Mastering Color Grading: Techniques, Tools, and Tips for Cinematic Excellence
Explore the art of color grading in cinema, from basic concepts to advanced techniques, and learn how to achieve stunning visual effects using modern tools.
File
Color Grading 101 - Everything You Need to Know
Added on 09/30/2024
Speakers
add Add new speaker

Speaker 1: From the strong and gritty looks of The Matrix, Mad Max Fury Road, and Blade Runner 2049, to the unique pink and soft atmosphere of the Grand Budapest Hotel, or the famous teal and orange look of Michael Bay films, color grading has always had an important place in cinema. Even in the days of black and white, different chemical treatments made it possible to give different atmospheres to the image. Most recently, in the 1960s, the Italian giallo style, a film genre that mixed thriller and horror, brought its influence of bright and flamboyant colors, which could feel almost surreal at times, and it was immediately linked with its lighting and production design. While back then, treatment to the final image was done through a chemical process on the actual film, today we have robust and easy-to-use tools to perform the color grading step. And today, that's exactly what we are going to be diving into, starting with some important basic concepts. Color grading is a post-production process, a final sweetening of the image. It shouldn't be the salvation of the image, so it is essential to lock in your look on the day, getting as much in camera as possible to create a consistent atmosphere through lighting, production design, composition, and so on. But once in post, there are two stages of color work, color correction and color grading. Color correction comes before our color grading step and consists of harmonizing our image to have a neutral base ready to be worked on. During this step, if necessary, we will correct the white balance, equalize the exposure, reduce the highlights, increase the midtones, and apply noise reduction. But also, we are able to match our shots between different angles and cameras. This step is very important, and when done well, will save a ton of time when we shift to color grading. Once our image is properly balanced, we do move into that color grading step. This is where we will give character and atmosphere to our scene. We can play with curves, masks, color wheels, and other very precise tools to land the look that we're looking for. We can also apply some effects such as grain to give an organic look and texture to our image or glow to create a light atmosphere, or even play with some chromatic aberrations. When we work on an image, whether for color correction or color grading, we are working in a limited color space. The most common profile that most screens use by default is Rec. 709. Its purpose is to ensure that all devices display the same image. However, since the standard was created in 1990 and is now outdated, it's being replaced by Rec. 2020 and Rec. 2100 for cinema. And with it, we get much more rich HDR color space. HDR, which stands for High Dynamic Range, allows for more brightness, very vivid colors, and deep blacks. Rec. 709, Rec. 2020, and Rec. 2100 are essentially broadcast color spaces. In practice, if you want to export for editing for distribution on, say, YouTube, you can use Rec. 709. For broadcast on Netflix, it's going to be necessary to use Rec. 2020, or more precisely, the P3D65, a more realistic version of Rec. 2020, since to date, no screen is capable of displaying all the nuances that Rec. 2020 provides. There are also acquisition-specific color spaces. Each camera manufacturer offers its own color space optimized for its sensor. For example, ARRI with their Alexa cameras provide Alexa Wide Gamut color spaces, RED offers the Red Wide Gamut RGB, and DJI the DJI-D Gamut. Not to be confused with Gamma, Gamut relates to color, while Gamma relates to brightness and contrast. In any case, the best way to see the color space is as a container. With most cameras, we have several options for how we capture our image. The first, and most standard, is to record in Rec. 709. Our image is immediately usable. However, we will obtain an image with limited amount of information, and are then limited in possibilities for color grading. Highlights will tend to clip quickly, and shadows will contain only a few details. This is useful for fast turnarounds, but not when you want complete control over your image. Another possibility is to record in RAW. RAW saves as much information received by the sensor as possible, which gives us a very rich image. But there are some drawbacks for the uninitiated. Firstly, it's size. RAW files tend to be much larger, which means the storage and overall workflow become much more complicated as a result. But you are getting a ton of a lot more useful data. RAW images are often recorded in 12 or even 14 bits, while in Rec. 709, we generally stay at 8 bits. The amount of bits refers to the number of colors that can be obtained. To give you an idea, in 8 bits, we have a maximum of 16 million colors. In 12 bits, it is almost 69 billion colors that we have at our disposal. The other problem with RAW, which is also what makes it interesting, is that the recorded image requires development or debayering. To be technically more exact, a RAW file is a mass of information and not an image itself. During this development process, it's us who will be able to choose which details we want to keep or not in our image. We will also be able to play with many parameters that would normally be baked in if we had recorded in Rec. 709. For example, white balance, tint, ISO or ASA, exposure, noise reduction, and so on. So RAW is definitely ideal, but again, there is the tradeoff of needing enormous storage and power. However, there is a good alternative to all of these constraints, which is LOG. Unlike RAW, which records all the information from the sensor in order to develop a usable image, LOG uses an optimized gamma curve to record an image containing as much information as possible in the highlights and shadows. So LOG is sort of a middle ground between RAW and Rec. 709. LOG uses a logarithmic curve rather than a linear curve in the way to record the highlights and shadows. The LOG curve will deliberately push the lowlights up to bring out more detail and pull the highlights down to avoid overexposures. This is done at the camera's sensor and before encoding in a video file, which usually will be 10 bits. So the details are truly preserved. And when shooting in LOG, for the most flexibility, it's best to try to expose your image as much as possible in the highlights just before clipping. The LOG image is recognizable by its desaturated appearance and its very grayish low contrast. It's an image that cannot be used as is. When importing a sequence taken in LOG in our color grading software, we must define a conversion. Generally, it's going to be a LOG conversion to Rec. 709, which again is the most used broadcast standard. Each camera manufacturer offers its own LOG profile adapted to its sensor. At ARRI, it's the LOG-C. At Sony, you have S-LOG, S-LOG-2, S-LOG-3. At RED, the RED-LOG film. Each LOG must be converted in a certain way. Fortunately, current tools allow these conversions with one click. Or if you don't have that, all of our LUT packs that we sell on our Triune Digital store do have utility LUTs included that will do this for you. But thanks to LOG, we have an image rich enough to perform advanced color grading, but without the huge weight of RAW files and its complex workflow. So it's really ideal for a colorist. ACES is a dynamic color grading process that allows you to work on original video files, whether in Rec. 709, LOG, or RAW, without having to conform each sequence to a destination color space. So we can work on the full range of colors that each shot gives us. The conversion to destination color spaces is done at the end of the chain, or at the end of the ACES pipeline. This makes it very easy to distribute a project on different platforms with different standards. For example, let's say you have an edit that includes clips recorded in LOG with a wide dynamic range and lots of highlights. If you work directly in Rec. 709, you will lose this dynamic range and your highlights will be burnt out because Rec. 709 is limited. For a YouTube upload, that's no problem, but if tomorrow you wanted to export your edit for HDR projection or Netflix streaming, then you will need to rework your entire edit. However, while working in ACES, the wide dynamic range of your video files will be preserved and considered throughout the color grading process, and at the end of the chain, you can choose to export either in Rec. 709, HDR, Rec. 2020, P3, and so on. With an ACES workflow, you benefit from all the color capabilities of your image right from the start, and keeping them throughout the color grading process. If there must be a limitation, it will be done on export. Let's pause there to thank today's sponsor and the one plugin that I've used on every project I've done for about six years now, and that's Film Convert. Film Convert Nitrate is a plugin that gives you the look of actual film stock, like Kodak 5207 or 5213, or some Fuji stocks and others. You select one and you immediately get that look applied in your Luma and Chroma, which you can adjust how much of the color and contrast are affected independently. It's also some of the best film grain out there. You can pick from different types as presets here, then can dive in further to customize exactly how you want it to look. More or less in the shadows or highlights, the size, amount, and so on. You also have great controls to dive in deeper and fully grade your piece from this one app if you wanted. It's a very powerful tool, and one that I use as the finishing touch on every project I grade, including our latest short, Century. Applying Nitrate at the end of the line really unifies everything and takes it that last step to fully feel filmic. If you want to try it out for yourself, check out the link in the notes below and use the code Film Riot to get 10% off of Nitrate. Logo. You've definitely already heard of LUTs, but what is a LUT? A LUT, or a lookup table, is a file containing instructions for replacing one RGB value with another, based on hue, luminance, and saturation. Really, you can think of it as a color filter. A LUT only affects colors and luminance and cannot contain effects, so while a LUT can adjust your color and contrast, a LUT cannot add grain, vignettes, reduce noise, correct chromatic aberrations, and so on. LUTs are often used to facilitate color space conversions, for example converting a log image to Rec. 709 like we talked about before. They're also used for calibration of screens, and these types of LUTs would be called technical LUTs. An excellent use of LUTs in production is to see something close to the final intended image while shooting in log. Most monitors do have the ability to load a LUT, and many cameras do as well, so you could create a unique LUT that will be your final look or use a standard Rec. 709. But then, of course, you have creative LUTs. Creative LUTs have many uses. They obviously facilitate the work of color grading by applying a ready-to-use look to the video clips that will harmonize the entire edit. They can also serve as a final touch or starting point for an intricate original look. And thanks to LUTs, we can easily switch from one style to another, auditioning looks and moving right along. LUTs also act pixel by pixel, so they are extremely precise. But it's important to remember that a LUT is just a tool, so its effectiveness depends on its use. A chainsaw is more efficient than a manual saw, as long as you know how to use it correctly and don't lop your own arm off. The common mistake with creative LUTs is being overdone with their use. Almost all tools that support LUTs allow you to play with the intensity of the LUT. So for example, if we use a LUT as a final touch, let's apply it to around 25-50% max for a solid end result without pushing it too far. Most video editing software offers built-in color grading tools. Some are more advanced than others, but all decent software will be good enough to get a workable end result. But the best bit of software for post-color is by far DaVinci Resolve. DaVinci Resolve from Blackmagic Design was initially exclusively dedicated to color grading and reserved mainly for pro-colorists. But of course, the software continued to evolve into what we have today, a complete post-production suite accessible to everyone since you can get it for free right now. There is a studio version that lets you get more, but the free version is loaded with everything that you would need to take your project all the way through post. There are several strengths that make Resolve one of the leading color grading tools. The first is its use of nodes. Like layers in Photoshop, nodes allow you to add adjustments and effects in an organized and non-destructive way. It's then easy to rework upstream and downstream of each step. The nodes also give you great visual workflow to facilitate the work. Then you have its masking capabilities and tracking tools. In a few clicks, we can select and correct the skin tones on an actor's face and track it to that area. We can also use power windows to isolate a part of the image and apply correction and effects to it. But you may not want to dive into Resolve or learn new software or get that advanced. In those cases, what you have is likely more than enough, like using the Lumetri panel in Premiere Pro or the tools in Final Cut Pro X. In the end, the software is just a tool. When we work on the color correction or color grading of an image, we absolutely have to control what we are doing with the scopes. Scopes may seem scary at first, but they are essentially tools that ensure that our modifications do not degrade the image so we can control the balance overall. Scopes are more reliable than our eyes, so it's perfect for that use. There are many different scopes, but let's take a look at the three main ones. The waveform is a graph that gives you important information about your image. Your highlights and lowlights are represented vertically, with everything at the top being pure white and at the bottom pure black. So here, we can easily check that our image does not include over or under exposure areas. If I exclusively increase the brightness of my image, you can see the waveform crashing against the top of the graph showing that clipping, and vice versa if I reduce the brightness. Then we have the RGB parade, which is nothing more than a waveform broken down into three graphs, red, green, and blue. With the RGB parade, we can check that our image is neither overexposed or underexposed, but also ensure that its RGB balance is correct. If our image contains a blue color cast, for example, we will see that blue graphics shift up compared to the red and green. So we can use the RGB parade to control our white balance. And finally, the vectorscope. It's quite a different chart that comes in a circular form. In the center of the graph is our image. All around that, we find the hues arranged in the same way as on the wheels. Red at the top, yellow on the left, blue on the right, green at the bottom. The more we move away from the center, the more saturation increases. If I deliberately push the blue to an extreme, we can see the graph moving away from the center. A properly balanced image will be represented by a well-positioned graph in the center of the vectorscope. The vectorscope has a tool that can be useful, the skin tone indicator. In theory, the skin tones of our image should be positioned on this line. Otherwise, they will either be too pink or too orange. To check this, we can isolate the skin tones with a power window or a mask and look at where they are positioned on the vectorscope. As you can see, the scopes are a very important tool since after a while, your eyes may lie to you, but the scopes never will. You probably already know the RGB curve, which allows us to correct the brightness of our image as well as its contrast, either globally or individually on each channel, red, green, and blue. But then we have slightly different curves. The Hue vs Sat curve allows you to modify the saturation based on a hue. For example, if I wanted to increase the saturation of the blues, I can move up just that area here. Or I could use the eyedropper to select a color directly from my image. The Sat vs Lume curve allows you to modify the luminance of a color based on its hue. So if I position a point in the red-orange, you can see I can lighten or darken the skin tones of this image right here. The Lume vs Sat curve allows you to reduce or increase saturation based on luminance, so we can adjust the color saturation in the highlights and shadows. And finally, the Hue vs Hue curve allows you to modify the hue based on a hue, like the skin tones. If I place a point right here, you can move up or down and you see the hue of the skin tones change drastically. It's a very convenient and quick way to do small adjustments. And now that we've had an overview, let's take a look at the entire process at work. Here we have a scene shot in Log. My timeline is Rec. 709, so I import my Log clip and apply a Log to Rec. 709 conversion at the last node. We place it at the end so that we can work on the wide dynamic range of the image before that shift. For the second step, we'll do our color correction. First, we will correct the white balance and adjust the exposure, contrast, and saturation. My waveform in the RGB parade tells me that my image is properly balanced with no over or underexposed areas. On this scene, I want to create a teal and orange blockbuster type look, and I'll help that along with a creative LUT. Here I'm using Bloodshot from our Cinematic V5 pack. This LUT is profiled for Rec. 709, so I apply it after my conversion node. And like I said before, we're using it as a final touch, so we're going to set it to about 50%. Now we can do our color grading, using the white balance to influence how the creative LUT looks. I'll push some warmth into the highlights, then reduce the saturation a bit. Correct skin tones here, then I'll add a bit of glow for a cinematic look. And that is a very simple process to quickly get a solid look for your project. And of course, we are using our LUTs here. If you are interested in finding out more about those, check the link in the notes below. We have all kinds of different looks, from horror to action and everything in between, so jump over to tryanddigital.com for more of that. But we're also going to be doing a lot more color grading tutorials, so if you aren't subscribed, consider doing that and hit the bell to be notified when we do put up new stuff. Until next time, don't forget to write, shoot, edit, repeat.

ai AI Insights
Summary

Generate a brief summary highlighting the main points of the transcript.

Generate
Title

Generate a concise and relevant title for the transcript based on the main themes and content discussed.

Generate
Keywords

Identify and highlight the key words or phrases most relevant to the content of the transcript.

Generate
Enter your query
Sentiments

Analyze the emotional tone of the transcript to determine whether the sentiment is positive, negative, or neutral.

Generate
Quizzes

Create interactive quizzes based on the content of the transcript to test comprehension or engage users.

Generate
{{ secondsToHumanTime(time) }}
Back
Forward
{{ Math.round(speed * 100) / 100 }}x
{{ secondsToHumanTime(duration) }}
close
New speaker
Add speaker
close
Edit speaker
Save changes
close
Share Transcript