NOTE: For those not familiar with my usual color management setup, I do it with nodes, using 2 Color Space Transforms. You can see this shown in some of my previous videos. The 1. one is at the very beginning and goes from the camera's color space to DaVinci Wide Gamut, DaVinci Intermediate. This is my working / timeline color space. The 2. CST is at the very end and takes the working color space signal and prepares it for the display, in this case Rec.709 Gamma 2.4.
These concepts are so important, yet so easy to miss when first starting out as a colorist - it's not like you get an error or warning when you move the point. Though what if you did? 🤔
Does this rule only apply on the look dev process, meaning post group level or do you also set your pivot point on the clip level to the mid grey, when adjusting contrast on the clip level? Thanks for the good info 👏
@manuelsweingarten I'd say on a clip level, you'd only use sth different if the content of the shot/scene is a lot darker. meaning the "middle" gray is lower, ie dark night scenes. For well lit shots, I'd pretty much always go with middle gray as a starting point.
Hi Kaur. I agree that 33,6% grey can be a reference point for balancing but it's more a cinematographer concern. And it applies if you want to reproduce the chemical process of film. But it's different for television also because of the viewing environment. The issue I have is that you can't talk about a middle grey according to this 33,6% point. The middle grey is at 18% on cinematographers' grey charts. It has been deducted from the Fechner-Weber-Stevens psycho-physics' laws by an institution like the CIE (or another, I'm not sure, I need to check). If you read Munsell you will also find papers about this distribution of light. The first log curves was used to transform the signal to bring the 18% in point of the signal to a 50% out point to make the stops evenly distributed in the electronic signal and get a proper IRE to play linearly with the signal. Then the technologies evolved with knees etc to retain some extra informations. So what you said is not completely wrong in one specific way, but it's not accurate enough to be true.
Hi Charles! The 33.6% is the same 18% gray that everyone talks about point when encoded with the DaVinci Intermediate transfer function. If we apply the inverse and get the linear signal, it'll be at 18% or 0.18. I should've made that more clear in the video that the middle gray point sits in a different place for every transfer function / color space.
@@KaurH Oh yes you're right: if I apply a transform curve from linear to the Davinci wide gamut, the 18% gray raise up to ≈33%. But I don't understand the purpose. Is it a convenient way to meet the exposure if the skin is exposed at 33% for the DCTL? Oh wait, I was watching again. You set your color management without an input but you also have CSTs that I didn't notice (360p...). So I suppose you transformed the source gamma to DVR intermediate? But you need to trust the exposure on set, or to correct it. The purpose of your 'CORR' node I suppose? If it's the case, I come back to the first question. But the ODT transforms the gamma to a 2.2 or a 2.4 which level up the gray from the DVR intermediate. So I don't get the point. I try to understand because I'm maybe an old fashioned guy and I say to my students to do it by eye and I give them exercices to do it by eye considering the deliveries and the viewing environment for the public. I understand that a more quantified way can be more convenient for some of them and what you show looks like this. But there's something I miss in your color journey. I miss something in the way you achieve this and the purpose of the achievement. I'm not an English speaker, I hope I'm clear.
@@charlesledlaire No worries! Color management can be confusing at times and it doesn't help that I kind of expect the viewers of this video to be somewhat familiar with node-based color management. I'll do a video on that next! Essentially I first go from the camera's log to DaVinci Intermediate, and finally to gamma 2.4 as a part of the picture rendering also known as Output Device Transform (ODT). Throughout the journey, the same 18% gray has different code values and in DaVinci Intermediate, where I tend to grade, it's .336. But yes, you're absolutely right that pivoting around middle gray only helps when the DP is exposing "correctly".
@koushikbhattacharya832 Hey! There's a known bug with the system we're trying to fix where if you've gotten the toolkit once before, it doesn't send it again. Please send an email to the admin: info@pixeltoolspost.com Sorry for the inconvenience!
NOTE: For those not familiar with my usual color management setup, I do it with nodes, using 2 Color Space Transforms. You can see this shown in some of my previous videos.
The 1. one is at the very beginning and goes from the camera's color space to DaVinci Wide Gamut, DaVinci Intermediate. This is my working / timeline color space.
The 2. CST is at the very end and takes the working color space signal and prepares it for the display, in this case Rec.709 Gamma 2.4.
Such a fundamental concept that many seem to overlook. Thanks for spreading the right information! 😁
These concepts are so important, yet so easy to miss when first starting out as a colorist - it's not like you get an error or warning when you move the point. Though what if you did? 🤔
Does this rule only apply on the look dev process, meaning post group level or do you also set your pivot point on the clip level to the mid grey, when adjusting contrast on the clip level?
Thanks for the good info 👏
@manuelsweingarten I'd say on a clip level, you'd only use sth different if the content of the shot/scene is a lot darker. meaning the "middle" gray is lower, ie dark night scenes.
For well lit shots, I'd pretty much always go with middle gray as a starting point.
Both DCTLs will be sitting from now on in all of my node trees! Thanks Kaur!
@@948ccm Awesome! 😎
Lovely ! Precise, concise and an alternative in Resolve, such a great video !!!!
Thanks so much, Chloe!
I am so deeply grateful for your channel and resources. I use several tips from here. Thank you!
You're very welcome!
Hi Kaur.
I agree that 33,6% grey can be a reference point for balancing but it's more a cinematographer concern. And it applies if you want to reproduce the chemical process of film. But it's different for television also because of the viewing environment.
The issue I have is that you can't talk about a middle grey according to this 33,6% point. The middle grey is at 18% on cinematographers' grey charts. It has been deducted from the Fechner-Weber-Stevens psycho-physics' laws by an institution like the CIE (or another, I'm not sure, I need to check). If you read Munsell you will also find papers about this distribution of light.
The first log curves was used to transform the signal to bring the 18% in point of the signal to a 50% out point to make the stops evenly distributed in the electronic signal and get a proper IRE to play linearly with the signal. Then the technologies evolved with knees etc to retain some extra informations.
So what you said is not completely wrong in one specific way, but it's not accurate enough to be true.
Hi Charles!
The 33.6% is the same 18% gray that everyone talks about point when encoded with the DaVinci Intermediate transfer function. If we apply the inverse and get the linear signal, it'll be at 18% or 0.18.
I should've made that more clear in the video that the middle gray point sits in a different place for every transfer function / color space.
@@KaurH Oh yes you're right: if I apply a transform curve from linear to the Davinci wide gamut, the 18% gray raise up to ≈33%.
But I don't understand the purpose. Is it a convenient way to meet the exposure if the skin is exposed at 33% for the DCTL?
Oh wait, I was watching again. You set your color management without an input but you also have CSTs that I didn't notice (360p...). So I suppose you transformed the source gamma to DVR intermediate? But you need to trust the exposure on set, or to correct it. The purpose of your 'CORR' node I suppose?
If it's the case, I come back to the first question. But the ODT transforms the gamma to a 2.2 or a 2.4 which level up the gray from the DVR intermediate. So I don't get the point.
I try to understand because I'm maybe an old fashioned guy and I say to my students to do it by eye and I give them exercices to do it by eye considering the deliveries and the viewing environment for the public. I understand that a more quantified way can be more convenient for some of them and what you show looks like this. But there's something I miss in your color journey. I miss something in the way you achieve this and the purpose of the achievement. I'm not an English speaker, I hope I'm clear.
@@charlesledlaire No worries! Color management can be confusing at times and it doesn't help that I kind of expect the viewers of this video to be somewhat familiar with node-based color management. I'll do a video on that next!
Essentially I first go from the camera's log to DaVinci Intermediate, and finally to gamma 2.4 as a part of the picture rendering also known as Output Device Transform (ODT).
Throughout the journey, the same 18% gray has different code values and in DaVinci Intermediate, where I tend to grade, it's .336.
But yes, you're absolutely right that pivoting around middle gray only helps when the DP is exposing "correctly".
hey kaur i have been trying to get the tool from yesterday but i did not get any email...i submitted the information several time...
@koushikbhattacharya832 Hey! There's a known bug with the system we're trying to fix where if you've gotten the toolkit once before, it doesn't send it again. Please send an email to the admin: info@pixeltoolspost.com
Sorry for the inconvenience!
@@KaurH I sent the email plz. check
@@KaurH i sent you an email
Many thanks
👍 nice one
Skilz bro! Thank you for all your videos and work.