What Color Is My Hoodie?

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
Okay, I'm just going to address the elephant in the room as I start, yep, we are three months into Britain's third lockdown at the minute, uh, I desperately need a haircut, and there is a sideways wind here, so we're just going to have to deal with what's going on up here. According to the manufacturer, my hoodie is grey. That was what I bought. I picked the most generic colour possible. It's not going to clash with anything I film against, it's not going to blend into a green screen or a blue screen. Although to be fair, if I am filming under a blank grey sky, it can make the shot look a little dull. But if I am filming against a grey sky, and we compare the colors between the clouds and my hoodie... then this looks just a little bit blue. Or maybe purple? It looks a bit different in all those thumbnails. I mean, technically those aren't all the same hoodie, I've got several, but they're all identical. To explain the magical color-changing hoodie, we need to talk about how color perception works, and also about the history of laundry. First up: a brief run through what you probably already know from school. Electromagnetic radiation is a spectrum, that includes x-rays and microwaves and radio waves and visible light. The difference between those types of radiation is the wavelength. We have special cells called "cones" in the back of our eyes, three types of them, and each type reacts to a small set of different wavelengths. Which are, oh-so-conveniently, the wavelengths that the atmosphere easily lets through. The wavelengths for each type of cone overlap a bit, it's not precise, but it's roughly equivalent to red, green and blue. So if these cells notice radiation but these mostly don't, you see blue. If these and these do, but these don't? Magenta. And if all of them do, at just the right balance? Well, that'll be white, or grey. But you have to remember that color is not a real thing. Colors are experiences, the philosophical term is qualia, your brain just creates color in your head based on the signals it receives from those cells. Depending on the context, you can see the exact same wavelengths as different colors, even at the same time. There are a lot of optical illusions that play with that. Color perception isn't just about wavelengths, it's about comparing those wavelengths to everything around. A camera doesn't do that, as I shall now demonstrate with a white wall. It's daytime as I record this, and the camera is set up for the daytime light that's shining in. Bright sunshine. Now, there is an official, precise definition of "average sunlight", it's called Illuminant D65, it's meant to simulate the middle of the day in Western Europe, and if you'd like to spend literal hours going down a rabbit hole of color theory that I couldn't understand, you're welcome to use those keywords as a starting point to your research, but the basic idea is: this wall is white, and therefore, the colour that comes out of  your screen should also be white. But if I come back at night, lit by  the artificial lighting in the room, and I don't change the colour  settings on the camera... ...this looks very, very yellow. Because artificial light usually has less blue in it. Maybe that's because it's imitating old firelight and candlelight and it feels more comfortable somehow. Maybe it's just because that was how old light bulbs happened to work. That yellow-blue distinction is called "color temperature". Yellow and blue are complementary colors: remove one and you'll start to see more of the other. If you're working with cameras, color temperature is something you have to watch out for, because colors seen on a screen are out of context, the audience is used to seeing certain wavelengths from their screen as "white", no matter the lighting used in the room. So the video colors have to be adjusted to match the light, so this wall still looks white to you. I didn't change the lighting just then, only the color settings on my camera. Or, actually, in my editing program.   But if I take these new settings  back to daytime and sunlight... ...it looks very, very blue in here. To be clear, in person this is  always obviously a white wall. Your brain establishes color from context. Which is why my hoodie looks grey against something brightly colored, well, as brightly coloured as you can get in the British countryside when spring hasn't quite arrived yet, but against something that's also grey or white... ...maybe not so much. Because there's one other factor here: if your clothing is slightly yellow, it looks stained and tired. Because usually, slightly yellow clothing is stained and tired. There used to be a common product called "laundry bluing". It's mostly obsolete now with modern detergents, I actually couldn't find it in shops around where I live, but it added a hint of blue dye to your washing. If your white shirt is actually very slightly blue, it can trick the eye into thinking that it's cleaner and brighter and whiter than the actually-white shirt of the person you're competing with for that promotion. So maybe, if a company's selling a grey hoodie, their customers will look better and like the product a little bit more if it is ever so slightly blue. And as a personal choice, when I tweak the color temperature on my camera or in my editing program, I'm going to favor going just ever-so-slightly towards blue, partly for that reason, and partly because I like the way it looks. Now, I would call this hoodie "grey". Even if it might seem to have a bit of blue in it. It's all subjective anyway. But there is a way to get some sort of definitive, objective answer. A calibrated color sensor. It will light up my hoodie with very precise amounts of light, and measure the wavelengths that are coming back into its sensor. Which means it can give me calibrated color numbers, but it'll also match the color to the descriptive names that paint companies use. Which means that if I test it on this spare hoodie, the color is... ...Amethyst Sky. I'm not sure that's helpful.
Info
Channel: Tom Scott
Views: 3,831,707
Rating: undefined out of 5
Keywords:
Id: HeKuX063bcs
Channel Id: undefined
Length: 5min 39sec (339 seconds)
Published: Mon Apr 12 2021
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.