Whether you’re a student or a seasoned veteran, selecting the right camera can make your head spin with questions. Does sensor size matter? What’s the difference between 4:2:0 and 4:2:2? Do I really need a camera that can record 4K? What about raw? We’ll explain the key components that help a camera produce cinematic images, and in doing so, hopefully help you make that all-important decision.
Evaluate Your Needs and Your Budget
Before considering certain cameras, you must know your needs and your budget. And be realistic. If you are a student or relatively new to filmmaking, consider starting with a more affordable option, rather than jumping into the deep end with something that may be overkill for your needs. Fortunately, obtaining cinematic results has become increasingly affordable, thanks to the rise in video capabilities of DSLRs and mirrorless cameras, and the decrease in price of larger, professional cameras.
You also need to be wary of hidden costs. The price of a camera might look great at first, but what if it needs a separately available electronic viewfinder and shoulder rig to function the way you need it to, or an external recorder to get the highest image quality? Let’s not forget about lenses, extra batteries, and media cards. In short, before you make your purchase, take a step back and consider the whole package. The last thing you want to do is spend your savings on a camera that will only serve as an expensive paperweight until you can afford to outfit it properly.
The Cinematic Look
It helps to look at the big boys-the top-of-the-line cameras being used to make major motion pictures. Why are they being selected? What do they offer that make directors, DPs, and producers pick them over others for their productions? When reading interviews, you’ll often discover it’s because the camera could produce cinematic images with high dynamic range that feel “filmic.” When discussing different aspects of digital cinema cameras, you’ll find me constantly referring to film as the reference point.
So, what does it mean to look “cinematic?” As a moviegoer, your eyes and brain have been trained over the years to know how a film should look. You also know-perhaps even more so-when something does not look cinematic. While professional set design, lighting, and shot selection play a significant role in achieving the cinematic look, many camera factors contribute, as well, including image-sensor size, resolution, dynamic range, and color sampling. These are some of the major factors that you’ll need to consider, along with more practical considerations, such as recording formats, connectivity, and form factor.
One of the fundamental aspects of cinematography is lens selection, so this should be one of the first factors you consider when selecting a camera. There are several standard lens mounts, including native DSLR mounts, such as Canon EF and Nikon F; mirrorless camera mounts, such as Micro Four Thirds or Sony E; and the PL mount found on high-end cinema cameras. Some mounts are more adaptable than others. For example, the short flange focal distance of mirrorless camera mounts makes them easily adaptable to other mount types using readily available adapter rings. This opens a seemingly endless selection of vintage and modern lenses, giving you the flexibility and freedom to select lenses based on your budget and the look you want.
While mirrorless camera mounts are more adaptable, this doesn’t mean you should avoid cameras with DSLR mounts. Today, many cine-style lenses are available for DSLR systems. For example, Zeiss CP.3 Compact Primes are available for Canon EF, Nikon F, Sony E, and MFT mounts, in addition to PL. DSLR systems also have a rich selection of native lenses, though these aren’t always the best choice for cinema use, due to the lack of manual aperture rings and electronic or focus-by-wire focusing systems that make them difficult to pull focus with and impossible-to-do smooth iris pulls.
Zeiss Interchangeable Mount Set EF
The PL mount is the standard for professional cinema cameras and lenses, though both tend to be on the expensive side. If it’s within your budget, or if you plan on renting lenses for each production, then there is nothing wrong with going the PL route. Some cinema cameras have mounts that can be swapped out, so you’re not roped into a single choice. Knowing the lenses you have available or plan on purchasing will play a major part in selecting a camera.
The image sensor is the heart of the digital camera. When talking about sensor size, many people will tell you that bigger is better. In many ways, they are right, though size isn’t everything. In the still photography world, the “standard” is referred to as “full frame,” which is roughly the same size as an image captured on 35mm film; anything smaller than that size is often described as having a “crop factor,” though this is a term of which I’m not overly fond.
In the cinema world, the standard sensor size has developed around the Super 35 (3-perf 35mm) frame, which has a crop factor of approximately 1.5x compared to full frame. For people coming from the stills world, this is about the same size as an APS-C sensor capturing 16:9 video. Part of the reason that I don’t like the term “crop factor” is that for most cinematographers, the motion picture 35mm frame is their standard, not full-frame. However, with digital sensors replacing 35mm film and full-frame cameras being used increasingly for video work, the distinction is becoming more of a gray area, so crop factor does, at the very least, give a standard reference point from which to compare sensor sizes.
Angle of View and Aesthetics
So why does the size of the sensor matter? Sensor size directly influences many imaging characteristics, including lens compatibility, angle of view at a given focal length, and apparent depth of field. All lenses project an image circle onto the sensor large enough to cover the format for which they are designed; a lens designed for a full-frame camera will cover full-frame sensors, an APS-C lens will cover APS-C sized sensors, and so on. Lenses will also cover sensors smaller than the format for which they are designed, but if you try to use them on larger sensors, then you may see vignetting or even the edge of the image circle.
The smaller the sensor size, the smaller the portion of the image circle is captured, which means that the angle of view-or width of the scene that is captured-decreases. For example, a 50mm lens on the full-frame sensor of a Canon 5D Mark IV will have about a 40-degree angle of view, while the same lens on the Super 35 sensor of the Sony F5 will only be around 25 degrees. To get the equivalent angle of view of the 5D, the F5 would need to use a lens around 32mm. Comparing or trying to match angle of view between two cameras with different sensors is where thinking in terms of crop factor can come in handy.
APS C Sensor
A “normal” lens is one that, for any given format, renders an image with perspective like the way you see the world. The general formula says that a “normal” lens has a focal length like the diagonal measurement of the sensor; on Super 16 this is going to be smaller than a normal lens on Super 35, which will be smaller still than a normal lens on full frame. The same can be said for wide-angle and telephoto lenses. This gives different formats/sensor sizes their own aesthetic-even though the angle of view changes, the perspective and how linear space is distorted does not.
Depth of Field
For many people, a chief benefit of a larger sensor is the ability to achieve shallower depth of field. Frame a shot with a lens at f/2.8 with a full-frame camera and it will appear shallower than a shot with the same lens, framing, and aperture on a Super 35 sensor. Being able to separate your subject from the background using shallow depth of field is a definite component to achieving a cinematic look. So, bigger is better then, right? Well, not necessarily. You can still achieve cinematic shallow depth of field on Super 35, Micro Four Thirds, and even Super 16 formats. Also, the shallower the depth of field is, the harder it is to keep a subject in focus during the shot, even for an experienced focus puller. Workable depth of field might mean stopping down to f/4 or even f/5.6 on full frame and losing some stops of light, while a Super 35 camera might be fine at f/2.8.
Having shot with Super 16, Micro Four Thirds, Super 35, and full-frame sensor sizes, I tend to find Super 35 to suit the way I shoot best, while giving me that classic cinematic aesthetic. It also opens a wider selection of lenses, including modern and vintage cinema lenses that won’t cover full-frame sensors. However, full frame is becoming more common in the video world and many cinema lenses, as well as re-housed DSLR lenses that cover full-frame sensors are becoming available to accommodate this trend.
The dynamic range of a camera is the range in luminance it can capture; essentially, how much shadow and highlight detail can be reproduced before hitting pure black or pure white. For many years, dynamic range was one of the main areas in which digital sensors fell short when compared to film. Think of all the digital footage you’ve seen where a window or sky was blown out and just looked white, with little to no information present. Now picture that same scene in a movie shot on film where there is still color in the sky and details through the window. Proper lighting and working within the useable range of a camera certainly helps, but the limitation of older digital cameras was plain to see.
Cameras with a wide dynamic range can hold onto more detail in the shadows and highlights. The image on the left shows what an image with wide dynamic range would look like and the image on the right shows what the same image would look like with narrow dynamic range
A wide dynamic range is one of the key components of the cinematic look. Modern 35mm motion picture film has upwards of 14 stops of dynamic range. While only the top-of-the-line digital cinema cameras tend to approach film in this regard, more affordable options, such as the Sony FS7, FS7 II, Canon Cinema EOS line, Blackmagic Design cameras, and even mirrorless cameras, such as the Sony a7S II and a7RIII, offer large dynamic ranges.
To capture the most dynamic range, look for cameras that offer “log” gamma curves. Rather than baking-in a contrasty, finished look in-camera, log gamma curves yield a flatter image that preserves more information in the shadows and highlights, and thus a larger dynamic range. This gives you-or the colorist-the ability to create the final look during post production and preserve the shadow and highlight details that would otherwise have been lost had it not been recorded using a log gamma curve. A flatter image also gives you greater control to adjust colors in post to make corrections and define the creative look of your movie.
Not each log gamma is the same, with each company offering its own flavor. For instance, Sony has its S-Log2 and S-Log3 gammas, Canon has C-log, Panasonic has V-Log and V-Log L, ARRI has Log C, and so on. Log gammas require experience to color-grade properly, so take that into consideration when purchasing a camera with a log gamma option. A good place to start would be to use a pre-built 3D LUT (Look Up Table) within your color-grading application to bring contrast and saturation back into your image and transform to the desired output color space, such as Rec. 709. From there, you can tweak your image and, in time, grade from scratch or even build your own custom LUTs. There is also a variety of available LUT packages and plug-ins that will mimic the look of popular film stocks, which can help give your images an even more cinematic quality.
When talking about resolution, we need to consider both the resolution of the recorded video file and the resolution of the sensor. Many professional cinema cameras will have sensors with resolutions that match the highest-resolution video that the camera can record. The Sony F5 and F55, for example, use an 8.9MP sensor that allows them to natively capture 4K (4096 x 2160) or Ultra HD (3840 x 2160) video. On the other hand, DSLRs and cameras designed for still photography can have sensors with resolutions of 24MP or higher. To record 4K or 1080p from those sensors, they often skip entire rows of pixels (line skipping), combine the information of adjacent pixels together to create a single pixel (pixel binning), or both. This can have a negative impact on the image, including the introduction of moir and aliasing, depending on the camera.
Not all DSLRs and mirrorless cameras resort to line skipping or pixel binning. The Sony a7S, for example, has a 12MP sensor that, in video mode, delivers full-pixel readout for Ultra HD (3840 x 2160) recording. In 1080p mode, it downscales the full UHD image to 1080p, doing binning on the image-processor level rather than on the sensor, for improved results. Other cameras, like the Panasonic GH4, use a crop of the sensor for native 4K resolution while even the Sony a7RII, with its 42MP sensor, offers a Super 35 crop mode with full pixel readout that is then downscaled to 4K, with excellent results. Please note that the Panasonic GH5 and GH5S do not crop when shooting 4K video.
When determining your resolution needs, you need to consider what your deliver market is. If your videos are just going up online, then you may not need anything higher than 1080p. Even if your production will have a theatrical release, you may not need to make the jump to 4K. The ARRI Alexa, a Hollywood staple, (when in 16:9 mode) only records 2.7K-resolution video (2880 x 1620), and the footage looks good and sharp when projected in theaters. Yet with the slow shift to Ultra HD televisions (3840 x 2160) and content underway, purchasing a camera with 4K resolution may help future-proof your investment, as well as make you more marketable should you decide to rent out your services or camera package.
Even if you’ll be delivering in 2K/1080p, purchasing a camera than can record 4K/UHD is a smart idea, as downscaling your 4K video to 2K almost always gives you sharper footage than if you recorded 2K in-camera. It also gives you a 4K master that you can archive and then export in full resolution in the future, should you need to. Additionally, you’ll find that many of the top video cameras these days offer 4K recording, so even if you only plan on shooting 2K/1080p, the camera will be ready for 4K when you are.
Resolution isn’t the entire story, and the ARRI ALEXA and AMIRA have been capturing perfectly acceptable images for years, at less than 4K resolution, that are shown at 4K. The ALEXA Mini is a versatile and popular camera at the high end of the Digital Cine platform. Still, some are a slave to the need for 4K origination, no matter how good the images may look, so ARRI has brought out the ALEXA LF, which meets and exceeds the 4K requirements of almost every distribution platform.
Sensitivity / Low-Light Performance
An important factor for many shooters may be the low-light performance of a camera. For larger productions with a full light package and crew, this might not be as important as it would be for a student, amateur, or indie filmmaker who plans on shooting with small packages or mostly natural light. A contributing factor is pixel size. Generally, the larger the pixel, the better it is at collecting light and reducing image noise. Thus, larger-sized sensors and sensors with lower pixel density tend to perform better in low light.
There are other factors, as well, such as sensor design and internal image processing and noise reduction. One technology that is starting to make its way into larger-sensor cameras is back side illumination. This doesn’t mean that there is a backlight on the sensor, but rather that the sensor wiring is placed behind the photodiode substrate rather than in front of it, to improve light-capturing capabilities. The technology is seen in the Sony a7R II and, combined the with camera’s full-frame sensor, allows it to have excellent low-light performance despite its 42MP sensor.
While we’re on the topic of sensitivity, it is worth mentioning that most cameras have a base or native ISO at which they perform best when recording with a log gamma; i.e., the ISO where they have the most dynamic range. For many cameras, like the Canon C300 Mark II, this is ISO 800. For other cameras, it’s even higher. The Sony F55 has a native ISO of 1250, the F5 ISO 2000, and the compact a7S II a staggering ISO 3200 (the minimum required to record using S-Log2).
For optimal performance, many DPs will keep cameras at their native ISO and adjust the lighting accordingly. When shooting outside, this often requires the use of ND (neutral density) filters to reduce the amount of light hitting the sensor, whether it is a variable screw-on filter, square filters in a matte box, or built-in filters if the camera has that feature.
Global Shutter versus Rolling Shutter
Another consideration is whether the camera uses a global shutter or a rolling shutter. Cameras that use a global shutter capture all the sensor information at once, like a film camera with its mechanical rotary disc shutter, albeit electronically. Cameras that use a rolling shutter scan across the sensor line by line rather than all at once. Part of the theory behind this method is that the sensor can continue to collect photons as the process happens, thus increasing the sensitivity. The downside is that footage can take on a noticeable wobble (commonly referred to as the “jello effect”) and motion skews with fast-moving subjects and camera movements due to the scene changing faster than the camera can scan through the pixels. The effect isn’t very pleasing, and various cameras’ rolling shutters can be a deal-breaker for some moviemakers.
Even many professional cinema cameras, including the ARRI Alexa, use a rolling shutter, but they use readout speeds of the sensor, effectively eliminating rolling shutter side effects in all but extreme conditions. DSLRs and mirrorless cameras tend to be some of the worst culprits-with some worse than others-so if fast-moving motion and handheld footage is part of your shooting style, you should do your research beforehand on the camera you’re considering. If you can, get your hands on the camera for a weekend and see if the camera’s rolling shutter is acceptable for your shooting style. Other cameras, such as the Sony F55 and Blackmagic URSA Mini employ electronic global shutters and some, such as Sony’s flagship F65, even use a mechanical rotary disc for its global shutter.
When examining a camera’s recording capabilities, there are a lot of things to consider beyond the resolution, including frame rates, codec, bit rate, color depth, and color sampling. Additionally, many cameras offer video outputs that render higher image quality when sent to an external recorder than can be recorded internally. Let’s look at frame rates.
It probably goes without saying that you’ll want a camera that can record either 24p, which is the frame rate at which motion picture films are shot in the United States, or the 23.98 (23.976, to be precise) video equivalent. Pretty much any camera these days offers this frame rate, as well as the 29.97p frame rate for television. If you’re in the UK or a PAL region country, you’ll want a camera that can shoot 25p.
In addition to these standard frame rates, if you plan on doing any slow-motion work, then you’ll want a camera than can shoot at higher frame rates, such as 60p, 120p, or 240p, depending on how slow you want to go. Often, the highest frame rates are offered at lower resolutions than a camera’s maximum resolution, or only in short bursts. For example, the Sony FS7 can record 4K/UHD up to 60p and 1920 x 1080 up to 180p, while the FS700 can record 1920 x 1080 up to 240p, but only in 8-second bursts.
Codecs and Bit Rates
Video captured by cameras is recorded internally to media cards using a variety of codecs. A codec compresses digital information to preserve storage space and makes it easier to play back quickly. Codecs can range from highly compressed with low bit rates (or data rates), such as AVCHD, to more robust, less lossy codecs, such as Panasonic’s AVC-Intra and Sony’s XAVC. The higher the bit rate, the more information is being recorded and, generally, the higher the image quality. With higher bit rates also come larger file sizes and the need for more computer resources to play back and edit the video.
There are two main categories of video compression to consider: interframe and intraframe. Interframe, sometimes referred to as long GOP (group of pictures) contains a set of grouped frames, or “key frames,” which reference each other to produce an image, resulting in smaller file sizes. With intraframe compression, each frame stands alone and contains all its own information. While intraframe files are larger, they also require less processing power to play back and edit, as the computer doesn’t need to look at a group of pictures all at once to display a single frame. Intraframe codecs tend to be used for higher bit-rate recording and are generally preferred for video origination, if available. Popular intermediate/editing codecs like Apple ProRes, Avid DNxHD, and Cineform are all intraframe.
Color Bit Depth
The camera’s ability to reproduce colors with smooth gradation is very important to creating filmic images. The number of possible colors a camera can record is determined by the color depth, or bit depth. In 8-bit video files, there are 8 bits of data for each color channel, resulting in 256 shades of red, of green, and of blue, for a total of 16.7 million possible colors. This may sound like a lot, but this still yields color banding in areas of gradation in tone. But, seeing as most consumer monitors and HDTVs are 8-bit, you can usually get away with cameras that record 8-bit.
Although 8-bit is fine for a lot of applications, if you’re going to be doing heavy color-grading work or compositing work, then you’ll want a camera that offers 10-bit recording, which yields 1,024 shades of red, of green, and of blue for a whopping 1.07 billion possible colors. Naturally, your file sizes will be much larger than with 8-bit, but you’ll have more color information to work with. Beyond 10-bit, you’ll also find 12-bit codecs in select high-end models, as well as 12-bit, 14-bit, and 16-bit raw formats.
Along with bit depth, you’ll also need to consider color precision, which relates to how many samples of each color channel are taken; the higher the sampling rate, the higher the precision. Without getting too technical, every pixel on a sensor registers a luminance value (brightness) and a chrominance value (color). Chroma sub-sampling is a process in which the luma value of each pixel is preserved, while groups of pixels are forced to share a same chroma sample, effectively becoming the same color.
A chroma sub-sampling of 4:4:4 means that there is no sub-sampling happening, the luminance and chroma information is recorded for each pixel, but this is only available in the highest-end camera models or via an external recorder, if the camera can output that signal (more on external recording later). 4:2:2 is a common subsample where every two pixels share a chroma sample while luminance is recorded for every pixel, resulting is plenty of color precision for most applications. Most consumer cameras, such as DSLRs and mirrorless cameras, use 4:2:0 sub-sampling, which throws away even more color information that makes it difficult to do chroma key work and can result in blocky artifacts.
If you want the most flexibility to adjust your images during post production, then you may want to consider cameras with raw recording capabilities. Raw is different from log gamma and uncompressed video in that it’s not really a video file; it is, as the name suggests, raw data. Much like a still photo shot in raw format, raw video doesn’t bake in any in-camera decisions, allowing you to adjust things like white balance and exposure in post. Raw data is usually at a higher bit depth, as well, between 12-bit and 16-bit, and often with 4:4:4 sub-sampling, giving you the most color information the camera offers.
Raw can be both uncompressed and compressed, and in proprietary or open source formats, such as CinemaDNG. Compressing raw may seem counterintuitive, but raw data already takes up a lot of space, so compressing it helps. Some cameras claim to have visually lossless compression, while other cameras let you choose how much compression is being applied, such as 3:1 to 18:1 on RED cameras. To be viewed, raw files have to first go through a de-bayer process to be converted to video. The video is usually converted to a log gamma format, allowing you to use similar LUTs that you use on log footage.
A raw workflow can be very costly and, unless you’re shooting a major motion picture or TV commercial, a simple log gamma recording may suffice. The amount of space that raw recordings take up is enormous, meaning you’ll eat through media cards and need plenty of storage drives, not to mention a computer capable of handling the footage. RED Digital Cinema offers compressed raw recording, which can save you a tremendous amount of storage space, while still providing the control in post that you get with raw recording.
If you want raw then there are certainly affordable cameras available, including the Blackmagic Pocket Cinema Camera 4K, but you should really have a compelling reason to shoot raw before jumping into it.
To record the highest quality with DSLRs and mirrorless cameras, it often means splurging for an external recorder. Many of these cameras offer uncompressed video output over HDMI, allowing you to bypass internal compression and low-bit-rate codecs in favor of higher-quality and edit-friendly codecs, such as Apple ProRes and Avid DNxHD. In addition to uncompressed video, some cameras that can only record 8-bit 4:2:0 video internally can output 8-bit 4:2:2 or even 10-bit 4:2:2 in some cases, such as the Panasonic GH4. For Sony’s a7S, its HDMI output provides the only way to record UHD video, making an external recorder a requirement for many shooters.
Not only DSLR and mirrorless cameras benefit from an external recorder. Take the Sony FS7 or FS7II, for instance. These cameras are able to output 4K and 2K 10-bit raw over SDI to Sony’s AXS-R7 Recorder or to Convergent Design’s Odyssey 7Q+ monitor/recorder via the XDCA-FS7 Extension Unit. The Sony FS5 with the RAW Output firmware update also benefits from an external recorder that can record RAW. Even for cameras that record raw or high-resolution and high-bit-rate video internally, there is still a benefit to using an external recorder, as they often record to fast and relatively inexpensive SSDs rather than costlier media, such as SxS or CFast 2.0 cards. Before purchasing a camera, be sure to note whether it would benefit from an external recorder, and if the added cost of one is worth the benefits that it brings.
Apple announced, in 2018, two flavors of ProRes Raw, available in conjunction with the Atomos Sumo 19 and Shogun Inferno recorders. At this writing, ProRes RAW recording is capable from the Sony FS5, FS5II, FS7, FS7II, Panasonic EVA1, Varicam, Varicam LT, Canon C300 Mark II, C500, and C700. Look for more cameras to join the list in the future.
Form Factor and Connectivity
While we have focused mostly on image quality, the form factor of a camera and how it handles might be a determining factor for you. Cameras come in a variety of shapes and sizes, each with its own pros and cons. Smaller cameras are easy to pack and travel with, but lack many of the conveniences of larger, professional video cameras, especially in terms of connectivity.
On the opposite side, larger cameras are just that-larger and heavier, and often require expensive accessories to get up and running. Even after outfitting a DSLR with a cage, 15mm rod support system, external audio recorder or adapter box, and even an external power solution, chances are that it will still be cheaper than many of the professional models available. Additionally, you always must be able to strip the rig down and modify it as needed. However, there is a reason professional cameras are larger and more expensive, as they are packed with higher-quality recording options and video-specific controls and features.
Ultimately, you’ll have to weigh the pros and cons of each camera available and decide for yourself which features are important to you and which you are willing to compromise on to stay within your budget. I hope that this article gives you some food for thought and helps you to select the camera that is right for you.