In the last twenty years, the craft of color grading has found itself at the nexus of massive shifts in the technologies, demands, and aesthetics of motion imaging. These shifts have democratized its tools, elevated its visibility, and given rise to innovative new workflows and techniques. But some unfortunate side effects have accompanied all this positive change: color grading has evolved and fractured so rapidly that most filmmakers have an incomplete, conflicted, and often misinformed understanding of it. That’s where this series comes in: I’m going to provide you with a ground-up education on the core principles and practices of color grading, empowering you to craft the best images possible.
So where do we begin? Today, we’re going to focus exclusively on understanding the fundamentals of human vision. There are a few reasons this is well worth our time:
- It allows us to understand how best to use our eyes as colorists: Without a basic understanding of human vision, we can’t know the strengths and limitations of our eyes as tools. For example, did you know that the longer you look at a shot, the less ability you have to make an objective assessment of its white balance? Neither did I, until I learned about the adaptive nature of our vision — which we’ll return to later in this article.
- It provides us with information we can use to manipulate the viewer’s gaze in our grades: The human eye isn’t just the primary tool for our work: it’s also the sole consumer of it. Understanding the way our eyes see and process images maximizes our ability to control and manipulate it with our grading choices.
- It gives us the ideal foundation for understanding cameras and displays: Human vision is the basis of every imaging system ever devised, from lenses to sensors to displays. The best way to understand these systems, and the role color grading plays within them, is to understand their common foundation.
With these motivations in mind, we’re going to overview the vision system as a whole, and then explore some of its key strengths, limitations, and biases. If you’re ready to take your first step toward better-looking, better-informed color grading, read on.
The Vision System
Let’s start with a broad overview of our vision system. How do we form images from light?
Light strikes the objects in our environment, and any wavelengths not absorbed are reflected back to our eyes.
- Once light reaches the eye, the iris opens or closes the pupil to admit more or less light as needed.
- The lens focuses the admitted light, and projects the resulting image onto the retina.
- Through light-sensitive photoreceptors known as rods and cones, the retina converts the image into electrical impulses which are carried via the optic nerve to the visual cortex.
If you’re at all familiar with the mechanics of cinematography, this process should sound familiar, because it’s very similar to the way a camera works. This similarity is no coincidence, as both systems have the same fundamental purpose: converting light into images, which are then processed and stored. Of course, unlike with our vision system, a camera’s images must be reproduced in some fashion before we can view them, converting the stored images back into visible light via a display.
But we’re getting ahead of ourselves. In order to understand and effectively work with man-made imaging systems such as cameras and displays, we first need to go deeper in our exploration of our biological imaging system. Now that we have a basic understanding of the overall process of vision, let’s look at some of its key properties.
The Visible Spectrum
What exactly is light?
In geek-speak, light (more specifically, visible light) is the range of frequencies within the electromagnetic spectrum which our eyes are sensitive to.
In layman’s terms, visible light is a particular type of radiation we happen to be able to see. There’s lots of other measurable radiation out there, from radio waves to x-rays, but only visible light is, well, visible.
This is a key attribute of any imaging system, whether biological or man-made: each has an effective range of wavelengths which it’s capable of measuring, and anything outside that range is invisible to that system. One way of measuring and expressing this effective range in a given system is to compare it to that of human vision: the larger the percentage of visible light it can capture, the more robust the system. This is a concept we’ll be revisiting throughout this series.
In the same way that our eyes can only perceive a finite range of wavelengths of light, they’re also limited to a finite range of luminance values. We’ve all experienced what happens when the amount of light in our environment falls below or above this range: we’re no longer able to resolve images. The good news is that this range of perceivable luminance values, known as dynamic range, is extremely well-adapted in humans, boasting upwards of 30 f-stops — far more than even the best cameras currently available. So, as with the visible spectrum, we can assess how robust a given system is by comparing its dynamic range to that of our vision.
After looking at the range of wavelengths and luminance values the human eye is capable of perceiving, it seems we’ve evolved a near-perfect imaging system. But while our vision is indeed extraordinary, these metrics don’t tell the full story. To better understand our vision as it relates to color grading, we need to look at a few of the adaptations and “hacks” it relies on, each of which has a direct impact on the way we create, manipulate, and perceive images.
Rods vs. Cones
We learned in our overview of the human vision system that the retina is responsible for converting focused light into electrical impulses. It does this through the use of photoreceptors, which come in two main varieties: rods and cones.
Cones are responsible for detecting color, but they need a significant amount of light in order to function, and there are relatively few of them spread across the retina (around 6 million). This means that our effective visible spectrum becomes far smaller in low-lit environments. Think back to the last time you stood under the moon and stars without artificial light: you could probably see reasonably well, but could you discern any particularly vivid colors? Probably not, because your cones need a stronger stimulus to function.
Rods, on the other hand, are far greater in number (around 120 million) and can detect light at much lower levels — these are the photoreceptors which allow you to see by moonlight. The catch? You guessed it: rods can’t perceive color.
Why does this matter? Because it gives us important clues about how to prioritize the capture and manipulation of our images. Knowing that the eye is far more sensitive to overall luminance and contrast than it is to color means that, rather counterintuitively, the most important decisions we make when color grading may have nothing to do with color at all. This is one of the key concepts to understand in color grading: Contrast is king.
Imagine you’re pulling a late night in a fluorescent-lit office, and you hand off a blue binder to a co-worker on your way out. The following morning, you bump into your co-worker in the parking lot, and she’s carrying a stack of multi-colored binders. Not having a free hand, she asks you to grab the binder you loaned her. Will you have any trouble recognizing it by color? Unless there’s more than one blue binder, you’ll have no issue.
This is actually a pretty remarkable feat, as in each lighting environment, the wavelengths of light bouncing off that binder are wildly different. Yet our eyes pull it off with apparent ease, thanks to a quality known as chromatic adaptation. What this essentially means is that our eyes are constantly using environmental cues to determine what “white” is in a given situation — think of it as an ongoing automatic white balance.
But despite being a huge advantage for our ability to perceive everyday color, this quality has several critical implications for filmmakers and colorists:
- In production, we need to be constantly mindful of the fact that cameras don’t have this same adaptive mechanism, and take care to explicitly tell them what temperature of light to capture as “white”.
- When grading, we need to work in an environment with fixed lighting which is consistent with the white point of our mastering display. If we’re grading in a room with a window, for example, our eyes will compensate for the changing color of daylight pouring in, and our grades along with them, allowing color casts and inconsistencies to sneak in.
- We have a limited ability to make an absolute assessment of an image’s overall balance, because our eye will find the neutrals in the frame and do the balancing for us. And the longer we stay parked on the shot, the worse this problem becomes!
Chromatic adaptation is also one of the key reasons movies are shown in a fully blacked-out theater — we of course don’t want light sources which compete for the viewer’s attention, but we also need to ensure they’re not getting environmental cues which cause the eye to adapt to a different “white” than that of the screen.
When it comes to human vision, all colors are not created equal. There are certain objects and environments we observe so often that we retain a highly specific mental image of what they should look like. These are called memory colors, and they include things like foliage, skies, and, most importantly, skin. When we’re presented with images of these objects which don’t match our internal memory color, we’re subconsciously repelled. This is an adaptation that runs far deeper than our personal mental “database” of these colors — it’s a trait that’s been selected by evolution. For our ancestors, it meant the ability to find healthy food, sense impending weather changes, and select the ideal mate.
This means that there are colors which deserve more attention than others. Your audience may not know what the color of a bedroom wall should be, but they’ll spot the wrong hue or saturation of a memory color every time. Understanding memory colors and prioritizing them in color is a vital concept to mastering pleasing images.
We’ve now covered the key aspects of the human vision system we’ll be referring back to throughout this series. If you’re like me, you may find learning these principles to be challenging at first, but once absorbed, they’ll prove well worth your time. Studying these concepts at the outset of learning color is like studying music theory when you begin to play an instrument: it’s tempting to skip to the hands-on stuff, and you can probably develop some decent chops without the foundational knowledge. But in both cases, sooner or later your growth is going to hit a wall, and the only option at that point is to go back to basics and re-train yourself with the proper concepts. Trust me when I tell you from experience that it’s far faster and more pleasurable to make this investment the first time around!
Now that we’ve got a fundamental grasp on human vision, we’re ready to do a deep dive on cameras in part 2, where we’ll break down how they work, how they differ from human vision, and how we can successfully navigate these differences.