Article by Florian Berthellot
Seminar at Camerimage: “Digital Cameras, Creative Workflows. Problems with invisible reasons, solutions with visible results.” Conference hosted with the help of Kazik Suwala (Camerimage), Ainara Porron (Sony Europe) & Irena Gruca (FilmPRO).
Amidst all the screenings and seminars at the Camerimage festival there was a drôle d’oiseau (funny bird, strange beast), a hybrid conference with an overview of flux de travail, (workflow) which in reality is not so obvious or easy to understand.
The conference was initiated by Roberto Schaefer and Philippe Ros, cinematographers, and Madelyn Most, a specialized journalist, and gathered an impressive panel and equally impressive audience: cinematographers, equipment rental companies (Panavision, ARRI rental, Clairmont Camera, Vantage Film), manufacturers of cameras (ARRI, Canon, Sony, Transvideo-Aaton), post-production companies (Digimage) and, of course, students. To have all these contributors to the art of cinematography in one room was a great success and a reassuring fact about the future of cinema.
Indeed, while the famous digital “revolution” has arrived as expected, it is clear that this revolution is still far from being unified. Each manufacturer, and each laboratory, worked independently to develop their own tools, often without the necessary consultation with the rest of the profession. This conference was thus intended to both inform about the possibilities of different workflows today and to try to find solutions to problems which begin to appear with the proliferation of these possibilities.
The conference started with an explanation of some key concepts of processing a digital image: the de-Bayering process, SDK (Software Development Kit), color processing, pipeline, and the notion of LUTs (Look Up Tables). This was intended to clarify the overall capture process, which is frequently managed automatically by the camera, and can escape the control of its user.
But the majority of cases are challenged today by the choice between a traditional workflow, and a raw workflow. The classic workflow uses the image processing offered by the camera when shooting in a given color space (RGB/Rec709), with a final image that has the same characteristics as that displayed on a monitor (at least if you use a well-calibrated monitor). It has the disadvantages of a more limited intervention for the post-production (de-Bayering is already performed, a limited color space even with a logarithmic gamma).
It is also possible to capture the image “raw,” which provides the possibility to de-Bayer in post-production, providing a better dynamic range and the chance to control sharpness and color gradingin a wider working color space.
At this point, there quickly arose the questions of choosing your tool, and the importance of those tools in the final result. Roberto Schaefer gave an example of the problems he encountered when shooting in raw, where, for reasons of production, he had to change lab during the post-production of his film. The two colorist worked on different software packages and he was unable at the second lab to find the images he had established with the first colorist. This was in fact the real issue of the conference: how to keep control of an image in the long process of making a film today. Or as the title of the conference said, how to handle problems with invisible reasons for solutions with visible results.
Faced with this problem, it is obvious that it would take much more than the two hours intended for the debate to end the discussion. However, some questions and expectations emerged from this meeting.
The implementation of the ACES protocol seems to be a way to greatly simplify exchanges, at least from the color point of view, between the images displayed at different stages, with different tools and cameras. Also, the generalization of the LUT (or in the case of ACES, an ODT) makes it easier to control the displayed and previewed images.
Nevertheless it is still necessary that the LUTs are properly established and verified at each stage of the workflow. The conference also enhanced the role of the DIT as a privileged intermediary in a raw workflow, able to keep the cohesion between each step. As an example, a cinematographer in the room explained that, after months of editing, the directors became accustomed to the wrong proxies.
In my opinion, one of the major issues raised by this conference was the difficulty of the industry to respond to the strange duality of the needs of the cinematographer, requiring both the most absolute control of the image (which leads today to shooting in RAW, with all the choices directly accessible by the user during shooting or in post- production) and at the same time a tool extremely simple to use and to handle. Shooting on film also came back recurrently as an example of a standard tool that the operators have learned to master and to manipulate to achieve their creative goal. The apparent simplicity and choice often allows much greater creative possibilities.
But, to be really honest, many cinematographers are not necessarily that well informed and specialized, or simply not curious enough about workflows and possibilities of different settings.
In response to this need of cinematographers to regain control of their image, manufacturers who were present explained one by one their own processes of image management, raw or standard shooting. But every manufacturer uses and makes their own SDK, which is also the strength and uniqueness of their images. That is why, in response to the question of a participant asking if it would not be possible to provide a free SDK currently protected by patents, the manufacturers stated that the SDK is the result of long and costly work, and that it would only complicate the actual situation more, with only a small benefit for the user. Indeed manufacturers all use different SDKs, and even different software to treat their own images. Moreover, laboratories have also developed their own SDKs, hence the multiplication of possible workflows today and the rather anarchic situation that emerges.
At this level of the discussion it seemed difficult to find a point of conclusion, and finally few words still made sense to me. Thus, a member of Panavision spoke of the Pantone Color chart (a reference in printing) and a physical object that could allow everyone to check in situ the correctness of the chosen color. It seems there may be a very interesting tool to develop, or a protocol, that allows cinematographers to quickly validate conditions of observing images, which can help to judge definition, dynamic and color accuracy of the images that are shown. But for such a tool to be established, it must be made available to the camera rental houses, the set or location, and all the way through to the post-production houses.
The cinematographers from IMAGO called for more transparency on the one hand, and more information on the other, which is indeed an indispensable prerequisite to improve the existing situation, even if it seems insufficient to me.
In conclusion, this conference, surely too short to truly succeed, was an encouraging sign of the chance for true collaboration between all the different participants involved in the work of cinematography. And while the entire movie industry seems to suffer from a technological bulimia that wishes to impose the constant mastering of new tools, it is also an evidence of a shared commitment to simplification and transparency of the process, about which most users can only rejoice. The question now is whether this call will be heard.
Sony Seminar: Digital cameras, Creative Workflows. Problems with invisible reasons, solutions with visible results. A conference hosted with the help of Kazik Suwala (Camerimage), Ainara Porron (Sony Europe) & Irena Gruca (FilmPRO).
Florian Berthellot is a graduate of the Louis Lumière School of Cinema. This was his first visit to Camerimage. Florian was part of the group of 12 film students brought to Camerimage by Transvideo and K5600.