Section 2
CHALLENGES IN REALIZING THE CREATIVE VISION TODAY
Modern digital creation workflows have struggled to keep pace with collapsing timelines for delivery and exploding VFX shot counts. Traditionally, network TV episodes were delivered weekly and produced on a staggered production timeline. At any moment, different episodes were at different stages and the post-production demands were consistent. However, the model for OTT and some cable distributors to deliver whole seasons of TV series at the same time has disrupted this cycle. For example, post-production may be postponed until principal photography is complete for all episodes, thereby creating a spike in demand for post-production services as 13 or more episodes appear at once, all needing color correction, sound mixing, finishing, dubbing, QC and encoding simultaneously. This results in a drought or flood scenario for vendors. The industry needs solutions that enable more elastic post-production resources, most likely through leveraging global talent availability, perhaps with AI-based automation to accelerate the processes.
Increased demand for content will require reuse of production assets. With the massive growth in scripted content and popularity of franchises, brands and universes, production elements are used in multiple features, across marketing material (e.g., multiple campaigns, press kits, and co-marketing material), ancillary material (e.g., bonus features, supplemental material, and XR experiences) and related works (e.g., video games and animated spin-offs). Each new deliverable requires a creative team to create, review and approve, often concurrently with the feature’s production and post-production. While ancillary material could hypothetically reuse CGI and other VFX assets to create other media types (game engines, XR experiences), reuse is currently not easy. Often, assets are recreated, draining resources and contributing to production delays. Late supplemental media lowers the impact of the title and franchise. A robust platform to store, find, share and archive production assets could enable reuse in sequels and spin-offs such that these assets are not recreated for each title.
Movie and TV productions are faced with significant security challenges today. The bigger the audience, the more likely that someone will try to steal the product. In common with all cybersecurity, this is an ever-escalating battle of technology, with defenders needing 100% success and attackers needing just one way in. Cybercriminals are making good use of cloud technology, and many of the tools they need are available on the dark web as SaaS services today, with the threat of more powerful tools like quantum computing coming. Managing cybersecurity is a complex task, and the aspects of cybersecurity that are unique to our industry make it even more complex. The production environment is fluid. Multiple vendors and sometimes sub-vendors are used, each with different security capabilities. As cloud workflows take over from traditional workflows, traditional perimeter security solutions are not well suited and we will need a new cybersecurity approach that is designed expressly for dynamic cloud infrastructure, shared compute models and software-defined workflows.
In common with all cybersecurity, this is an ever-escalating battle of technology, with defenders needing 100% success and attackers needing just one way in.
Identity management, the basis of access controls, is becoming increasingly difficult to implement without getting in the way of users. Video production employs a large number of people in development, management, crew, marketing, VFX, post and so on. These people are employed by studios, production companies, agencies and vendors. Many are individual contractors who could be working on multiple productions simultaneously. It becomes difficult to track everyone, and even harder to have effective security management. One studio IT team estimates that an individual working on multiple productions across different studios can have up to 50 different applications and systems to access, each with its own minimum password requirements, multi-factor authentication methods and expiration schedules. This creates a nightmare of usernames and passwords for team members to manage – often by reusing them or writing them down, thus negating the gains of securing access. As workflows become more distributed, effective identity management becomes even more important. The industry must find a way to simplify identity management and do so in a way that does not frustrate users.
Managing the media elements used throughout productions has become an enormous burden – a typical tentpole movie can include hundreds of thousands of discrete media elements, including CGI assets, audio stems (sound effects, music, dialogue tracks), lookup tables (LUTs), composited layers of video and physical assets such as props, some of which have both physical and CGI manifestations. Management of these assets spans multiple systems; sometimes, they are even still managed using “pencil and paper” systems, which are archaic by today’s standards and for most other industries that have benefited from huge gains in productivity by automating their production systems. In addition, each of these media elements may have a specific creator with specific contractual license terms that constrain usage. When the asset becomes part of the final content, the ability to identify where an asset was used and the rights associated with it may be lost. Usage and licensing rights are often tracked in contract documents, guild license terms or spreadsheets, and often in a format that does not facilitate tracking compliance and other obligations. We have an opportunity at this time to move teams and processes into the digital age using modern technologies and enable tracking and linking of our complex asset and workflow systems.
Despite advances in computing technology, rendering of complex VFX-heavy scenes in productions is still not possible in real time with full ray-tracing. Some scenes require hours or days to render a single frame, even using massive render farms filled with parallelized and powerful servers running billions of complex ray-tracing calculations. These “offline” renders constitute an enormous cost in modern filmmaking in money and time. The time required to render limits productions to a multi-day loop (render, receive feedback, iterate, render, receive feedback, iterate, etc.). With up to 30% of the budget of a major Hollywood movie going to render costs, any additional efficiencies can have a huge impact on the profitability of titles. While the number of render passes can be minimized, certain rigging, lighting, animation and simulation decisions can only be made when a scene has been rendered at final output quality with ray-traced rendering. Ultimately, productions will need to migrate to a world where these renders can be achieved with photorealistic quality in real time for rapid iteration cycles and allow creatives more time for their work.
Continuing to scale the content output from our industry requires dealing with some of the inefficiencies that are inherent in embracing unique creative workflows.
Finally, the industry takes great pride in the fact that the creative process is a culmination of experiences and workflows and each production is a unique combination of talents, processes and tools. Creatively, every production should be a “snowflake.” However, from a systems perspective, “every production is a snowflake” creates a nightmare of complexity for studios managing and finding files, configuring tools and authorizing processes. Continuing to scale the content output from our industry requires dealing with some of the inefficiencies that are inherent in embracing unique creative workflows.
The MovieLabs 2030 Vision foresees an industry that collaboratively addresses these challenges, increasing efficiency, quickening production cycles, lowering costs, improving security and increasing profitability while creating opportunities for storytellers to realize their vision and delighting audiences.
Identity management, the basis of access controls, is becoming increasingly difficult to implement without getting in the way of users. Video production employs a large number of people in development, management, crew, marketing, VFX, post and so on. These people are employed by studios, production companies, agencies and vendors. Many are individual contractors who could be working on multiple productions simultaneously. It becomes difficult to track everyone, and even harder to have effective security management. One studio IT team estimates that an individual working on multiple productions across different studios can have up to 50 different applications and systems to access, each with its own minimum password requirements, multi-factor authentication methods and expiration schedules. This creates a nightmare of usernames and passwords for team members to manage – often by reusing them or writing them down, thus negating the gains of securing access. As workflows become more distributed, effective identity management becomes even more important. The industry must find a way to simplify identity management and do so in a way that does not frustrate users.
Managing the media elements used throughout productions has become an enormous burden – a typical tentpole movie can include hundreds of thousands of discrete media elements, including CGI assets, audio stems (sound effects, music, dialogue tracks), lookup tables (LUTs), composited layers of video and physical assets such as props, some of which have both physical and CGI manifestations. Management of these assets spans multiple systems; sometimes, they are even still managed using “pencil and paper” systems, which are archaic by today’s standards and for most other industries that have benefited from huge gains in productivity by automating their production systems. In addition, each of these media elements may have a specific creator with specific contractual license terms that constrain usage. When the asset becomes part of the final content, the ability to identify where an asset was used and the rights associated with it may be lost. Usage and licensing rights are often tracked in contract documents, guild license terms or spreadsheets, and often in a format that does not facilitate tracking compliance and other obligations. We have an opportunity at this time to move teams and processes into the digital age using modern technologies and enable tracking and linking of our complex asset and workflow systems.
Despite advances in computing technology, rendering of complex VFX-heavy scenes in productions is still not possible in real time with full ray-tracing. Some scenes require hours or days to render a single frame, even using massive render farms filled with parallelized and powerful servers running billions of complex ray-tracing calculations. These “offline” renders constitute an enormous cost in modern filmmaking in money and time. The time required to render limits productions to a multi-day loop (render, receive feedback, iterate, render, receive feedback, iterate, etc.). With up to 30% of the budget of a major Hollywood movie going to render costs, any additional efficiencies can have a huge impact on the profitability of titles. While the number of render passes can be minimized, certain rigging, lighting, animation and simulation decisions can only be made when a scene has been rendered at final output quality with ray-traced rendering. Ultimately, productions will need to migrate to a world where these renders can be achieved with photorealistic quality in real time for rapid iteration cycles and allow creatives more time for their work.
