A dialogue between the Director / DP of the demo video “A little part of me inside my photos”, Masahiro AIDA, and the Fujifilm Technical Team that worked on the X-H1.
Masahiro AIDA J.S.C.
Born in 1965 in Chiba, Japan. Made his debut in the 90s as a camera operator, has worked on numerous film productions, and was the DP on the popular TV series “Aibou (Partners)” from its inception in 2002. He always incorporated the latest digital techniques on set while shooting in film on productions such as “Young Man H” (2013, director, Yasuo FURUHATA). The theatrically released movie “Aibou (Partners): The Movie III” (2014) was the first domestic movie filmed entirely in 4K. He is a member of the Japanese Society of Cinematographers (J.S.C.).
R&D Management Headquarters
Optical Device & Electronic Imaging Products Development Center
Sales & Marketing Group
The lure of ETERNA is “standard”
AIDA: For this shoot, in order to maximize ETERNA’s features, we tried to keep the tones as natural as possible. I felt it was especially easy to film facial tones. There was no color degradation, and it produced bright tones. I’ve made many movies using ETERNA film. Based on those experiences, I got a sense of the ETERNA’s connections, tones, and color reproduction from the X-H1 film simulation. The link between the shadows and the highlights remained smooth. Even when I shot with an extremely classic shooting style, the skin tones still retained a rich tone and smooth highlights – I got the impression that it was very ETERNA-like, and the image quality was a rank above the rest. Even in challenging environments, the highlights didn’t turn yellow, and it produced beautiful gradation. One of the features of ETERNA is that you can create texture-rich images that have a cinematic look.
IRIE: We are so happy that you rated our product so highly, as you have been using this since the days of film. At our company, we have always striven to achieve natural skin tones as much as possible from the time when we were working with film. While the actors are performing, it’s natural for them to move into different lighting situations. What we want most to avoid in those instances is a degradation in color. Even when the lighting changes, it’s important for skin tones to remain consistent. When it came to ETERNA’s Film Simulation, we were particular about getting the skin tones right.
It doesn’t distract from the acting: “what sets it apart is its lack of character”
AIDA: What makes ETERNA’s Film Simulation unique? What are some of its features?
IRIE: What we were most conscious of was that “what sets it apart is its lack of character.” One major idea behind this was to achieve the “standard” cinematic look with this product.
In terms of what is the standard in video, above all, it is not to distract from the actors’ performances. A major feature is to tell the story, but not overdo it.
In the digital world, it is our understanding that these images undergo numerous edits that you wouldn’t necessarily do on photographs. From that point of view, we aimed for a standard color reproduction.
AIDA: I see. Actually, my colleagues and I chose ETERNA back in the days of film, and it’s for precisely that reason. It doesn’t overshadow the storytelling process, but it expresses the story and it’s easy to use. We also talked about how the color timing was easy to use. It has inherited that spirit.
IRIE: The craft behind ETERNA’s film is a reflection of that idea. We have now changed over to digital, but whether it’s the film era or digital film simulation, that idea remains the same.
The Creative Process of Choice
AIDA: The X-H1’s signal processing rate is 4:2:0 8-bit video. In all truth, if it were 10-bit, the color settings probably would have been much easier for the Fujifilm engineers. That said, in terms of shooting, the balance was great all the way to the end. It was easy to shoot. I had no issues with the dynamic range during this shoot, and I thought that the images looked great. I was really pleased with it, and I’m sure there are a lot of people who are hoping to achieve that look. You’ll get some beautiful shots if you balance everything out within 8 stops and don’t get tied down by all the specs. I think we got close to achieving 10-bit expression.
In reality, though, there isn’t much point in shooting just to make use of all the specs. I choose my cameras and film with every new project, but it doesn’t have anything to do with the specs. Of course the greater your specs, the more you can achieve, but no one uses all the specs.
In movies, the important elements you need to consider when choosing your equipment are whether the color, the tonality, and the mount are suited to the movie you want to make. In that sense, the X-H1 is extremely well-balanced.
This time, because the “simulation” had a brand name, I was glad that I didn’t have to worry about so many things. Since I was working with Eterna, I could simply say, “That’s the color of the sky” and visualize it as such. This is something I can’t do with other brands. Speaking of which, what’s the difference between film simulation and LUT (Look Up Table)?
IRIE: With film simulation, what you are able to express with the subtle color gradations that you calculate for each image differs completely from what is generated with LUT. LUT is based on a mathematical formula that takes what is input (the original source image) and converts it. However, once you’ve made your decision, even with 8-bit, you will have 16,770,000 pairs of reference data, and you will need 50MB (in the case of 3D LUT). If it’s 10-bit, the reference data increases to over a billion and requires 4GB. It becomes virtually impossible to do any complex, high-speed processing in-camera. For this reason, the camera will perform “interpolation”, which actually culls the reference data, but errors can occur because of this. Ordinarily, the reference data (grid) numbers are 33 and 65. The higher the grid number, the less likely it is that an interpolation error will occur, and there will be a higher rate of accuracy, but at the same time more memory and computer processing are necessary. It has its drawbacks: the file size and required memory size increase, it takes longer to read and record data, and the whole system becomes costly.
The X-H1 operates with the ideal 16-bit in-body without the LUT process. In the case of a 65-grid LUT, in the gradation stage for 16-bit, it can deliver a 65,535 grid, for example, but obviously the look will differ. Most of the time, LUT relies on “interpolation” calculation software that performs the calculation processing. At our company, we use the X-Processer Pro Engine with film simulation, which relies on hardware, so it can perform high-speed processing and can reproduce velvety color.
However, all this is not to dismiss LUT. We have also re-released the F-log LUT, which is currently on the market. But the market is flooded with LUT, and there are product quality issues; it’s not uncommon to see such errors like the ones mentioned above. When the number of grid points and the position aren’t ideal – and you can’t really tell from just a glance – there’s the possibility that the art will be destroyed. If you do a thorough job with your LUT, you will have something you can use.
AIDA: This time round I felt like we were able to capitalize on Eterna’s features when shooting with the X-H1. Personally, I think it’s a waste to sacrifice your film know-how for the sake of the camera and the specs, or because of numbers. I hope we’ll be able to use the important know-how we gained in crafting art during the film era.
A robust 5 axis IBIS (In Body Image stabilization)
AIDA: What does the “H” in X-H1 stand for?
WATANABE: The “H” is for High performance. With the X-H1, we wanted to create a line of high performance equipment that was geared towards professionals, so we listened to what the professional photographers and videographers had to say. In response to their requests, we upgraded the design to improve the autofocus and video performance, increased the grip size, and changed the feel of the release. What most of our customers wanted was in-body image stabilization.
MIZUTA: While there are products that reflect the results of R&D regarding image stabilization and there have been lenses with image stabilization built in, we have finally arrived at achieving this for the body. With this built-in 5-axis image stabilization, now you can use lenses that do not have this built-in feature. In particular, we particularly wanted to achieve 5.0 (in compliance with CIPA guidelines) for lenses without this capacity, and laid out a detailed design for each lens.
By the same token, when using the gimbal as a stabilizer, one of the advantages is that the gimbal’s ability to achieve image stabilization in various directions differs from the X-H1’s ability to achieve image stabilization, and you can eliminate even the most minute camera shake when it’s mounted on a gimbal.
AIDA: When it comes to attaching weighty lenses like the MKX, which is being launched at the same time as the X-H1, have there been improvements to the X-mount accessories?
TAKADA: We have reinforced the strength of the body where the mount is located. Compared to past models like the X-T2, which has a magnesium alloy body, it is 25% thicker, and we re-examined the interior framework of the mount and improved its strength so that it can cope with a heavy lens.
In addition, we are upgrading various hardware devices geared towards professionals. For example, we have upgraded the visibility of the exterior crystal screen compared to what we had in the past. We were able to decrease the surface reflection via direct bonding between the cover and the screen (it adheres by means of a eliminating the air between the cover and the crystal screen).
Introduction movie of IBIS (compared to when with the stabilizer used)
※All in Japanese
Upgraded AF feature when shooting in video
OKAMOTO: The video AF setting comes with a menu for video, and you can customize the settings to change the focus speed by 10 different levels. When the subject cuts across the camera, and when you need it to move in a split second, or conversely, when you don’t want it to move, you can slow it down. There is a trade-off between stability and tracking, and that’s where things get challenging. Until now, during shooting, you could pinpoint your focus on a still subject, but with the AF upgrade in the X-H1, we have made it so it won’t make these kinds of movements.
AIDA: Something that bothered me when watching footage shot with previous X series lenses was searching for a still subject with AF. I recommend using AF, but I think my ideal would be to have it be partially manual. When you’re shooting many actors and you know who is the focus, if you can switch to manual operation at that point, it would make shooting so much easier. When you are shooting with film simulation, I appreciate that the recorded image is reflected in the viewfinder.
WATANABE: Partially manual? Thank you for the interesting proposal!
With the X-H1, you will of course find that the film simulation’s true colors will be reflected in the viewfinder. The trademark of the X series is the high performance 3.69m-dot EVF, and we upgraded the response as well.
After the launch of the X-T2, we couldn’t have predicted the public response to 4K video – in particular, film simulation shooting. This time we created the optimum film simulation to go with it: ETERNA. We upgraded the features and operability to some degree in response to the demand. As a high performance machine for professionals, the X-H1 is being sold alongside the X-T2 as part of a separate series.