Company3, Tinting blacks/neutrals, Contemporary blue look

[Rihards]

Hi folks.

I've been chasing and trying to understand how to achieve a certain blue look very so often seen from the top shelf houses and their colorists. 

I will use references from a music video, YG -Scared Money

 
This was graded by one of the senior colorists at Company3, Joseph Bicknell. https://www.company3.com/artists/joseph-bicknell/

While they don't seem to be very eager to disclose whatever techniques they use, there is a finite amount of techniques that can be utilised and I suppose they are not wizards, well arguably at least. 

I've tried various acrobatics with curves and color spaces, LGG bouncing, change mixer, and color warper which got me the closest.

I'm curious how you guys would approach this sort of challenge?

Please check out the references.
Notice the blue tinting in neutrals as well as the overall blue tone.

 
Ref.2

Ref. 3

 

My Results, Color Warper + CST+Balance+PFE 2393

 

CST+Balance+PFE 2393

 

Node Tree and Color Warper Adjustment 

[Marc Wielage]

You may be going at this in a much more complex way than the people at CO3 did it. Bear in mind that a lot depends on Art Direction and the original photography, and it's hard to force a look on a shot that doesn't lend itself to that approach.

Veteran LA colorist Stefan Sonnenfeld is generally credited for popularizing the so-called Orange/Teal look, and he was doing it in the mid-to-late 1990s for commercials and home video projects on the old daVinci 2K. The 2K didn't lend itself to extreme secondaries or keys the way Resolve does now, and he actually did a lot of it through just careful Primary manipulation. 

You can argue now that it's possible to use layer mixers to retain more of the original skintone from the initial shot, then change everything else to cyan and use secondaries (or Color Warper) to minimize magenta. Basically, what you're looking for on the Vectorscope is a straight line from a little to the left of Red and then directly down to the Cyan box on the display... and not much else. Look at Transformers and a number of other Sonnenfeld features, and you can at least appreciate the consistency of what he does with the image, both on scopes and on the monitor.

@Stefan Ringelschwandtner has reverse-engineered some of these grades by taking the results, then pushing them back to a "Normal" look, which I think is a very imaginative thought process. His Mononodes are actually very interesting, but I'll let him comment further.

BTW, lest anybody criticize Stefan for this look, know up front that the ultimate decision is always made by the director and DP. It's our job as colorists to merely give them options to choose from. If they want orange/teal for maximum color contrast, then so be it -- it's their film. It's not a look I always like, but you can't deny the success of CO3 or Sonnenfeld. 

[Rihards]

2 hours ago, Marc Wielage said:

You may be going at this in a much more complex way than the people at CO3 did it. Bear in mind that a lot depends on Art Direction and the original photography, and it's hard to force a look on a shot that doesn't lend itself to that approach.

Veteran LA colorist Stefan Sonnenfeld is generally credited for popularizing the so-called Orange/Teal look, and he was doing it in the mid-to-late 1990s for commercials and home video projects on the old daVinci 2K. The 2K didn't lend itself to extreme secondaries or keys the way Resolve does now, and he actually did a lot of it through just careful Primary manipulation. 

You can argue now that it's possible to use layer mixers to retain more of the original skintone from the initial shot, then change everything else to cyan and use secondaries (or Color Warper) to minimize magenta. Basically, what you're looking for on the Vectorscope is a straight line from a little to the left of Red and then directly down to the Cyan box on the display... and not much else. Look at Transformers and a number of other Sonnenfeld features, and you can at least appreciate the consistency of what he does with the image, both on scopes and on the monitor.

@Stefan Ringelschwandtner has reverse-engineered some of these grades by taking the results, then pushing them back to a "Normal" look, which I think is a very imaginative thought process. His Mononodes are actually very interesting, but I'll let him comment further.

BTW, lest anybody criticize Stefan for this look, know up front that the ultimate decision is always made by the director and DP. It's our job as colorists to merely give them options to choose from. If they want orange/teal for maximum color contrast, then so be it -- it's their film. It's not a look I always like, but you can't deny the success of CO3 or Sonnenfeld. 

Thanks for your reply Marc. No doubt it's a very popular look and done by everyone and their dog however at varying degrees of finesse. 

I did too think that It shouldn't be too complex given it being around for that long however by observing this sort of looks through the years and with advancement of software ability a greater level of control is apparent.

In this specific example the luminance values that are low in cromacity have a controlled blue tint. Once that is achieved it's easy to point the vectors to their respective areas as you suggest. But the question remains how do we tint these luminance values in the best way while still keeping the image looking "Correct"?

Also just to clarify, in this exercise I hold reds and skin tones to little importance as the objective is tinting the neutrals.

[Marc Wielage]

Yeah, I have to say there's probably about 4 or 5 different ways to do it. One issue I've had is when suddenly there's a "flesh-colored" surface behind or near one of the actors, and suddenly a teal-look isn't possible for the background. I've had to explain to directors, "this is an art direction problem: you need to give me something to work with around the actor, and not pick a wall that's too close to the actor." We wound up rotoscoping the actor a bit in order to force more separation into the shot, but it was a bandaid fix at best and I'm not proud of the results. But: the director was pleased that I'd at least made the effort, and we moved on.

That's a valuable lesson I've told students before: never tell a client NO. At the worst, you could tell them, "that's a challenge because of this specific situation, but let's see how close we can get." Sometimes, they realize after a few minutes that you've gone down a rabbit hole and it could require lots and lots of time, and they'll take what we have and go on to the next scene. 

The key note is that Orange & Teal is not automatic, and a LUT alone won't do it. I had one memorable scene a few years ago where prisoners in a jail cell were just naturally orange & teal, and it worked great with almost zero effort. I used to get a little antsy about putting color in the shadows, but sometimes it's warranted and it can work to a point. 

[Rihards]

4 minutes ago, Marc Wielage said:

Yeah, I have to say there's probably about 4 or 5 different ways to do it. One issue I've had is when suddenly there's a "flesh-colored" surface behind or near one of the actors, and suddenly a teal-look isn't possible for the background. I've had to explain to directors, "this is an art direction problem: you need to give me something to work with around the actor, and not pick a wall that's too close to the actor." We wound up rotoscoping the actor a bit in order to force more separation into the shot, but it was a bandaid fix at best and I'm not proud of the results. But: the director was pleased that I'd at least made the effort, and we moved on.

That's a valuable lesson I've told students before: never tell a client NO. At the worst, you could tell them, "that's a challenge because of this specific situation, but let's see how close we can get." Sometimes, they realize after a few minutes that you've gone down a rabbit hole and it could require lots and lots of time, and they'll take what we have and go on to the next scene. 

The key note is that Orange & Teal is not automatic, and a LUT alone won't do it. I had one memorable scene a few years ago where prisoners in a jail cell were just naturally orange & teal, and it worked great with almost zero effort. I used to get a little antsy about putting color in the shadows, but sometimes it's warranted and it can work to a point. 

Would you mind sharing your interpretation of those techniques if you have the time? 

Hahah yes, I know that one. Just make the orange brick building blue or something 🤣 it will be fine. Must have been fun rotoscoping that...

You are so right  about the NO. Next time they'll just go to someone who can "figure it out" .

[Marc Wielage]

There are some good tutorials out there for on the net "Twin-Tone Looks" and things like that, and Lowepost has some good Look Development tutorials that I would recommend. So much depends on the nature of the original material, lighting, and exposure, that there's no one way or a "best" way to do it.

[Stefan Ringelschwandtner]

First: I would not try to "normalize" the final Grade back to "rec709" look. I tried it using Auto correcition tool but the "grade" (the look) is still in the neutrak image. So the best way to recreat a look is to find BTS Images. (in my opinion) .
And the best way is to analyze the Vectorscope and rebuild the shape of it. 
There are many ways...

My approach. (Using my "Soft 2 Strip" PowerGrade Node Tree)

1. Take a look at the "Vectorscope" .. and you will see it is a "Teal & Orange" or "Two Strip" (Soft Two Strip) Look.
2. I took the BTS image and applied my PowerGrade to it. Tweaked it a little (but no Hue vs. Hue, no curves, no Color Warper, not even Lift Gamma Gain)

It is not perfect, the lighting in the BTS image is different. But you can see at the vectorscope that it is going in the right direction.
This was a "quick and dirty" match to get the behind the scenes image into the ballpark. From there you can use "hue" to rotate the "2 strip" line
or you can use Lift Gamma Gain / Warper ... to tweak it further.

[Rihards]

55 minutes ago, Stefan Ringelschwandtner said:

First: I would not try to "normalize" the final Grade back to "rec709" look. I tried it using Auto correcition tool but the "grade" (the look) is still in the neutrak image. So the best way to recreat a look is to find BTS Images. (in my opinion) .
And the best way is to analyze the Vectorscope and rebuild the shape of it. 
There are many ways...

My approach. (Using my "Soft 2 Strip" PowerGrade Node Tree)

1. Take a look at the "Vectorscope" .. and you will see it is a "Teal & Orange" or "Two Strip" (Soft Two Strip) Look.
2. I took the BTS image and applied my PowerGrade to it. Tweaked it a little (but no Hue vs. Hue, no curves, no Color Warper, not even Lift Gamma Gain)

It is not perfect, the lighting in the BTS image is different. But you can see at the vectorscope that it is going in the right direction.
This was a "quick and dirty" match to get the behind the scenes image into the ballpark. From there you can use "hue" to rotate the "2 strip" line
or you can use Lift Gamma Gain / Warper ... to tweak it further.

Thanks Stefan, that looks quite close. I'll presume that you're not very eager to disclose the sauce of you powergrade 🤣

[Stefan Ringelschwandtner]

4 minutes ago, Rihards said:

Thanks Stefan, that looks quite close. I'll presume that you're not very eager to disclose the sauce of you powergrade 🤣

sorry. Not giving away my sauce for free.
But it is super easy and very minimal and easy adjustable, and I really think it works with a lot of different movies (as you can see on my website).It did not cause any bandings and artefacts. You can really push and shift the grade in a direction of a reference image. There is nothing special about my grade. it is a "simple trick".  I've seen some people using this technic in youtube videos. I´ve just combined a very simple technic with 1, 2 other knobs. I love minimal setups / minimal node trees. The pack is not just about: "twist this knob ... done". The User Guides also will teach you how to use the minimal setup, how to read scoped, how to analyze footage, and more. I hope it helps people to think differently about grading and say: oh, with this grade... I can go into that direction, but also into another direction.  

[Jamie Neale]

Lots of ways to achieve the look but does also depend on where you're placing the look (locally or globally). Working globally then think broader strokes working with less aggressive tools (avoid keys etc). Working locally then follow Marc's advice or as Stefan mentioned creating a 2 strip look works well. The 2 strip look limits the colours based on an old subtractive print process, it's biased towards blue/green vs red/orange. That bias is what gets us the beloved teal/orange look but if neither of those colours are present in the frame then you will get mixed results that probably won't look like the C3 example.

A really quick and dirty way to test the look on your Morocco shots would be to key the red hat in a layer mixer then on the bottom node (which is the top node) put Lum mix to 0 and then add cyan by lowering the red channel in Offset. Then use Curves to dial in the orange/red for the hat.

I'm not saying that's the way to do it but at least you'll get close enough to see if the look is a good fit for your project. 

Some techniques that will help and worth looking into are split toning, 2 strip process and subtractive colour. It will help you build a more usable look/power grade that you could use again and again

 

[Nicolas Hanson]

On 2/10/2022 at 4:46 PM, Rihards said:

I've been chasing and trying to understand how to achieve a certain blue look very so often seen from the top shelf houses and their colorists. 

If you have access to Ravengrade for Resolve, there are several film print emulations there that will work perfectly as a starting point for these types of looks. The rich and bold film contrast, tinting and strong restricted primary colors.

[Jamie Neale]

2 minutes ago, Nicolas Hanson said:

If you have access to Ravengrade for Resolve, there are several film print emulations there that will work perfectly as a starting point for these types of looks. The rich and bold film contrast, tinting and strong restricted primary colors.

Good shout. I'm really enjoying Raven Grade, the volumetric highlight/shadow adjustment is a really nice feature too.

[Nicolas Hanson]

6 minutes ago, Jamie Neale said:

I'm really enjoying Raven Grade, the volumetric highlight/shadow adjustment is a really nice feature too.

Yeah that's one of my fav features too, I used to key the highlights and decrease contrast within that range but now Ravengrades Vol/Ratio handles it for me - with a roll of that makes it look good every time.

Edited February 15, 2022 by Nicolas Hanson

[Adéyẹmi]

If you are quite new to colour, my advice to you is to not bother at all about Plug-ins and LUTs. Stick to your LGG, sat and colour curves. Knowing how to tweak things manually, colour seperation, tonal range gradient. Thats where real practice lies. You need to train your eyes and grow your taste. These are the things that makes you become an actual colourist. And then later, you can experiment. Believe u me, as far as grading is concerned, what matters is the photography, your monitor and computer speed. Everything else is very minor.  This look can be created with 2 to 3 nodes using LGG primaries, and colour curves only. 

 

Edited February 16, 2022 by Adéyẹmi

[Jamie Neale]

22 hours ago, Adéyẹmi said:

If you are quite new to colour, my advice to you is to not bother at all about Plug-ins and LUTs. Stick to your LGG, sat and colour curves. Knowing how to tweak things manually, colour seperation, tonal range gradient. Thats where real practice lies. You need to train your eyes and grow your taste. These are the things that makes you become an actual colourist. And then later, you can experiment. Believe u me, as far as grading is concerned, what matters is the photography, your monitor and computer speed. Everything else is very minor.  This look can be created with 2 to 3 nodes using LGG primaries, and colour curves only. 

 

Really solid advice that, becoming a jedi of the LGG and extended tools will make for a much more versatile colourist. A good exercise is take your favourite LUTs deconstruct them and try emulating them using primaries. When I started I definitely went plug in/LUT heavy but now I find most of what needs to be done can be achieved with mostly primaries.

LUTs/Plug ins still play a part for me but they end being for something very specific. That said, Raven Grade is a great addition to the toolbox and well worth checking out even if it's just to understand what the different sliders do to your image.

[Keidrych wasley]

2 hours ago, Jamie Neale said:

Really solid advice that, becoming a jedi of the LGG and extended tools will make for a much more versatile colourist. A good exercise is take your favourite LUTs deconstruct them and try emulating them using primaries. When I started I definitely went plug in/LUT heavy but now I find most of what needs to be done can be achieved with mostly primaries.

LUTs/Plug ins still play a part for me but they end being for something very specific. That said, Raven Grade is a great addition to the toolbox and well worth checking out even if it's just to understand what the different sliders do to your image.

Try recreating Yedlins LUT, or Mitch’s 2383 with LGG, you’ll soon come up against a wall of complexity. 

Recreating a complex print film LUT using primaries is impossible.  The maths available in LGG can’t get you even remotely close, because the underlying math in LGG is too simple / basic. 

This is a good reason why studying color science and LUTs has value because it helps to understand the underlying math and what is happening ‘under the hood’. 

Learning to use LGG and learning about LUT building (along with the more complex color science that LUT building entails) are equally important and both of high value. 

The sliders in Ravengrade from memory are mostly simple operations, eg adding gain in linear for exposure, changing tone curve for ‘volume’ etc. The complexity is in the data collection and the implementation of the data used to build the looks. For example Mitch supplied a print film emulation based upon a very large data set and implemented extremely smoothly. Take a look at some of those looks as a graphical 3D cube, there’s some very advanced color science on display such as extremely smooth outer gamut curvature etc that made me smile.
 

[Adéyẹmi]

30 minutes ago, Keidrych wasley said:

Try recreating Yedlins LUT, or Mitch’s 2383 with LGG, you’ll soon come up against a wall of complexity. 

Recreating a complex print film LUT using primaries is impossible.  The maths available in LGG can’t get you even remotely close, because the underlying math in LGG is too simple / basic. 

This is a good reason why studying color science and LUTs has value because it helps to understand the underlying math and what is happening ‘under the hood’. 

Learning to use LGG and learning about LUT building (along with the more complex color science that LUT building entails) are equally important and both of high value. 

The sliders in Ravengrade from memory are mostly simple operations, eg adding gain in linear for exposure, changing tone curve for ‘volume’ etc. The complexity is in the data collection and the implementation of the data used to build the looks. For example Mitch supplied a print film emulation based upon a very large data set and implemented extremely smoothly. Take a look at some of those looks as a graphical 3D cube, there’s some very advanced color science on display such as extremely smooth outer gamut curvature etc that made me smile.
 

True that, but what I am simply putting out for someone new in colour is to start small. I have seen ppl two months into colour looking for tutorials on how to make their own DCTL. Its totally unnecessary to go complex when you dont understand the basics. Like any other artistic practice. You shouldn't jump to playing Chords if you dont know your basic Tonic Solfa. Ppl are too overwhelmed with the technology and tools.  They want to start big rather than basic.  The other day I had a conversation with a film student, she was quick to talk about kodak this and that, but didn't know what printer lights is. The look above can be achieved with a combination of colour curves and colour seperation if the photography lends its self. Whether scene or display referred. 

Edited February 17, 2022 by Adéyẹmi

[Keidrych wasley]

7 minutes ago, Adéyẹmi said:

True that, but what I am simply putting out for someone new in colour is to start small. I have seen ppl two months into colour looking for tutorials on how to make their own DCTL. Its totally unnecessary to go complex when you dont understand the basics. Like any other artistic practice. You shouldn't jump to playing Chords if you dont know your basic Tonic Solfa. Ppl are too overwhelmed with the technology and tools.  They want to start big rather than basic.  The other day I had a conversation with a film student, she was quick to talk about kodak this and that, but didn't know what printer lights is. 

Wholeheartedly agree. A simple tone curve, sat and LGG can be immensely powerful. 

I just think understanding a tools limitations is an important part of learning how to use them.

[Jamie Neale]

39 minutes ago, Keidrych wasley said:

Try recreating Yedlins LUT, or Mitch’s 2383 with LGG, you’ll soon come up against a wall of complexity. 

Recreating a complex print film LUT using primaries is impossible.  The maths available in LGG can’t get you even remotely close, because the underlying math in LGG is too simple / basic. 

This is a good reason why studying color science and LUTs has value because it helps to understand the underlying math and what is happening ‘under the hood’. 

Learning to use LGG and learning about LUT building (along with the more complex color science that LUT building entails) are equally important and both of high value. 

The sliders in Ravengrade from memory are mostly simple operations, eg adding gain in linear for exposure, changing tone curve for ‘volume’ etc. The complexity is in the data collection and the implementation of the data used to build the looks. For example Mitch supplied a print film emulation based upon a very large data set and implemented extremely smoothly. Take a look at some of those looks as a graphical 3D cube, there’s some very advanced color science on display such as extremely smooth outer gamut curvature etc that made me smile.
 

You're obviously right about the hidden complexities of LUT but my comment was more about getting someone excited to dive into the look creation rather than making it sound impossible. I'm not sure you do need to know the math to be a great colourist. But that's the beauty of being a colourist. You can be both a scientist and an artist or one or the other. There's plenty of room (dark ones) for us all

[Claudius Neal]

My version using blue