three color records that are near fully exposed. As the most strongly saturated colours have the highest disparity between colour records, that’s why density (or exposure) is so strongly associated with saturation on film. I look forward to hearing your comments.
Рік тому
@@team2films Thanks for your answer. True, but on the other hand, if you would take a picture of a super saturated blue sky and absolutely underexpose the frame, no matter how much saturated would be the sky, the negative would be super thin :D cheers.
@@team2films yes always nice to see the faces behind these things and understand where they're coming from. I've seen a few vids on this area of saturation but this was by far the best. Thanks again, love your channel.
Really interesting - I know both Jason and Kaur - really helpful and smart guys. I proposed the question to Cullen of why we tend to not fix the saturation in footage first, then decide on the amount to apply? My opinion is pretty much what Jason and Kaurs tool actually does. I don't believe that saturation is something that can be a one knob fix in regular sat tools using any color space. I also have the opinion that camera manufacturers, and pretty much every job on set is to nail skin tones - especially in camera, the digital saturation and how it is used is to push the best ( what they think is the best ) skin tone in their saturation decisions. Camera manufactures know this, and will target skin as a selling point. Now because of that fact - a lot of the rest of the image may, or may not have the correct balance in saturation, and become either too saturated or undersaturated. Where as I mentioned - one knob isn't going to fix it. The filmic aspect - the fact that light is being pushed through the film - you want something redder or greener - a film camera can't just add saturation - they need to increase the amount of color in the process, by adding density ( literally ) which of course will let less light through and it becomes darker. Keep up these great videos - you are going to get a large following really quickl and the sucess will be well deserved. cheers
Thanks for watching Jim and for your encouraging comment. Jason and Kaur are great. Yeah, it's great to have such a variety of cool tools for addressing saturation. We are spoilt! And Kudos once again to Cullen for his work on this. Thanks again for watching. Great to have you here.
I saw some guys use the second method, but instead of using nodes to separate the V-Channel they disabled these channels in the nodes drop-down menu. I'm not entirely sure if this approach has any differences, but it is way easier as you only need one node and it gives you subtractive saturation and similar results.
@@team2films Yeah in a single node set colorspace to HSV with channels 1&3 disabled leaving only channel 2 enabled to control saturation via the gain wheel. Scopes behave the same way as method two so looks to be subtractive.
In fact you can see it in this video! When you adjust the green gain channel in your HSV node, the scopes look identical to when you're adjusting the gain at the end@@team2films
Invaluable information, that's so cool! Especially compositing of color nodes and layers explanation. Huge thanks for sharing knowledge! The irony is that one video interview about advanced color techniques in 4 bit color with banding and like shoot on a potato... 🤣
Absolutely amazing advanced tutorial here! I find it very difficult to find expert knowledge in this field because the rise of influencers lately.. I am curious if you have a paid course in this format? Also I would love some help in the realm of sharpening! There are many ways to tackle it but most leave me feeling with gopro like footage. Thanks again!
Hello! Yes, we have a course! It focuses on professional editing in DaVinci Resolve but it also tackles basic colour and audio work. There will be more details about that next week. We love requests and suggestions, thank-you.
Ah thank-you Eric. That's really kind of you to say. When you are ready, it's there for you! In the meantime thanks for watching. It's great to have you here.
I think for someone like me who personally don’t like film grain or halation when emulating a film look, sub sat is the best way I can make my modern digital video look filmic combine with proper contrast curve that never hit true black at the bottom and true white at the top of the curve. I personally will not 100% staying away from Sat vs Lum in HSL it’s sometimes can work as complimentary knob for HSV color space achieving that deep rich filmic look. You don’t want to over do it tho.
@@team2films One would expect though that a world class software like davinci would already have such a plug-in!(Simplyfied) let’s hope to an uptade soon!
@@blertonsaraciYeah agreed, it would be great if these features were integrated in the future. In the meantime though, It’s pretty amazing though that DaVinci’s node based engine makes it possible to perform these kind of operations. That’s also why Blackmagic has developed the DCTL api. So other developers can augment Resolve with additional functionality. It would perhaps make the program more difficult to use if it tried to incorporate tools for every style of colour grading.
@@team2films Absolutely, you are right! Just that with the time,(as you briliant guest mantioned) you start feeling/understandig the foundamentality of density, and it’s definitive belonging in the DaVinci world!
So I've played around with this and found out that setting a node to HSV and just raising the green channel yields the exact same result. When set to luminosity on a single hsv node it's identical to the presented node tree when just raising y gain. What I experimented with is splitting HSV and feeding Saturation into the Value channel and subtracting it that way. Beforehand I multiply the Value channel with the Saturation channel to avoid ultra saturated colors at the very extremes of the luminance ranges. The result is - in my opinion - the most flexible and the most pleasing looking. The corresponding node tree can easily be saved as a power grade and just needs one or two nodes to be touched to get nice results. It also has the benefit of lowering the gain of saturated colors, without adding saturation you don't want. Just my 2 cents. Cheers!
Thanks so much for watching! Yes we’re aware of that technique. There is a comment in the description about it. Doesn’t mention the multiply technique though. You can also use a curve instead of gain. That also gives you a roll off. Subtractive saturation is also now implemented natively in resolve with the new color slice tool, and the film look creator.
The interesting part is, that I use a curve to bias how the luminance gets affected, the sky is the limit here. The gain then is just the to dial in an amount of said curve.
Hi, wanted to point out an error when you were explaining about the "fake" methods. The second "fake" adjustment was hue vs luminance, but when you were explaining it you said "hue vs saturation globally desaturates all blue, regardless of whether it is saturated or not". Would like to clarify if what you meant was that hue vs luminance globally darkens all blue, regardless of whether it is saturated or not" ? Also, you pointed out that sat vs luma desaturated unwanted parts of the image. Shouldn't it be the case that is darkens unwanted parts of the image, rather than desaturating? Since sat vs luma targets the luminance across a range of saturation, it does not objectively alter the saturation. Pls correct me if i'm wrong but i needed to clarify this source of confusion. Love the well elaborated video btw!!🙏🏻
Having watched your great tutorial, and played with it a bit, and looking a lot ... I'm now frustrated that I can't work out how to replicate this in Photoshop! 🤣
Hahah! That's so funny! I'm sure there must be a way of doing it! For sure though there are times that I wish Lightroom or Photoshop worked more like Resolve. Would love to have tools like lift gamma and gain controls or Vs Curves in photoshop. Thanks for watching.
@@SeanHuangRealEstate Yes of course!!! When you change the node to HSV color space, you can also right-click the node and select the option "Channels" right below... this will open a window with 3 channels already checked, being channel 1, 2 and 3 respectively for H, S and V. Then you can simply uncheck channels 1 (for Hue) and 3 (for value)... now you have only Saturation channel working. Simply control the amount of saturation with your "gain" wheel on log wheels. (Gain wheel gives the best results, I tried using other wheels and they also work but breaks the image considerably. (hope it helps my friend)
@@arthurangenendt8959 Thank you very much for the detailed explanation! It looks like using this method we can change the overall subtractive saturation. If I want to change the saturation of a certain color, then I will need to use a qualifier and then repeat the above?
Excellent tutorial, thank you! I have just one question - is it not easier to switch the node to HSV, turn off channels 1 & 3, and apply Gain/Gamma to the same node instead of creating the layer mixers?
Thank you very much for the video. I have one question. How does the “hard way” 5:35 differ from switching the node to HSV color space and then manipulating the S channel, i.e., the Green wheel and/or curve? I ask because I learned this very same technique from Cullen Kelly’s TouTube channel and it does appear to yield better results than additive saturation. Thank you again and congratulations on your excelent videos.
Pleasure, thanks so much for watching and commenting. That's a great question! Please post a link to the @CullenKelly video. I'd love to take a look. He's a fantastic source of information on colour grading. It sounds like you a describing a method that still produces additive saturation but in a way that is more pleasing than the stock saturation tool in Resolve.
@@team2films Thank you so much for answering my question! Here's the Cullen Kelly video I was referring to: ua-cam.com/video/bAAjyNHDpyo/v-deo.html I find the work you've been doing on your channel complements quite well what I learn from Cullen Kelly's videos, so I would love to hear back from you on this technique. Thank you again.
@@dlaimer Hello, thanks for your patience. We did some testing, and yes the single hSv method yields the same result as the complex node tree. Our apologies for not having included it in the video. If we ever make a follow up on subtractive saturation, we'll include that.
That's a great question. It's hard to say. Blackmagic are always innovating and adding new features to the Resolve but DCTLs will always have a place though.
They did just add a feature a day ago that people were asking for. I have a feeling we're going to see something like this for Studio users. I prefer to learn the hard way because knowing how coloring works gives one better control for unforeseen problems. The images he's working with were properly shot with correct lighting and white balance. Working with terrible footage may be more challenging even with the DCTLs plugins he's promoting.
Perhaps it will provide a little more clarity when considering the Cullen Kelly 1 hSv node solution, considering that he often balances the saturation on this node with gamma AND gain.
The hSV method is great. It's capable of producing the same result as the more complex node tree we created. Sorry we didn't include it in the video too!
@@team2films No reason for an apology. I just wanted to respond to some comments that don't see the difference between your method and Cullen's method. Cullen then balances the saturation manually, while your node tree delivers a slightly better result out of the box.
@@jensgeumann7564 Ah thanks. That's kind of you! We did some more testing and it's possible to replicate the complex node trees results with the hSv method, but also at the same time there are some unique things you can do with either method too. We're horrified by the prospect of sharing incorrect information 🤣 So we always appreciate people pointing out errors and helping us to give a more rounded impression of a given topic.
Sweet Jesus who sleeps with Mary in heaven!!! Tried this on a recent grade and it's like next level film-like! I always knew the saturation knob had something "off" about it. Always increased saturation with great misgivings
I have been trying both this "Hard way" method and the Bleach Bypass tutorial. I can't seem to get either to work. I have gone over the above video dozens of times, ensuring that I have each node and each adjustment/setting on each node correct. I have tried putting this node tree before the CST, LUT, Contrast, CST and 'Cineon Film Look' Nodes that I use as a default on all my Sigma FP footage. I've tried putting it after the FP default node tree, I've tried it between the LUT and Contrast Nodes. Which is where I believe it should best reside. But still I get nothing. Absolutely nothing happens when I adjust the Gain or either gain colour channel in the last Node before the "Add" mixer. Where am I going wrong? I've even tried starting a new project, with a blank slate, absolutely no nodes in the colour tab, and still I get nothing when using the above described "Hard Way" node tree. I'm befuddled.
I am total new to DVR, got the Studio version though. I got one question: What is the difference between the native DCTL tool and the HUEShift tool? Sorry if this may appear a stupid question, as mentioned I am new to these kind of things. Thank you already.
Great to have you here. The DCTL is for loading DCTLs. You can’t do anything with it unless you load a DCTL. HueShift is a DCTL that you can load with it. There are many DCTLs, some free, some that you need to pay for. Hope that helps
Isn’t the issue the same in photography ? I wonder if lightroom supports substractive saturation or if there are workarounds / plugins for better film look emulation. (NOT shitty presets)
is this the same way the new color slicer works in 19? i was trying it out last night and noticed the scopes were lowering when i increased the sat, which was confusing me. if so , do you still see a reason to set your nodes up this way or will the channel slicer work ?
That's a good question. They use different maths, and the Mononodes plugin is split into different DCTLs. So there are differences in workflow and there will be differences in the results too. I guess it's like two different cars, they have similar functions on paper. You need to go test drive them to see the differences. Hope that helps.
Ooo, that's a good question. There are some similarities, i.e. it allowing you to manipulate primary colours. But printer lights won't left you shift hue, change density or saturation. So very different in that regard.
I've simply been using the green gain channel on an HSV node to emulate film density on the advice of a few professional colourists. Can you explain why that much simpler method is different from your four-node solution? As far as I can tell, increasing Saturation in the HSV color space only lowers luminance values on my waveform, and renders the deeper, richer colours you get with your method.
@@MrWarwickJones Hmmm... I think I need to apologise here! I had tested both methods previously and yielded different results. However, I just tested it again and like you said it's yielding identical results. Perhaps I'd overlooked something in my first tests. Although the video doesn't specifically mention the hSv method, I'll add a pinned comment to the video suggesting the hSv method as an alternative to the complex method outlined in the video as that will give a better impression of different methods of producing sub sat. Thanks for watching and commenting and for your persistence in helping this method bubble to light! PS I removed my incorrect comment so I don't leave bad information here in the comments!
@@team2films No problem. The price is fair. I'm just used to seeing ~10% discount from sponsorships. Thanks for getting back to me, and great video. You managed to simplify some difficult concepts.
Рік тому+1
Can't you just make one node: HSV, add qualifier to the same node and change density with green channel on wheels or curve?
Yes, that's a great point. But please consider the following. It can yield similar results and give the appearance of subtractive saturation in ideal scenarios, but it does not perform the same mathematical operations as the complex node tree or DCTL. It's just another way of faking subtractive saturation. In particular, it often produces a halo effect around transitions between colours and luminance values. Finally the need to use a qualifier or curve to limit sat operations to the brightest and most strongly saturated colours makes it more challenging to apply consistently. As a result I'd use the complex node tree or DCTL to apply a sub sat adjustment.
Рік тому
@@team2films Yeah but what is the difference? I tried the simple way and complex node tree. It gives me the same results.
@ Hello Piotr, just amending my earlier comment. Yes, the hSv method does yield similar results. Sorry it wasn't included in the original video. If ever we do a follow up on subtractive saturation, we'll include it. I had tested the method when I replied to your comment! But I think I must have overlooked something in the node tree that was skewing the results. My apologies.
Thanks for the video! Remembers me to the Technicolor Nodetree by the German channel Farbkanal. Does this have issues in a color managed DWG or aces project? I got this to work in rec709 timelines only jet :/
It works fine in different colour spaces. As always though, remember the tool will respond slightly differently. A grade created in one colour space won't copy to another, but it's broadly capable of doing the same things in either colour space.
Thanks, appreciate you saying that. We really try to teach topics thoroughly so people understand why and not just how and it’s nice when people like yourself notice that. Great to have you here.
Thank you for the high quality instruction on saturation! I have a question, however, on the comment that Resolve Does not use subtractive saturation. When I use the HDR color wheels to saturate by pushing the Sat slider to the right. The waveforms move in what appears to be the same manner they move when using the node setup "shenanigans" demonstrated in The Hard Way (or the one HSV node technique that others have already pointed out, which does the same thing). Am I missing something? It appears the HDR wheels saturation IS subtractive. Thanks again for the time and effort to put this tutorial together!
Hello, sorry for the delayed reply. Wanted to take the time to test this. I can't seem to replicate subtractive saturation with the HDR wheels. Is there anything I should to replicate your setup. When I use the HDR Sat slider and view a Y waveform it's true there is an overall decrease in the luminance of the image. However when you use the parades, you'll see that it is increasing the intensity of individual channels. HDR Saturation has a roll off in the highlights and shadows which will cause it to give a slightly different result than the regular primaries saturation slider, however beyond that I still appears to be operating as additive saturation in my tests, albeit with a slight decrease in overall luminance as the saturation increases. Looking forward to hearing your feedback. Thanks so much for watching and for your kind comment. It's great to have you here.
Hey guys. Every time I build a subtractive saturation tree, I find the image breaks down. I get large, blocky noise when its turned down or past 1.4. Any thoughts?
Hello, thanks so much for watching! Firstly, this video shows you how to do Subtractive saturation for free! The DCTL is presented as an option, but we wanted to give you a completely free option too! We're trying to inform our audience, not just push a given solution. Correction: You are correct! The hSv method does yield the same result! I did some further testing and I think I'd overlooked something in my initial test. My apologies for the misinformation and for not having included this method in the video.
Hello. Thank you for such a detailed explanation Perfect theory demonstration. But if you will - can't this be done simply in one node? When you switch to HSV color space - just increase gain in G channel - and it will add subtractive saturation. I compared results and they looks identical. It is also possible to apply qualifier to the same HSV node to keep everything in single node, or just add second node with alpha qualifier, similarly to what you did in the video. In that case there will be just two nodes. What are the cases when we need to build this nodetree with 6-7 nodes to achieve this result instead of directly increasing S channel in HSV node? Also, am I right that you lowered red channel in HSV node just as a demonstration, because it doesnt affect the final result even if we leave hue channel untouched? Or I'm doing something wrong? Thank you!
You're welcome! Yes the single node method you describe does work. There's a pinned comment about that. Go ahead and use the simpler version. Also, now you can just use color slice in DR19. Technology has moved quickly since this video was made!!!
@@team2films oh, my bad! Don't know how i missed pinned comment! Anyway your node tree explained greatly how to substract and add channels, which may be helpfull in other more complex cases. Yeah, new Color Slice is really nice tool. But it's great to know how to do this manually. Thanks again!
@@tomibegabsolutely no problems. Yeah, you get our point exactly…. We wanted to show the ‘science’ behind it, to know how it works manually even if you got and use more automatic tools. Appreciate you watching and commenting. Thanks so much.
Very useful thanks but would have been better if you just concetrated on how to use the dctl with examples rather than talking about theory for half the video. Would like to see how you graded those shots from scratch
Our apologies! We always do our best to explain the how and why rather than just showing people what to do. The grades were really simple though, a testament to the effectiveness of subtractive saturation. Worked a treat with those beautiful blue skies. Thanks for watching and for commenting. Always eager to improve our videos.
@@team2films no worries. Just the bit when you got the dctl makers to talk about it seemed like a sales pitch where that time which is the majority of the video could have been used practically. Just a thought for future vids as your channel could be extremely helpful. Thanks!
Not convicted, as you undercontrasted (or didn't match the contrast) between the HueLum Sat and Manual Sub Sat compared to HueShift DCTL. You can have the same ou close result with the pluggin by contrasting (like just look the hairs and sunglasses on the 2 first method, the missing contrast is a joke). I feel this comparaison unfair cauz the pluggin includes in its code contrast or similar stuff. Just an opinion :)
Hello, thanks so much for watching and sharing your thoughts. Maybe the below will help clear things up. Look forward to hearing your thoughts. It's better to think of sub sat as an operation rather than a look. It's a tool for achieving a look. It's not the look itself. As sub sat affects luminance, it will (in particular scenarios) increase the perceived contrast in an image. However it's not adding contrast which is technically a luminance only operation. There are other things that can be done to an image that will effect our perception of the image's contrast. The plugin is not adding contrast. It's using a very similar mathematical operation to the manual node based sub sat method that was demonstrated. It might be possible to emulate the look of sub sat using non-sub sat methods in particular scenarios ... but it will be more difficult to apply that look consistently across a variety of shots. Subsat is difficult to emulate. The best way is to actually perform sub sat. That's the main point of the comparisons. Thanks again for watching and commenting. Great to have you here.
Great video, I'm hoping you can answer a question I haven't seen answered in Cullen Kelly's video or this one: how do we implement this feature with higher compression footage? I shoot Sony cameras (FX3 and FX6) and while both cameras color holds up very well in grading, as soon as I try to implement these subtractive color tricks, the footage completely falls apart. Macroblocks appear every where, colors get ruined, skin tones, flickery pixels, etc. It seems to be an issue with compressed footage, and much less of an issue with high quality prores or RAW. I figure there must be a way to use this trick with sony footage or other h264 based footage without completely breaking the colors. Do you know anything?
@AxelKozber It's difficult to provide a solution without more information but here's some things to consider. Most cameras employe a compression technique called chroma sub sampling where higher levels of compression are applied to the colour information than to the luminance information. As subtractive saturation operates exclusively on the colours in the image, it's more likely to reveal colour compression artefacts. Your cameras are capable of recording to different flavours of XAVC. It might be helpful to check your manual to work out which ones preserves the most colour information in the final file. It is however perfectly possible to grade low bit depth - high colour compression files. Images cannot typically be pushed as hard though. Broader corrections are more likely to work rather than corrections with narrow qualifications to target a specific colour range, for example skin tones. There are 'tricks' to lessen these artefacts, like blurring qualifications, or qualifying based on luminance only... but at the end of the day, garbage in, garbage out (I mean that in the nicest possible way!) There are compromises that must be made. Thank-you both for watching. I hope the comments above were helpful. Great to have you here.
@neilseiffer5155 Thanks for the comment. Yes, h.264 is a super common delivery format... but the codec might be more capable for professional work that you might have realised. Please allow me to share some thoughts and forgive me if they are things you are already aware of: There are lot's of professional colour grading and editing workflows that extensively use h.264. Firstly because many prosumer grade cameras are used to supplement other cameras in production, but secondly because many professional cameras use h.264 compression. For example Sony's XAVC format (found in cameras like the FS7, FX9, etc) uses h.264 compression. In fact, XAVC is capable of recording 4:4:4. As a result h.264 is not just a delivery codec. It's extensively used as a capture codec in professional workflows. It's more accurate to say that H.264 is not typically used in high-end production. Thank-you both for watching and commenting. It's great to have you here.
What about changing the color space of one node to color space hsv and then disabling channel 1 and 3. Now you can use the gain tool to manipulate saturation. Not a good method?
I like the way you are thinking... but the gain operation only has the desired effect thanks to the the unique structure of the rest of the node tree. I'm sure there are other ways of building subtractive saturation node trees, but this is the method I'm familiar with.
@@team2films thank you for sharing it with us. I like when we learn from each other and grow as a community of creatives. Really appreciate your channel. Aesthetically pleasing and very well put together. Keep it up👌🏻👍🏻🙏🏻
Thanks so much @@The_Daliban Yeah, it's a great community and it's nice to be able to highlight the work of other awesome creators out there. I'm still digesting the maths, but I believe the gain operation on negative values is what makes it work.
Understanding color grading well means getting to know the technical parts of it. Using test images can show you the small differences between power grade methods and DCTL. Often, these can look like they give the same results, but there are small differences if you look closely. Instead of just asking which one is better, a good colorist will try out both methods and set up tests. By doing this, they can see and understand the differences better and know which one to use for different needs. So, it’s more about learning and trying out different things to see what works best, rather than just choosing one method and sticking to it. This helps colorists to get better at their job and to be more creative in color grading.
@@allenpayne9182 that’s definitely a good approach i also follow. Sometimes there is a technical reason for preferring one method over another. If that’s the case, i‘d like to know. I need to see a difference in the image that is in improvement or have a technical reason to chose a method. All else equal i‘m gonna go with the more efficient method👍🏻😉
hmm I’d argue projectors do not simply work in additive colours, at least internally, since they start with a white light source, and by using monochromatic filters and micromirrors, shave off unwanted light...
I see where you are coming from, but whether using a single bulb or not… a projector can be considered as having three separate light sources - Red, Green and Blue. Regardless once that red, green and blue light hits the wall it is definitely creating colours through additive colour. Does that make sense? Thanks so much for watching and commenting.
AI will replace this job in 5 years. Priority #1 for all digital media creators should be to slow down AI development. Discussion of any subject other than the existential threat AI poses is a waste of our time. We need a unified protest movement NOW.
That's an interesting perspective, but the future might not be as grim? Technological advances have disrupted industries before but after a period of adjustment has helped to automate mundane jobs and free up workers for more skilled work. Hopefully these new technologies will help editors be even more creative and productive! There will always be work for skilled editors and creatives.
Thanks for watching! Definitely not a commercial though. The DCTL is there to use if you want. Otherwise we demonstrated free ways of achieving the same effect. You don’t even need to pay for resolve. There’s also some great tips down in the comments for doing this other ways.
I knew I could trust this guy because his video has very good coloring. No bright red and blue LED lights blasting out the background like most of the UA-cam How-to crowd. Thanks!
Haha, thanks so much. It's great to have you here, we really try to focus on the teaching rather than the dazzle! Btw, please note there is a simpler way to do Sub Sat using the hSv method described in the comments.
What colour space are you using for your timeline on min 7:20? And have you already transform the log footage with a LUT? I cannot make it work on my project.
Hello! We’re using RCM to normalise the image. The timeline working colour space is DaVinci Intermediate. If you’ve not seen it, please check out our color management video :)
Unfortunately, that method does not yield the same result. Subtractive saturation should have a non-linear response across different luminance and saturation values. The method you described applies saturation in a more linear manner, affecting darker and less saturated colours more than desired. It yields broadly similar results with subtle adjustments, but does not look as good when stronger adjustments are made. Hope that helps! Feel free to try both methods and see how it affects the image and parade scopes.
This channel just breaks the internet with such huge information and knowledge. it's basically a masterclass for free. thanks
Thanks so much! Kudo's to all the people we mentioned in the thanks at the end though.
@@team2films 0:37 green is not a primary color lol. The primary colors are yellow, red, blue
@@RustyShackleford9000 Thanks for watching and commenting. I can understand why you'd think that, but green actually is a primary colour. Primary colours 'are sets of colors that can be combined to make a useful range of colors'. Different colour systems use different primary colours. For example in print Cyan, Magenta and Yellow are the primary colours used. In additive colour Red, Green and Blue are used as the primary colours.
You might find this article helpful: simple.wikipedia.org/wiki/Primary_color
Thanks again for watching and commenting.
@@RustyShackleford9000 lol
THIS is how Color Grading Tutorials should be done! Thorough explanation of how Things Work and why you do what and compare different methods and explain the effects of the techniques on the footage. Great channel
Thanks so much. You get us! 😄 It’s great to have you here.
@@team2films its just so good to NOT See someone pushing colors around wildly talking about "Juice"....If you know who I mean 😀
Hello, a big thanks to everyone who has pointed out that subtractive saturation can also be created with a single node set to hsv and then changing the gain of just the green (saturation) channel. We're calling that method hSv. It's a great tip and our apologies for not including that in the video. It doesn't change the fact that HueShift is a great tool that adds value to the process and speeds up your work! Thanks once again to everyone who has watched and commented. We appreciate you all!
You’re great🙏🏻
Thank you for sharing.
We are all here to learn.
Keep it up👍🏻🙏🏻🫶🏻
I'm two films in with Kaur, Jason, and Stefan Ringelschwandtner's DCTLs and I'm never going back to the "old ways" that involved complex node trees. The bang for your buck is a no-brainer.
That's great to hear. The convenience they bring to your workflow is well worth it.
One of the greatest tutorial channels for DaVinci! Remarkable job! One little correction i must mention because I am shooting on film a lot. 3:26, The density on a film is not coming from the saturation or the color richness on a negative film. The density is coming from the exposure. More overexposed the negative, the dense (thick) it will be. Underexposed means less dense (thin). Again. GREAT GREAT job guys. thanks for this wonderful video!
Correction. It is somewhat relative with the color richness also, but mainly the density is exposure.
Heya. Thanks so much for watching the video and leaving such a kind comment.
Some thoughts. Yes, density is a consequence of exposure, but exposure is also a consequence of saturation. Highly saturated colours are also typically highly exposed. Hence, the reason why density is being associated with saturation.
Saturation occurs when there is a difference in the exposure of the colour records. The greater the disparity the higher the saturation. So in the instance of a highly saturated blue... the saturation is caused by a dense magenta and cyan record.
Would love to hear your thoughts.
@@team2films "Highly saturated colours are also typically highly exposed". If you think about light reflectance, it should be other way round. Objects with highly saturated colors reflect less light, because they absorb more wavelengths that they reflect in comparison with pale and white colors. So, they should be less exposed on film.
@@Fedor_Dokuchaev_Color thanks for commenting and joining this thread. Yes, you are right that saturation does technically decrease exposure, many highly perceptually saturated colours in an image have above average exposure.
Please allow me to explain what I mean.
It all depends on your frame of reference. Yes, a highly saturated colour has less exposure than pure white therefore it has a lower exposure. But in the example provided above of a strong blue it has strongly exposed magenta and cyan records. That’s 2 out of three color records that are near fully exposed. As the most strongly saturated colours have the highest disparity between colour records, that’s why density (or exposure) is so strongly associated with saturation on film.
I look forward to hearing your comments.
@@team2films Thanks for your answer. True, but on the other hand, if you would take a picture of a super saturated blue sky and absolutely underexpose the frame, no matter how much saturated would be the sky, the negative would be super thin :D cheers.
Thank you as always for the nuanced run through and explanations - the interviews were a great addition too!
Thanks Ed. Much appreciated. Yeah, the interviews were a nice surprise. Really enjoyed hearing Jason's perspective on grading.
@@team2films yes always nice to see the faces behind these things and understand where they're coming from. I've seen a few vids on this area of saturation but this was by far the best. Thanks again, love your channel.
Yea this is insane. Color is insane. Thanks for overloading my brain. I'm having to replay this 42 times to get the message in my thick skull.
Ah thanks so much! Yeah it's pretty packed, so replay as many times as you want!
As the beginner colorist,this channel is very helpful for learning about colorgrading. Thank you very much for making this content
So glad it's helping on your colour grading journey.
Really interesting - I know both Jason and Kaur - really helpful and smart guys. I proposed the question to Cullen of why we tend to not fix the saturation in footage first, then decide on the amount to apply?
My opinion is pretty much what Jason and Kaurs tool actually does. I don't believe that saturation is something that can be a one knob fix in regular sat tools using any color space.
I also have the opinion that camera manufacturers, and pretty much every job on set is to nail skin tones - especially in camera, the digital saturation and how it is used is to push the best ( what they think is the best ) skin tone in their saturation decisions. Camera manufactures know this, and will target skin as a selling point.
Now because of that fact - a lot of the rest of the image may, or may not have the correct balance in saturation, and become either too saturated or undersaturated. Where as I mentioned - one knob isn't going to fix it.
The filmic aspect - the fact that light is being pushed through the film - you want something redder or greener - a film camera can't just add saturation - they need to increase the amount of color in the process, by adding density ( literally ) which of course will let less light through and it becomes darker.
Keep up these great videos - you are going to get a large following really quickl and the sucess will be well deserved. cheers
Thanks for watching Jim and for your encouraging comment. Jason and Kaur are great. Yeah, it's great to have such a variety of cool tools for addressing saturation. We are spoilt! And Kudos once again to Cullen for his work on this.
Thanks again for watching. Great to have you here.
I saw some guys use the second method, but instead of using nodes to separate the V-Channel they disabled these channels in the nodes drop-down menu. I'm not entirely sure if this approach has any differences, but it is way easier as you only need one node and it gives you subtractive saturation and similar results.
I can't say for sure without seeing the method you are referring to, but it might be performing a type of additive saturation (not subtractive).
@@team2films Yeah in a single node set colorspace to HSV with channels 1&3 disabled leaving only channel 2 enabled to control saturation via the gain wheel. Scopes behave the same way as method two so looks to be subtractive.
@@InfiniteRealms this is what I do as well
In fact you can see it in this video! When you adjust the green gain channel in your HSV node, the scopes look identical to when you're adjusting the gain at the end@@team2films
Update: Yes, that method does work. @MrWariwckJones kindly helped us out with that.
I met an editor/colorist a few weeks ago and this video is spot on. He also talked alot about great DCTL is and owns several plugins for it.
Thanks so much Tom. Appreciate you commenting.
You guys are till heavily under appreciated. I hope you gain the traction you deserve. Always top-notch quality. Thank you!
That's really kind of you. We're happy to know people are enjoying the videos.
Need to watch this vid on loop
Help yourself! Thanks for watching.
A phenomenal video! I learned a lot, thank you!
So happy to hear Martin. Thanks for watching.
Invaluable information, that's so cool! Especially compositing of color nodes and layers explanation. Huge thanks for sharing knowledge!
The irony is that one video interview about advanced color techniques in 4 bit color with banding and like shoot on a potato... 🤣
Hahaha, yeah, those are some webcam zoom interviews. Thanks for watching. Jason and Kaur are awesome.
Absolutely amazing advanced tutorial here!
I find it very difficult to find expert knowledge in this field because the rise of influencers lately..
I am curious if you have a paid course in this format?
Also I would love some help in the realm of sharpening! There are many ways to tackle it but most leave me feeling with gopro like footage.
Thanks again!
Hello! Yes, we have a course! It focuses on professional editing in DaVinci Resolve but it also tackles basic colour and audio work. There will be more details about that next week.
We love requests and suggestions, thank-you.
This is above my skill level, but still, what a beautiful tutorial. Thank you for such elegant teaching.
Ah thank-you Eric. That's really kind of you to say. When you are ready, it's there for you! In the meantime thanks for watching. It's great to have you here.
I want to see more videos like this.
Seeing these videos of you, feels like i did not miss and cinema school. Huge respect and a huge thank you.
Thanks. Glad it's helpful. More topics coming soon!
Loved this. Thank you!
Our pleasure.
Guys! This knowledge is changing my life!
Thanks so much Vlad.
Mind blowing! Bless you and your team for this info!
Our pleasure, thanks for watching.
I cannot like this video enough. Great info and expertly presented!
Thanks so much! Glad you enjoyed it so much.
Wow. This is a lot of knowledge shared.
Thank you!
Subscribed.
Welcome aboard!
The knowledge you shared for us is another level.Thank you so much
Our pleasure. Thanks for watching.
I think for someone like me who personally don’t like film grain or halation when emulating a film look, sub sat is the best way I can make my modern digital video look filmic combine with proper contrast curve that never hit true black at the bottom and true white at the top of the curve.
I personally will not 100% staying away from Sat vs Lum in HSL it’s sometimes can work as complimentary knob for HSV color space achieving that deep rich filmic look. You don’t want to over do it tho.
Awesome! Thanks so much for watching and commenting.
My brain melted!! Amazing tutorial!! Thanks for sharing it!!! :)
Oh no! We gotta get you back in the freezer! 🤣 Thanks so much for watching and commenting.
@@team2films Really amazing Video!!!
Thanks for these jewels 💎 man! 👏🏼👏🏼👏🏼
Glad you liked it :)
@@team2films One would expect though that a world class software like davinci would already have such a plug-in!(Simplyfied) let’s hope to an uptade soon!
@@blertonsaraciYeah agreed, it would be great if these features were integrated in the future. In the meantime though, It’s pretty amazing though that DaVinci’s node based engine makes it possible to perform these kind of operations. That’s also why Blackmagic has developed the DCTL api. So other developers can augment Resolve with additional functionality. It would perhaps make the program more difficult to use if it tried to incorporate tools for every style of colour grading.
@@team2films Absolutely, you are right! Just that with the time,(as you briliant guest mantioned) you start feeling/understandig the foundamentality of density, and it’s definitive belonging in the DaVinci world!
So I've played around with this and found out that setting a node to HSV and just raising the green channel yields the exact same result. When set to luminosity on a single hsv node it's identical to the presented node tree when just raising y gain. What I experimented with is splitting HSV and feeding Saturation into the Value channel and subtracting it that way. Beforehand I multiply the Value channel with the Saturation channel to avoid ultra saturated colors at the very extremes of the luminance ranges. The result is - in my opinion - the most flexible and the most pleasing looking. The corresponding node tree can easily be saved as a power grade and just needs one or two nodes to be touched to get nice results. It also has the benefit of lowering the gain of saturated colors, without adding saturation you don't want. Just my 2 cents. Cheers!
Thanks so much for watching! Yes we’re aware of that technique. There is a comment in the description about it. Doesn’t mention the multiply technique though. You can also use a curve instead of gain. That also gives you a roll off.
Subtractive saturation is also now implemented natively in resolve with the new color slice tool, and the film look creator.
The interesting part is, that I use a curve to bias how the luminance gets affected, the sky is the limit here. The gain then is just the to dial in an amount of said curve.
❤ This video was awesome 👏 Thank you 🙏
Thanks so much ❤️
I use colour density DCTL from mononodes.
I use it too, I think the math involved is impressive, is very hard to break the image. the result is very nice colors especially with Hue Shift.
Is that the Color Shift package?
@@ronlinehfx yes.
Glad you are enjoying it. Stefan has done great work on those tools.
So much great information, delivered in such a clear way!
Thank-you very much, appreciate you watching and commenting.
Incredible! Thanks for sharing.. I've bookmarked this for future reference :D
So glad you enjoyed it. Thanks so much.
This is incredible technique, thank you so much!
Thanks so much. Appreciate you watching and commenting.
Thank you for sharing knowledge.
Our pleasure. Great to have you here.
Brilliant! Thank you ! I liked and sub'd.
Thanks so much. Welcome!
Hi, wanted to point out an error when you were explaining about the "fake" methods. The second "fake" adjustment was hue vs luminance, but when you were explaining it you said "hue vs saturation globally desaturates all blue, regardless of whether it is saturated or not". Would like to clarify if what you meant was that hue vs luminance globally darkens all blue, regardless of whether it is saturated or not" ?
Also, you pointed out that sat vs luma desaturated unwanted parts of the image. Shouldn't it be the case that is darkens unwanted parts of the image, rather than desaturating? Since sat vs luma targets the luminance across a range of saturation, it does not objectively alter the saturation.
Pls correct me if i'm wrong but i needed to clarify this source of confusion. Love the well elaborated video btw!!🙏🏻
Having watched your great tutorial, and played with it a bit, and looking a lot ... I'm now frustrated that I can't work out how to replicate this in Photoshop! 🤣
Hahah! That's so funny! I'm sure there must be a way of doing it! For sure though there are times that I wish Lightroom or Photoshop worked more like Resolve. Would love to have tools like lift gamma and gain controls or Vs Curves in photoshop. Thanks for watching.
love the content!!
Thank-you so much! Appreciate you commenting.
It’s very amazing … great videos
Thanks so much.
you can do HSV subtraction with only 1 node.... change it to HSV color space, disable Chanel 1 and 3, use gain to saturate colors. done!
Yes! Apologies that wasn't included in the video. Thanks for watching and commenting.
Could you be mor specific where to disable Chanel 1 and 3?
@@SeanHuangRealEstatethere’s no need to disable the other channels. Just only manipulate the gain in the second channel using the primaries bars.
@@SeanHuangRealEstate Yes of course!!!
When you change the node to HSV color space, you can also right-click the node and select the option "Channels" right below... this will open a window with 3 channels already checked, being channel 1, 2 and 3 respectively for H, S and V. Then you can simply uncheck channels 1 (for Hue) and 3 (for value)... now you have only Saturation channel working. Simply control the amount of saturation with your "gain" wheel on log wheels. (Gain wheel gives the best results, I tried using other wheels and they also work but breaks the image considerably. (hope it helps my friend)
@@arthurangenendt8959 Thank you very much for the detailed explanation! It looks like using this method we can change the overall subtractive saturation. If I want to change the saturation of a certain color, then I will need to use a qualifier and then repeat the above?
Excellent tutorial, thank you! I have just one question - is it not easier to switch the node to HSV, turn off channels 1 & 3, and apply Gain/Gamma to the same node instead of creating the layer mixers?
Thanks for watching. Yes, you're correct. That's another great way of producing a similar effect. 👍
I never got round to asking, but did you both make it to the top?
Got to admit that the view was great when I did the Jungfrau many many years ago.
Hello, no we didn't summit any of the peaks. We walked down the glacier towards the Konkordiaplatz. Absolutely stunning walk.
@@team2films It is indeed a beautiful part of the world. Its been a pleasure chatting with you both.😀
This is experience❤
Mononodes has a great density dctl
Yeah, Stefan's tools are great too.
Great ! Thank for your Work!
Thanks for watching!
να σαι καλα, φιλε μου!
Ευχαριστηση μας
Thank you very much for the video. I have one question. How does the “hard way” 5:35 differ from switching the node to HSV color space and then manipulating the S channel, i.e., the Green wheel and/or curve? I ask because I learned this very same technique from Cullen Kelly’s TouTube channel and it does appear to yield better results than additive saturation. Thank you again and congratulations on your excelent videos.
Pleasure, thanks so much for watching and commenting. That's a great question! Please post a link to the @CullenKelly video. I'd love to take a look. He's a fantastic source of information on colour grading.
It sounds like you a describing a method that still produces additive saturation but in a way that is more pleasing than the stock saturation tool in Resolve.
@@team2films Thank you so much for answering my question! Here's the Cullen Kelly video I was referring to:
ua-cam.com/video/bAAjyNHDpyo/v-deo.html
I find the work you've been doing on your channel complements quite well what I learn from Cullen Kelly's videos, so I would love to hear back from you on this technique. Thank you again.
@team2films It's actually the same technique as described by @D_Aliban in an earlier comment 🙃
@@dlaimer Hello, thanks for your patience. We did some testing, and yes the single hSv method yields the same result as the complex node tree. Our apologies for not having included it in the video. If we ever make a follow up on subtractive saturation, we'll include that.
Great video, thank you!
Thanks
Thanks very much for an interesting and informative tutorial. Do you think that DR will add this kind of functionality natively to a future update?
That's a great question. It's hard to say. Blackmagic are always innovating and adding new features to the Resolve but DCTLs will always have a place though.
They did just add a feature a day ago that people were asking for. I have a feeling we're going to see something like this for Studio users. I prefer to learn the hard way because knowing how coloring works gives one better control for unforeseen problems. The images he's working with were properly shot with correct lighting and white balance. Working with terrible footage may be more challenging even with the DCTLs plugins he's promoting.
Perhaps it will provide a little more clarity when considering the Cullen Kelly 1 hSv node solution, considering that he often balances the saturation on this node with gamma AND gain.
The hSV method is great. It's capable of producing the same result as the more complex node tree we created. Sorry we didn't include it in the video too!
@@team2films No reason for an apology. I just wanted to respond to some comments that don't see the difference between your method and Cullen's method. Cullen then balances the saturation manually, while your node tree delivers a slightly better result out of the box.
@@jensgeumann7564 Ah thanks. That's kind of you! We did some more testing and it's possible to replicate the complex node trees results with the hSv method, but also at the same time there are some unique things you can do with either method too.
We're horrified by the prospect of sharing incorrect information 🤣 So we always appreciate people pointing out errors and helping us to give a more rounded impression of a given topic.
Sweet Jesus who sleeps with Mary in heaven!!! Tried this on a recent grade and it's like next level film-like! I always knew the saturation knob had something "off" about it. Always increased saturation with great misgivings
AWESOME! 🤩 Glad it helped.
Thank you for the helpful insight on DCTL.
Are the DCTL parameters mapped to the Blackmagic Mini Panel?
Yes, they can be controlled with a mini panel
@@team2films Thank you.
great video!! curious if there are any subtractive saturation for adobe plug-ins
Good question, I've not researched it. Sorry!
I have been trying both this "Hard way" method and the Bleach Bypass tutorial. I can't seem to get either to work. I have gone over the above video dozens of times, ensuring that I have each node and each adjustment/setting on each node correct. I have tried putting this node tree before the CST, LUT, Contrast, CST and 'Cineon Film Look' Nodes that I use as a default on all my Sigma FP footage. I've tried putting it after the FP default node tree, I've tried it between the LUT and Contrast Nodes. Which is where I believe it should best reside. But still I get nothing. Absolutely nothing happens when I adjust the Gain or either gain colour channel in the last Node before the "Add" mixer. Where am I going wrong? I've even tried starting a new project, with a blank slate, absolutely no nodes in the colour tab, and still I get nothing when using the above described "Hard Way" node tree. I'm befuddled.
I am total new to DVR, got the Studio version though. I got one question: What is the difference between the native DCTL tool and the HUEShift tool?
Sorry if this may appear a stupid question, as mentioned I am new to these kind of things. Thank you already.
Great to have you here. The DCTL is for loading DCTLs. You can’t do anything with it unless you load a DCTL. HueShift is a DCTL that you can load with it. There are many DCTLs, some free, some that you need to pay for. Hope that helps
Isn’t the issue the same in photography ? I wonder if lightroom supports substractive saturation or if there are workarounds / plugins for better film look emulation. (NOT shitty presets)
Hello, Sorry, I'm not sure if there are tools for doing this in Lightroom. Thanks so much for walking.
is this the same way the new color slicer works in 19? i was trying it out last night and noticed the scopes were lowering when i increased the sat, which was confusing me.
if so , do you still see a reason to set your nodes up this way or will the channel slicer work ?
Yeah, the new color slice tool uses subtractive saturation. Check out our video covering it.
Do you know how I could apply the subtractive saturation look to still images? Ideally using Lightroom CC on iPad?
No, Sorry, I'm not familiar with methods for doing that in Lightroom.
Which is better? Hueshift or Colorshift? 2 different DCTL's but they seem to do the same thing....
That's a good question. They use different maths, and the Mononodes plugin is split into different DCTLs. So there are differences in workflow and there will be differences in the results too. I guess it's like two different cars, they have similar functions on paper. You need to go test drive them to see the differences.
Hope that helps.
@@team2films makes sense. Thanks for clarifying.
thanks, is there a free trial for hueshift?@@team2films
This tool is kinda like using the printer lights?
Ooo, that's a good question. There are some similarities, i.e. it allowing you to manipulate primary colours. But printer lights won't left you shift hue, change density or saturation. So very different in that regard.
Amazing tutorial Leon! What is your keyboard model?
Thanks Gustavo. My keyboard is made by EditorsKeys. It's the Resolve model.
In case i use the middle method on my grade, on what position would you advice me to put it, in the beggining, middle or end of the tree node??
Sub Sat is usually a secondary operation. So after your primaries.
Targeted adjustments, non-breaking, comprehensive scope of the tools: the way of the future?
Yeah, it's pretty good. Great work from Jason and Kaur.
I've simply been using the green gain channel on an HSV node to emulate film density on the advice of a few professional colourists. Can you explain why that much simpler method is different from your four-node solution? As far as I can tell, increasing Saturation in the HSV color space only lowers luminance values on my waveform, and renders the deeper, richer colours you get with your method.
@@team2films i did try both, and pushed them to an extreme. The waveform is identical for both methods 🤷♂️
@@MrWarwickJones Hmmm... I think I need to apologise here! I had tested both methods previously and yielded different results. However, I just tested it again and like you said it's yielding identical results. Perhaps I'd overlooked something in my first tests. Although the video doesn't specifically mention the hSv method, I'll add a pinned comment to the video suggesting the hSv method as an alternative to the complex method outlined in the video as that will give a better impression of different methods of producing sub sat.
Thanks for watching and commenting and for your persistence in helping this method bubble to light! PS I removed my incorrect comment so I don't leave bad information here in the comments!
What is the spherical colour model that Kaur mentions at the very end? The gen, zen, jen...couldn't catch it.
Chen (I think!). Try message him on his channel if you want more details.
@@team2films Thanks, yes he indeed confirmed it was called (Tieling) Chen👍
how can i do this in premiere pro?
Hello! We’re not aware of how to do this in Premiere. Maybe someone makes a plug-in that does this?
damn!
Hi guys. Did PixelTools offer a discount code?
It's currently $20 off. No further discount that we are aware of at the moment though. Sorry. But it's a great price for what you are getting.
@@team2films No problem. The price is fair. I'm just used to seeing ~10% discount from sponsorships.
Thanks for getting back to me, and great video. You managed to simplify some difficult concepts.
Can't you just make one node: HSV, add qualifier to the same node and change density with green channel on wheels or curve?
Yes, that's a great point. But please consider the following.
It can yield similar results and give the appearance of subtractive saturation in ideal scenarios, but it does not perform the same mathematical operations as the complex node tree or DCTL. It's just another way of faking subtractive saturation. In particular, it often produces a halo effect around transitions between colours and luminance values. Finally the need to use a qualifier or curve to limit sat operations to the brightest and most strongly saturated colours makes it more challenging to apply consistently.
As a result I'd use the complex node tree or DCTL to apply a sub sat adjustment.
@@team2films Yeah but what is the difference? I tried the simple way and complex node tree. It gives me the same results.
@ Hello Piotr, just amending my earlier comment. Yes, the hSv method does yield similar results. Sorry it wasn't included in the original video. If ever we do a follow up on subtractive saturation, we'll include it. I had tested the method when I replied to your comment! But I think I must have overlooked something in the node tree that was skewing the results. My apologies.
Could you give a more detailed explanation of your procedure? I would like to compare this to the complex node tree too. Thanks!
@@SeanHuangRealEstateheya! It’s as simple changing the node’s color model to HSV and then manipulating the green gain primaries bar. Hope that helps.
I'm only 3 mins in and already feel blown away by how well this information has been presented and explained. Subscribed!
Thanks so much! We're so glad it's hitting right. Great to have you here.
Thanks for the video! Remembers me to the Technicolor Nodetree by the German channel Farbkanal. Does this have issues in a color managed DWG or aces project? I got this to work in rec709 timelines only jet :/
It works fine in different colour spaces. As always though, remember the tool will respond slightly differently. A grade created in one colour space won't copy to another, but it's broadly capable of doing the same things in either colour space.
This one just blew my mind. That wasn't some middle of the mall info you'll see the casual influencer spewing. Massive respect.
Thanks, appreciate you saying that. We really try to teach topics thoroughly so people understand why and not just how and it’s nice when people like yourself notice that. Great to have you here.
Thank you for the high quality instruction on saturation! I have a question, however, on the comment that Resolve Does not use subtractive saturation. When I use the HDR color wheels to saturate by pushing the Sat slider to the right. The waveforms move in what appears to be the same manner they move when using the node setup "shenanigans" demonstrated in The Hard Way (or the one HSV node technique that others have already pointed out, which does the same thing). Am I missing something? It appears the HDR wheels saturation IS subtractive. Thanks again for the time and effort to put this tutorial together!
Hello, sorry for the delayed reply. Wanted to take the time to test this. I can't seem to replicate subtractive saturation with the HDR wheels. Is there anything I should to replicate your setup.
When I use the HDR Sat slider and view a Y waveform it's true there is an overall decrease in the luminance of the image. However when you use the parades, you'll see that it is increasing the intensity of individual channels. HDR Saturation has a roll off in the highlights and shadows which will cause it to give a slightly different result than the regular primaries saturation slider, however beyond that I still appears to be operating as additive saturation in my tests, albeit with a slight decrease in overall luminance as the saturation increases.
Looking forward to hearing your feedback. Thanks so much for watching and for your kind comment. It's great to have you here.
Hey guys. Every time I build a subtractive saturation tree, I find the image breaks down. I get large, blocky noise when its turned down or past 1.4. Any thoughts?
Heya, sorry you are having problems. What kind of footage are you working with? It might perhaps be limitations of the codec you are using.
thank you for the reply! It's XAVC HS 4k on Sony FX3. After a CST converting slog 3cine to Davinci wide gamut @@team2films
Anyone have an idea how to do this in Photoshop? I am about to sleep, but will try out some stuff, I guess working in CMYK mode is a start.
No, I'm so sorry! Not sure how to do this in Photoshop.
Is this applicable in photography
Yes, it is applicable! But I'm not sure how you would replicate it in a program like Photoshop. Perhaps there are similar plugins available?
8:30 I don't understand what is the difference in just changing to HSV and increasing saturation (S), skipping all the other "doing/undoing" nodes.
Hello, thanks so much for watching! Firstly, this video shows you how to do Subtractive saturation for free! The DCTL is presented as an option, but we wanted to give you a completely free option too! We're trying to inform our audience, not just push a given solution.
Correction: You are correct! The hSv method does yield the same result! I did some further testing and I think I'd overlooked something in my initial test. My apologies for the misinformation and for not having included this method in the video.
Hello. Thank you for such a detailed explanation Perfect theory demonstration. But if you will - can't this be done simply in one node? When you switch to HSV color space - just increase gain in G channel - and it will add subtractive saturation. I compared results and they looks identical. It is also possible to apply qualifier to the same HSV node to keep everything in single node, or just add second node with alpha qualifier, similarly to what you did in the video. In that case there will be just two nodes.
What are the cases when we need to build this nodetree with 6-7 nodes to achieve this result instead of directly increasing S channel in HSV node?
Also, am I right that you lowered red channel in HSV node just as a demonstration, because it doesnt affect the final result even if we leave hue channel untouched? Or I'm doing something wrong? Thank you!
You're welcome! Yes the single node method you describe does work. There's a pinned comment about that. Go ahead and use the simpler version. Also, now you can just use color slice in DR19. Technology has moved quickly since this video was made!!!
@@team2films oh, my bad! Don't know how i missed pinned comment! Anyway your node tree explained greatly how to substract and add channels, which may be helpfull in other more complex cases. Yeah, new Color Slice is really nice tool. But it's great to know how to do this manually. Thanks again!
@@tomibegabsolutely no problems. Yeah, you get our point exactly…. We wanted to show the ‘science’ behind it, to know how it works manually even if you got and use more automatic tools. Appreciate you watching and commenting. Thanks so much.
The vanilla saturation tool😂
Hahah, I do love vanilla though!
Very useful thanks but would have been better if you just concetrated on how to use the dctl with examples rather than talking about theory for half the video. Would like to see how you graded those shots from scratch
Our apologies! We always do our best to explain the how and why rather than just showing people what to do. The grades were really simple though, a testament to the effectiveness of subtractive saturation. Worked a treat with those beautiful blue skies. Thanks for watching and for commenting. Always eager to improve our videos.
@@team2films no worries. Just the bit when you got the dctl makers to talk about it seemed like a sales pitch where that time which is the majority of the video could have been used practically. Just a thought for future vids as your channel could be extremely helpful. Thanks!
Not convicted, as you undercontrasted (or didn't match the contrast) between the HueLum Sat and Manual Sub Sat compared to HueShift DCTL. You can have the same ou close result with the pluggin by contrasting (like just look the hairs and sunglasses on the 2 first method, the missing contrast is a joke). I feel this comparaison unfair cauz the pluggin includes in its code contrast or similar stuff. Just an opinion :)
Hello, thanks so much for watching and sharing your thoughts. Maybe the below will help clear things up. Look forward to hearing your thoughts.
It's better to think of sub sat as an operation rather than a look. It's a tool for achieving a look. It's not the look itself.
As sub sat affects luminance, it will (in particular scenarios) increase the perceived contrast in an image. However it's not adding contrast which is technically a luminance only operation. There are other things that can be done to an image that will effect our perception of the image's contrast.
The plugin is not adding contrast. It's using a very similar mathematical operation to the manual node based sub sat method that was demonstrated.
It might be possible to emulate the look of sub sat using non-sub sat methods in particular scenarios ... but it will be more difficult to apply that look consistently across a variety of shots. Subsat is difficult to emulate. The best way is to actually perform sub sat. That's the main point of the comparisons.
Thanks again for watching and commenting. Great to have you here.
Thanks for the complete reply :) @@team2films
Great video, I'm hoping you can answer a question I haven't seen answered in Cullen Kelly's video or this one: how do we implement this feature with higher compression footage? I shoot Sony cameras (FX3 and FX6) and while both cameras color holds up very well in grading, as soon as I try to implement these subtractive color tricks, the footage completely falls apart. Macroblocks appear every where, colors get ruined, skin tones, flickery pixels, etc. It seems to be an issue with compressed footage, and much less of an issue with high quality prores or RAW. I figure there must be a way to use this trick with sony footage or other h264 based footage without completely breaking the colors. Do you know anything?
H 264 is a delivery codec, not for pro color grading or pro editing.
@AxelKozber It's difficult to provide a solution without more information but here's some things to consider. Most cameras employe a compression technique called chroma sub sampling where higher levels of compression are applied to the colour information than to the luminance information. As subtractive saturation operates exclusively on the colours in the image, it's more likely to reveal colour compression artefacts. Your cameras are capable of recording to different flavours of XAVC. It might be helpful to check your manual to work out which ones preserves the most colour information in the final file.
It is however perfectly possible to grade low bit depth - high colour compression files. Images cannot typically be pushed as hard though. Broader corrections are more likely to work rather than corrections with narrow qualifications to target a specific colour range, for example skin tones. There are 'tricks' to lessen these artefacts, like blurring qualifications, or qualifying based on luminance only... but at the end of the day, garbage in, garbage out (I mean that in the nicest possible way!) There are compromises that must be made.
Thank-you both for watching. I hope the comments above were helpful. Great to have you here.
@neilseiffer5155 Thanks for the comment. Yes, h.264 is a super common delivery format... but the codec might be more capable for professional work that you might have realised. Please allow me to share some thoughts and forgive me if they are things you are already aware of:
There are lot's of professional colour grading and editing workflows that extensively use h.264. Firstly because many prosumer grade cameras are used to supplement other cameras in production, but secondly because many professional cameras use h.264 compression. For example Sony's XAVC format (found in cameras like the FS7, FX9, etc) uses h.264 compression. In fact, XAVC is capable of recording 4:4:4. As a result h.264 is not just a delivery codec. It's extensively used as a capture codec in professional workflows. It's more accurate to say that H.264 is not typically used in high-end production.
Thank-you both for watching and commenting. It's great to have you here.
🌟🌟🌟🌟🌟
Thanks so much Cornelius.
What about changing the color space of one node to color space hsv and then disabling channel 1 and 3.
Now you can use the gain tool to manipulate saturation.
Not a good method?
I like the way you are thinking... but the gain operation only has the desired effect thanks to the the unique structure of the rest of the node tree. I'm sure there are other ways of building subtractive saturation node trees, but this is the method I'm familiar with.
@@team2films thank you for sharing it with us. I like when we learn from each other and grow as a community of creatives.
Really appreciate your channel.
Aesthetically pleasing and very well put together. Keep it up👌🏻👍🏻🙏🏻
Thanks so much @@The_Daliban Yeah, it's a great community and it's nice to be able to highlight the work of other awesome creators out there. I'm still digesting the maths, but I believe the gain operation on negative values is what makes it work.
Understanding color grading well means getting to know the technical parts of it. Using test images can show you the small differences between power grade methods and DCTL. Often, these can look like they give the same results, but there are small differences if you look closely.
Instead of just asking which one is better, a good colorist will try out both methods and set up tests. By doing this, they can see and understand the differences better and know which one to use for different needs. So, it’s more about learning and trying out different things to see what works best, rather than just choosing one method and sticking to it. This helps colorists to get better at their job and to be more creative in color grading.
@@allenpayne9182 that’s definitely a good approach i also follow.
Sometimes there is a technical reason for preferring one method over another.
If that’s the case, i‘d like to know.
I need to see a difference in the image that is in improvement or have a technical reason to chose a method.
All else equal i‘m gonna go with the more efficient method👍🏻😉
If I create a LUT from this grade, does the subtractive sat work?
It should do, but remember a LUT is an approximation of a grade. Results will vary as you feed the LUT different images.
@@team2films okay, but is it good for screen prep?
@@film.jonathanelia Probably. But I'd recommend testing the workflow first before relying on it for production.
@@team2films thank you
hmm I’d argue projectors do not simply work in additive colours, at least internally, since they start with a white light source, and by using monochromatic filters and micromirrors, shave off unwanted light...
I see where you are coming from, but whether using a single bulb or not… a projector can be considered as having three separate light sources - Red, Green and Blue. Regardless once that red, green and blue light hits the wall it is definitely creating colours through additive colour.
Does that make sense? Thanks so much for watching and commenting.
The pricing for these plugins are downright silly.
Hello. Thanks for watching. I’m sorry you feel that way about the pricing. May I ask why do you think it’s silly?
AI will replace this job in 5 years.
Priority #1 for all digital media creators should be to slow down AI development.
Discussion of any subject other than the existential threat AI poses is a waste of our time.
We need a unified protest movement NOW.
That's an interesting perspective, but the future might not be as grim? Technological advances have disrupted industries before but after a period of adjustment has helped to automate mundane jobs and free up workers for more skilled work. Hopefully these new technologies will help editors be even more creative and productive! There will always be work for skilled editors and creatives.
Interesting until it became a commercial.
Thanks for watching! Definitely not a commercial though. The DCTL is there to use if you want. Otherwise we demonstrated free ways of achieving the same effect. You don’t even need to pay for resolve. There’s also some great tips down in the comments for doing this other ways.
way too much for a lil sat.
Fair enough! Some jobs do demand a higher level of finesse though.
Color grading is... weird...
How so!?
Selling a DCTL is such a joke
Thanks for watching and commenting. I’m curious to know why you think selling a DCTL is a joke?
I knew I could trust this guy because his video has very good coloring. No bright red and blue LED lights blasting out the background like most of the UA-cam How-to crowd. Thanks!
Haha, thanks so much. It's great to have you here, we really try to focus on the teaching rather than the dazzle! Btw, please note there is a simpler way to do Sub Sat using the hSv method described in the comments.
Welcome DaVinci Resolve 19, where you can do it with one slider. Awesome update, and awesome content from you. Cheers!
Hahha, yeah almost! Some of these new features are pretty wild!
Is it really now one slider ?
What colour space are you using for your timeline on min 7:20? And have you already transform the log footage with a LUT? I cannot make it work on my project.
Hello! We’re using RCM to normalise the image. The timeline working colour space is DaVinci Intermediate. If you’ve not seen it, please check out our color management video :)
You can do this easily with one node in davinci without the plugin. Create a node in HSV color space, deselect channel 1 & 2 and increase the gain.
Unfortunately, that method does not yield the same result. Subtractive saturation should have a non-linear response across different luminance and saturation values. The method you described applies saturation in a more linear manner, affecting darker and less saturated colours more than desired. It yields broadly similar results with subtle adjustments, but does not look as good when stronger adjustments are made.
Hope that helps! Feel free to try both methods and see how it affects the image and parade scopes.
You Guys are soooo informative! What an amazing channel!
Thanks so much. Kind of you to say.
Where is the best position int the node structure to bring in hue-shift tool ? After the first CST or after the last?
I'd usually put it where you normally put your secondary corrections.
@@team2films okey cool, i try out also its works stronger or in a other way when use it after the CST
@@team2films before your output to display space.
Thank so much for this! I was having big trouble making my skin tones looking good and not so orange, this fixed my issue! Keep up the good work mate!
Ah! So glad to hear. Thanks for watching.