A Brief History of Full Frame
"Full Frame" has become one of the most popular ways to reference an acquisition format that has an area that is roughly 36mm x 24mm. This is the same dimension as 35mm still photography film as well as the digital sensors in DSLRs like the Canon 5D, or mirrorless cameras like the Sony a7s. There are also a host of digital still photography cameras that have image sensors smaller than 36mm x 24mm. Formats like Canon's "APS-C" and Nikon's "DX" have sensors that measure 25.1mm ×16.7mm, and 24mm x 16mm respectively. These smaller sensors are often referred to as "cropped sensors" and the larger 36mm x 24mm sensors are therefore called "full frame" sensors.
Comparing the image area and field of view differences between a "Full Frame" digital sensor and an APS-C "cropped" digital sensor
The term "Large Format" has recently been popularized thanks to ARRI's introduction of the ARRI Alexa LF and Alexa Mini LF which both have a sensor that measures 36.7mm x 25.54mm. RED resurrected the name VistaVision, calling their flagship camera the "RED Monsto VV" which has a sensor measuring 40.96mm x 21.60 mm, which is a wider but not quite as tall as Full Frame. All these names, sensors, and formats can make things confusing. So here's a brief introduction.
ARRI Alexa LF - camera used for The Ultimate Anamorphic Lens Test
In the 1920s the 35mm motion picture film format had been in use for years, with the 35mm film running vertically through the camera, just as it still does today through cameras like the ARRIFLEX 235 and Panavision Millennium motion picture film camera. In those modern film cameras, the standard Super 35mm film gate measures 24.89mm × 18.66mm. Oskar Barnack while working for Leica took 35mm motion picture film and rotated it 90 degrees so that it would run through a camera horizontally, and created a much larger film gate that was 36mm wide by 24mm tall, which gave us the 135 still photography format which is still the standard for photography today. The creation of larger film formats for motion picture capture began decades later, but in a similar way.
The introduction of larger film formats came from a need to get people to buy more movie tickets, especially with the growing popularity of television in the 1950s. One such format was Cinerama which captured images using 3 separate film cameras stacked side-by-side-by-side and then projected using 3 separate film projectors, which naturally gave the audience 3 times the detail and allowed for larger screens to project onto. Anamorphic processes like Cinemascope allowed a widescreen format captured using specially designed optics to squeeze an image onto a larger 4-perf negative, which was then de-squeezed for theatrical release.
In 1952 the Todd AO system used 70mm film with a film gate of 48.56mm x 20.73mm. Paramount took a different approach and in 1954 developed a format called VistaVision which utilized 35mm motion picture film running through the camera horizontally instead of vertically, resulting in a film gate that was 37.72mm x 24.92mm. Both formats' use of a much larger negative resulted in images that had a higher resolution and finer film grain.
The introduction of larger film formats came from a need to get people to buy more movie tickets, especially with the growing popularity of television in the 1950s. One such format was Cinerama which captured images using 3 separate film cameras stacked side-by-side-by-side and then projected using 3 separate film projectors, which naturally gave the audience 3 times the detail and allowed for larger screens to project onto. Anamorphic processes like Cinemascope allowed a widescreen format captured using specially designed optics to squeeze an image onto a larger 4-perf negative, which was then de-squeezed for theatrical release.
In 1952 the Todd AO system used 70mm film with a film gate of 48.56mm x 20.73mm. Paramount took a different approach and in 1954 developed a format called VistaVision which utilized 35mm motion picture film running through the camera horizontally instead of vertically, resulting in a film gate that was 37.72mm x 24.92mm. Both formats' use of a much larger negative resulted in images that had a higher resolution and finer film grain.
Oskar Barnack
Full Frame Lenses
With the birth of these larger film acquisition formats, came the need for lenses that could cover the larger negative. New lenses were designed, and in some cases still photography lenses were repurposed, and often rehoused, or at least re-mounted to work with these new film cameras. Here we are 60+ years later, and we are in the same exact position. New Full Frame motion picture cameras are available including the ARRI Alexa LF and Mini LF, RED Monstro, Sony Venice and FX-9, and the Canon C500 MKII, to name a few. These cameras need lenses to cover their larger sensors. Industry standard lenses like Zeiss Master Primes, Cooke S4s, Zeiss Super speeds, were not designed to cover such a large format. These lenses , in particular the wider focal lengths cannot completely illuminate Full Frame sensors. The results are images with heavy vignette in the corners, and even a hard porthole effect.
Using lenses designed for different formats on a Full Frame Sensor
Today, all major lens manufacturers are offering new lens designs that cover these larger formats including Zeiss Supreme primes, Cooke S7i primes, and ARRI Signature Primes. For those looking for lenses with more vintage character there are many older still photography lenses that are being rehoused by companies like Zero Optik, True Lens Services, and GL Optics. Since these still photo lenses were designed to cover 36mm x 24mm still photography film, they are perfectly capable wen being paired with modern Full Frame digital sensors. By now you may have seen someone pairing vintage Leica R primes, Canon FD primes, or Nikon AI-S primes with modern cameras. This new era of Full Frame digital motion picture cameras take full advantage of the larger image circle of these vintage full frame primes.
Comparing 35mm still photography film (left) to 4-Perf Super 35mm film (right)
Elvis Presley with a VistaVision Camera in 1957
The Full Frame Look
Where you will see the biggest aesthetic differences between images captured with various sensor sizes are the magnification of film grain, (and the magnification of a digital sensor's noise), and the ability to blur backgrounds relative to the subject of an image.
One of the simplest differences to understand is the magnification of the recorded image. When we finally show our recorded images to an audience, they are magnified from their original size (the size of the sensor or of the film used to capture the images). Whether the images are being watched on a cell phone or an IMAX theater screen, the original recorded images will be magnified to some degree. With film, when projected onto let’s say a 60-foot-wide theater screen, a Super 35mm print has to be magnified more than an IMAX print to illuminate the entire 60-foot screen. If both of these hypothetical films were shot using the same film stock, they would both have the same size film grain. If you then projected each film onto the same size screen, although the film grain would be the same size in both stocks, due to its smaller physical size, the Super 35mm print will need to be magnified more than the IMAX print for projection, and so the size of the film grain would also be magnified more. So the film grain in 35mm film would appear larger and more noticeable when compared to the film grain in the 35mm film.
It’s the same in digital. If you shot two theoretical projects, one on an ARRI Alexa Mini shot at 3.2K, and one shot on an ARRI Alexa LF in 4.5K, both camera’s sensors have the same size photosites. So if you kept the camera’s settings the same (aspect ratio, ISO, shutter, codec, etc), and projected the two films onto the same size screen, like with film, you will see that same aesthetic differences in regards to grain (or "noise" in the case of digital). Because the Alexa LF sensor is so much larger than the Alexa Mini sensor, the Alexa LF’s images would need to be magnified less than the Alexa Mini’s images, therefore the appearance grain or “noise” would be less in the Alexa 65. In the images below you can really see the presence of noise in the Super 16mm format example. It's more difficult to see the differences between the Super 35mm image and the Full Frame image because of the limitations of presenting these small, compressed images on your computer screen.
In the example below we compare 3 images with the same field of view, shot at the same T stop, and you can really see the difference in the appearance of grain/noise as well as the difference in how defocused the background can be if field of view and T stop are the same.
One of the simplest differences to understand is the magnification of the recorded image. When we finally show our recorded images to an audience, they are magnified from their original size (the size of the sensor or of the film used to capture the images). Whether the images are being watched on a cell phone or an IMAX theater screen, the original recorded images will be magnified to some degree. With film, when projected onto let’s say a 60-foot-wide theater screen, a Super 35mm print has to be magnified more than an IMAX print to illuminate the entire 60-foot screen. If both of these hypothetical films were shot using the same film stock, they would both have the same size film grain. If you then projected each film onto the same size screen, although the film grain would be the same size in both stocks, due to its smaller physical size, the Super 35mm print will need to be magnified more than the IMAX print for projection, and so the size of the film grain would also be magnified more. So the film grain in 35mm film would appear larger and more noticeable when compared to the film grain in the 35mm film.
It’s the same in digital. If you shot two theoretical projects, one on an ARRI Alexa Mini shot at 3.2K, and one shot on an ARRI Alexa LF in 4.5K, both camera’s sensors have the same size photosites. So if you kept the camera’s settings the same (aspect ratio, ISO, shutter, codec, etc), and projected the two films onto the same size screen, like with film, you will see that same aesthetic differences in regards to grain (or "noise" in the case of digital). Because the Alexa LF sensor is so much larger than the Alexa Mini sensor, the Alexa LF’s images would need to be magnified less than the Alexa Mini’s images, therefore the appearance grain or “noise” would be less in the Alexa 65. In the images below you can really see the presence of noise in the Super 16mm format example. It's more difficult to see the differences between the Super 35mm image and the Full Frame image because of the limitations of presenting these small, compressed images on your computer screen.
In the example below we compare 3 images with the same field of view, shot at the same T stop, and you can really see the difference in the appearance of grain/noise as well as the difference in how defocused the background can be if field of view and T stop are the same.
Another big difference between smaller formats and larger formats can be the depth of field you are able to achieve, and how much you are able to blur the background relative to the subject. If you compare the 3 images above, the camera was placed in the exact same position for all 3 shots. The woman was placed in the exact same spot in the room. All 3 were shot at the same T stop, with the same field of view, and focus was set to her eyes. The 3 images were captured with camera that have different sensor sizes, therefore 3 different focal lengths were required to capture the same field of view. You can really see how much more background blur you can achieve when using longer focal lengths on larger sensors compared to using wider focal lengths on smaller sensors. To explain this a little further, images captured by cameras with Full Frame sensors and Super 35 format sensors, using focal lengths that will give you the same field of view and the same amount of background blur relative to the subject, is about 1 full stop. For instance, if you capture an image with an ARRI Alexa Mini using a 50mm lens at f1.4, to get an image with the same field of view and the same depth of field on an ARRI Alexa LF, you’d need to use a 70mm lens shot at f2. If you want to go the other direction, and capture a third image with the same field of view, and the same depth of field, but this time using the ARRI Alexa Mini’s S16 HD format (Super 16), you would need to use a lens that was 26mm f0.7. As you can see, it is theoretically possible to achieve extremely shallow depth of field in smaller sensor size and formats, but it might be difficult or impossible to find lenses that have such huge maximum apertures. There aren’t too many f0.7 lenses out there. So one advantage of shooting on larger formats like Full Frame is the ability to achieve shallow depth of field, on images with a wide field of view, with the lens options that are available to filmmakers.
The Future
Full Frame is not a fad. It is here to stay. That does not mean however that it is replacing smaller formats. It's important to understand the benefits and the aesthetic differences, but it's also important to keep in mind what is best for your project. These larger format cameras and optics deliver stunning images, but they also can be more expensive to buy or rent. Those big sensors fill up hard drives really fast. Your favorite set of lenses, might not fully cover a Full Frame camera. We hope that this latest chapter in our lens test will not only be used to compare full frame lenses, but also to compare the lenses and formats used in our other 2 lens tests. Full Frame digital cinematography is just the latest tool in your toolbox, and we hope this growing lens library will help you to decide which tools are best for your project.
