8 minute read

The Ins and Outs of Video Quality

How GFX system cameras’ larger-format sensor and X-Processor 4 combine to produce professional quality video

If you want to create high-quality video, you’ve got lots of choices. After all, we live in an age that’s dominated by hybrid cameras, and where 4K video is commonplace both on pro-level and serious enthusiast models. But for serious videographers and directors, the FUJIFILM GFX system offers something different: the combination of a large format sensor with professional level 4K imaging. For a number of reasons, larger format camera systems can offer results that look distinctly different from smaller cameras. Of course, professional users need a full suite of options alongside that pure quality, which includes versatile output options, connectivity, stabilization, and more. For now, let’s take a look at why the combination of the GFX sensor and high-speed image processing makes GFX100 and GFX100S so well-suited for video.

Uncropped quality

Let’s start with the sensor size. One of the ways that quality is improved in video is by using the full width of the sensor area for movie recording, rather than capturing the footage from a cropped portion of it. So, although the aspect ratios of modern video formats mean some of the GFX sensor’s regular 4:3 shape is lost at the top and bottom, GFX100 and GFX100S still use the entire 43.8mm of the sensor horizontally. The recorded information is then down-sampled from the original capture of over 60 million pixels, into a DCI4K 4096×2160 or UHD 3840×2160 output. There’s a big benefit to this, because using more pixels to create the original source data can give you better detail, noise performance, and accuracy of colors and tones.

Many cameras apply an additional crop to the sensor area when recording in 4K resolution. On top of lowering the output quality, compared to a full-width capture, this also leads to a loss in the field-of-view from the camera’s usual lens set. What’s more, not only is it a given that the sensor area used by the GFX system for 4K-class video is larger than full-frame cameras, but it’s also greater than that of most professional cinema cameras on the market today. And better still, it’s all packaged in the compact weather-resistant body designs of GFX100 and GFX100S.

Increased bit depth

Another hugely important factor in terms of video quality is the bit depth of the files. Rather than relating to resolution, the greater the bit depth, the more accurate and enhanced the tonal gradient and color is in a recording. Many cameras, at an output of 8-bit depth per color channel, limit their recordings to only 256 steps or gradations between black and white, for a maximum of 16.8 million colors. However, by using the 10-bit recording options on GFX100 and GFX100S cameras, the output files are made up of 1,024 gradation steps for a possible range of over 1 billion colors.

This increase in quality is very useful, not only when recording scenes that have very disparate highlights and shadow tones, but even more so when holding onto subtle gradations in softer colors and skin tones is needed. So, you get smooth and natural-looking footage, rather than the ugly ‘steps’ or ‘banding’ you can see in lower bit-depth video. And even though most screens and TVs only support 8-bit display, the important thing is that the original 10-bit source information can be used to create higher-quality footage after processing in the camera or during post-production, especially when creative decisions call for dramatic changes to color and contrast in scenes.

More color information

Another quality factor to consider is the process of color sub-sampling. This is a measure of how color information is compressed for recording and, therefore, it also relates to image quality and accuracy. Video compression methods are quite different from those used in stills imaging, and in fact multiple kinds of compression are used in combination during video recording. The level of color subsampling is written such as 4:2:0 or 4:2:2, etc. These numbers by themselves are not significant, but 4:2:2 means more color information is retained compared to 4:2:0, and for some things this can give you better results during post-processing the video.

GFX100 and GFX100S record in 10-bit 4:2:0 internally, but provide an uncompressed 10-bit 4:2:2 signal via the HDMI port which can be recorded to an external device. The internal 4:2:0 color compression is actually quite adequate for most purposes, but having the even better option available when needed is a good thing.

Expanded options

Video from GFX100 and GFX100S can also be recorded in a variety of formats and encodings, offering great flexibility and ensuring the highest quality when it’s needed. One commonly used encoding scheme is something known as “Log”, short for logarithmic, and FUJIFILM cameras utilize the specially engineered “F-Log” variety. Characterized by a distribution curve that preserves the widest-possible dynamic range recorded, F-Log has very low contrast and more retained detail in highlights and shadows. That means filmmakers start with an extended range of picture information, which increases quality in post-capture manipulation and interpretation during color grading.

F-Log mode is very useful if you need the highest quality cinematic video, though it needs to be graded to bring the best out of it. For comparison, here you’ll see the Astia Film Simulation mode on the left and F-Log on the right, ungraded (first image) and graded (second image)

F-Log mode is very useful if you need the highest quality cinematic video, though it needs to be graded to bring the best out of it. For comparison, here you’ll see the Astia Film Simulation mode on the left and F-Log on the right, ungraded (first image) and graded (second image)

GFX100 and GFX100S also support Hybrid Log Gamma (HLG) output, which is an internationally-recognized high-dynamic range (HDR) format. Related to F-Log, HLG is pitched for use on HDR TVs, and uses 10-bit color information to create broadcast-ready footage that covers an extensive array of brightness levels and color saturation. Importantly, HLG can be played on both standard screens as well as HDR-compatible models, and it can be manipulated in post in a fashion similar to F-Log, but with somewhat more flexibility in highlights. Although one needs a true HDR-capable display in order to see the full brightness values of HLG format, almost all video editing and color correction software provides the proper workflows for it.

Should it be necessary to have even more versatility in post-production image manipulation, both GFX100 and GFX100S can supply Apple ProRes RAW video data output via HDMI to external devices, such as Atomos Ninja V field monitor/recorders. Providing an unprocessed 12-bit feed from each camera’s sensor, ProRes RAW video data makes the most of the sensor’s dynamic range, providing huge flexibility in terms of exposure adjustment and color grading. Handily, simultaneous output of ProRes RAW along with Film Simulation/F-Log/HLG recording is available, letting users enjoy the best of processed and unprocessed footage. Having ProRes RAW as a backup can be used for the rare situations where the image might not have been exposed properly, but in practice, F-Log is more than enough even for HDR distribution.

Flexible formats

As regular footage is recorded, it’s compressed, and the method of compression used – the “Codec” (short for compression-decompression) – affects the image quality, as well as how large the files are. Overall file sizes are determined by the amount of data being pushed out of the camera’s processors, and are quoted in Megabits Per Second (mbps). Driven by the high-speed X-Processor 4 engine in both cameras, this means you can get up to 4K/30P 4:2:2 10-bit data via an HDMI connection, and record 4K/30P 4:2:0 10-bit data to one of the SD cards at up to 400 Mbps using the H.265 codec.

Data rates to the external recorders are governed by the recorder itself, and can reach over 1 Gigabit per second. While many cameras’ internal recording options max out at 100 Mbps, GFX100 and GFX100S’s 400 Mbps settings mean fantastic image quality without overwhelming file sizes. And once again, this may present cinematographers the best route for both heavy editing in post-production; when filming against a green screen environment; or when the subject matter is moving quickly. Conversely, lower data rates can be chosen to suit different subjects and workflows.

Similarly, GFX100 and GFX100S’s codec options allow either “Long GOP” or “All-Intra” encoding, providing filmmakers with a choice that suits their subject matter and editing software. While Long GOP uses a long string of multiple frames that relate to each other, allowing lower Mbps settings, All-Intra has no interpolation and uses the maximum 400 Mbps bit rate. Because in All-Intra, each frame is compressed on its own, which minimizes artifacts especially with moving subjects. You won’t always notice the difference, but subjects such wind-blown fields, or rushing water really benefit from All-Intra encoding.

Putting it all together

In summary, the combination of GFX100 and GFX100S’s X-Processor 4 engine and the fact that the signal is created from a larger format sensor offers a lot of benefits for filmmakers. Not only do they get the distinctive look from a larger format sensor, with the potential for superb sharpness, very shallow depth-of-field, greater color accuracy, and lower noise, there are also multiple quality settings to support all sorts of subjects and workflows. Overall, it’s a very complete package.

Learn more about GFX system here or contact your nearest FUJIFILM Authorized Dealer for more information.