Owning a RED camera can be quite a treat; it can also be a royal pain in the butt. I would know because I currently own One. The camera has its ups and downs. I’m here to provide you with my knowledge of the camera, and what I strongly believe you should and should not do when it comes to owning this brand of camera and basically how it operates.
What is RAW? It’s the equivalent of a film negative in a digital format. It’s basically all of the information in the image being taken into the sensor as ones and zeroes; I’m talking about all that juicy detail from the bark on a tree to the pores in a human being’s skin. This is important because we, as photographers and cinematographers, are trying to recreate what’s in front of us to the best of our ability. Humans can’t read images in the form of ones and zeroes, so the image is translated into a digital intermediate called a Log image (Move to the next section for an explanation on Log images) .
This series of cameras captures RAW. While some cameras like Nikons and Canons capture RAW with multiple ISOs, the RED cameras do not. Try and think of the REDs as digital cinema cameras with a single film stock; changing your ISO value does not change the camera’s sensitivity to light that’s hitting the sensor; the camera’s sensor only becomes more or less sensitive to the light you’re feeding it; electrical signal boosts aka ISO do not exist for these cameras. For the Mysterium X sensors, the ISO is 320 for daylight (250 for Tungsten), and as for the Dragon, well the Dragon is ISO 250. I believe the original Mysterium sensor was ISO 100.
Linear Image Capturing and Logarithmic Data On The RED
Now, in the stage when the camera ingests the information, which is named the acquisition stage, the image is stored in a log curve. The middle curve in the image above is that curve. All important information is stored in the middle to preserve image detail. Highlights are stored in the shoulder of the curve, and the shadows reside in the toe. Once, in post, the image will look like a gray, contrast-lacking image that isn’t aesthetically pleasing to anyone. The colorist’s job is to take that Log image and turn it back into a Linear image for the post production pipeline. So, you capture the image linearly, bring it into post as a Log image , which is a preservation of image detail, and the image is finally transformed back into a linear image.
Images on today’s digital sensors are captured in a linear fashion; there’s a ceiling for the highlights and a floor for the shadows. If you hit your highlight ceiling, you’ll blow out your highlights; if you blow out your highlights while recording your image, they will become irrecoverable. And, if you crush your shadows, hitting your shadow floor, you will be recovering image detail from your image in post with a lot of noise or fixed pattern noise in the shadows, which is not what you want. That’s why it’s your job as the photographer or cinematographer to capture the most well-exposed image your can, so you don’t have such a rough time in post trying to clean up your mistakes. Underexposure can be fixed with Neat Video, but at the cost of more post production time, but lost highlights are gone forever.
The Bayer Sensor behind the RED Cameras
While each iteration of the RED may have a differently sized sensor, it doesn’t change how the chip has been built for optimal sharpness and flexibility in post. The RED’s sensor is silicon-based sensor. Pretty much all technology has some form of silicon incorporated into its structural design due to silicon’s abundance. How it effects the RED camera is that silicon is very sensitive to warm light and least sensitive to daylight. This basically means that when warm light strikes the sensor, it results in quicker image blowout and exposure loss for the camera because the blue channels are being neglected.
The sensor itself is a Bayer design. Basically, all the photosites on the sensor are divided into quads that include two green photosites, one red and a blue photosite. The reason for two green photosites is to mimic the human eye’s sensitivity to green waves of light, which is what it’s most sensitive to. Therefore, the doubling of the green photosites make the image appear sharper to our eyes. Fun fact: the human eye is least sensitive to the color blue, which means we humans have a heavy resistance to that wavelength of light, due to evolution; the atmosphere is blue, so our eyes are very resistant to that color.
The Red Piece of Glass In Front of The Sensor Aka The OLPF
That red piece of glass in front of the bayer sensor is called an optical low pass filter (OLPF). This filter softens the image so the sensor does not create aliasing artifacts. In other words, the sensor is so sharp that it needs to be softened in order to prevent moire (the wavy stuff in the image), which can never be removed from the image once it has been introduced.
Metadata And How It Pertains To The RED Cameras
Metadata is really important for everyone involved in post production, and is basically a footprint that the cinematographer leaves behind for the colorist and editors; the colorist can look at the footage in REDCINE-X PRO and see exactly what went wrong with your exposure. It tells the post production crew what aperture, ISO and shutter speed you shot at. It even shows the exact timecode, which can help the DIT sync the picture even more easily. The great part about the metadata that comes from the RED is that it is changeable in post. For cinematographers whom are used to physical celluloid, this is really bad because it essentially allows the colorist to change whatever he or she wants about the image… except for physical lighting, aperture and focus that is. We, the cinematographers, capture the image and the colorist further influences it in post. Instead of a one-step process, the captured image is now truly being divided into two main stages: production and post production. Pre production is obviously important as well, but this article is mainly focusing on how the camera operates.
Why The Mysterium X Sensor is Rated At ISO 800 By RED Digital Cinema
Now that you know the Mysterium X sensors has a film stock equivalency of ISO 320, you may be questioning why RED recommends ISO 800 as a starting point for your filming. Well, here’s the thing; ISO 800 doesn’t actually exist on the camera. Think about if you were using a film stock, and it was ISO 320; shooting for ISO 800 would be equivalent to underexposing the stock. When it’s developed in post, the stock will have to be pulled up 1 and 1/3 stops to compensate for the underexposure. In turn, the grains in stock will be pulled up and and more prevalent in the image. The same is applies to the RED One MX camera. Here’s the exposure theory.
When looking at the camera in “ISO 800”, the user is tricking him or herself into thinking the image is more overexposed than it actually is; in turn, you’ll end up underexposing the image to compensate for the blowout that he or she is seeing on the monitor; if you’re exposing via the monitor, you’re in real trouble. By underexposing the image, you’re providing the camera with one extra stop of over exposure latitude to work with. But, as a trade off, you’re underexposing the shadows, which in my opinion, is really unfair because you gain a lot more image noise in your image when you’re pulling it up in post, and your goal is to capture a noiseless, clean image… with a little noise sprinkled in for texture. I personally believe it’s a useless tool that only confuses the user and makes him or her want to straight-up abandon the RED; I know I did at first.
You MUST Hit The Camera’s Sweet Spot
No silicon-based, digital sensor likes to be underexposed, even the highly acclaimed Alexa. While the Alexa loves to be overexposed because its roll-off is just lovely, the RED does not. The Mysterim X sensor needs to be exposed perfectly. It can’t be slammed to the left or the right. It’s a very sensitive camera that needs to be lit just right, which takes time, and that may not work for everyone, especially those who need to work quickly. These days, many are seeking a low-light camera that will provide them with good latitude. The RED One, Epic, Scarlet and Dragon will not be very helpful in lowlight unless you bring the image up in post and noise reduce, which is very time consuming in my opinion.
How To Avoid Under or Overexposing The RED
While I would normally recommend a light meter, I would have to advise against that method for the RED unless you’re using it as a loose reference or taking contrast ratio measurements for your subject and the set around you. The reason being is that the camera and light meter don’t always seem to one-hundred-percent agree with one another when it comes to exposure; this applies to the Alexa as well. So, which tools should you use to judge the camera’s exposure?
Use The In-Camera Tools
The RGB histogram on the bottom left of your RED Monitor is subject to change due to the multiple “ISO” settings you choose; trust it only when you’re viewing it in RAW mode. Always expose for a full histogram. Slamming it to the left will lead to underexposure, and slamming to the right will end up in overexposure.
The goal posts, the two posts on the left and right ends of the histogram, are highly trustworthy because they are directly tied to the sensor itself. If the left goal post is filled up with red then you are underexposing the camera, and if the right post is being filled with red then overexposure is occurring.
There is one tool that I highly suggest you use in combination with the goalposts; that tool is the false color mode. This tool shows your blown out highlights as red and your underexposed shadows as purple; trust that tool with your life because it’s connected directly to the sensor. It’s telling you what the camera needs. So, once again, trust it.
The best Light Sources To Use With The RED
Now that you know the camera’s sensor is a 5,000-degree-kelvin sensor, you should be guessing that the camera’s sensor works best under daylight sources. If you guessed that then you are correct. It doesn’t matter if you use Kinos, HMIs, Tungsten with Full CTB, LEDs or even THE SUN; as long as they source is balanced to daylight and you’re filling the RAW histogram, you’ll be providing the sensor with all of the light that it needs to attain that beautiful image that you’ve always wanted, and you’ll end up making friends with you DIT.
Tungsten will only serve to underexpose your camera, and give you muddy image quality because there’s no blue in tungsten, and when that happens, the sensor boosts its blue channel in order to compensate for the lack of blue leading to unwanted image noise because the camera is being underexposed. Underexposure and image noise can be more easily avoided by simply using daylight-based sources.
To sum it up, only trust your RAW view. Use the in-camera tools to gauge your exposure, and do NOT expose your image just by looking at the monitor. Finally, if the image is noisy or not the way you want it, it’s YOUR fault and not the camera’s. It is your job to understand this camera to the best of your ability. Know your camera’s strengths and limitations, or you as an operator are useless.