As technology continues to progress at a rapid pace, post-production is set to change dramatically. Besides great new software features, compute (and therefore render) power continues to increase with faster processors and more powerful xPUs (GPUs, TPUs, etc.). This means real-time editing of HD, 4K, 8K, AR and VR is fast becoming a reality, ultimately shortening post-production timeframes.
Over the next few years, we’ll continue to see the commoditization of hardware and virtualization of software, which will not only lower the cost to operate post-production tools but remove the stipulation that the user is near the computer power or even storage.
AI leads the pack. AI is no longer a budding technology in the media industry. The technology has provided several bespoke solutions from video editing to media assets management, automating mundane tasks and freeing up time for producers to focus on work that truly matters.
If there’s one thing all the experts agree on, is that we’ve probably only just scratched the surface when it comes to how AI can change post for the future. There’s so much that’s exciting. Things like voice understanding, for example. Why should an artist not be able to say ‘take those import clips and process them like this’? That’s the long-term vision.
How we develop AI for post is probably going to change too. In post-production, you can’t just say if the results aren’t good enough, we’ll simply improve the data that feeds into the AI algorithm’ because everybody has a different idea of what is good enough. That’s why in the future, we’ll see more creatives informing how the AI works and what problems it’s meant to solve.
It’s also crucial to consider how AI systems can introduce bias that can negatively impact diversity for the future.
Since the arrival of AI (artificial intelligence), post-production houses and streamers have tapped on AI to automate content workflow, saving US$1 billion a year. However, integrating AI to existing legacy systems needs a careful balance of cost and efficiency.
In 2020, Reuters applied AI technology to 100 years of archived videos, enabling faster news discovery. With the support of Google Digital News Innovation (DNI) Fund, Reuters was able to extract vast amounts of data from the videos to provide accurate information for its editors.
At any 24/7 news agency, efficiency is crucial in delivering up-to-date news, and AI technology has enabled searchable video archives, thus reducing tedious work and enabling an intuitive workflow.
Aside from highly ambitious attempts to let AI take over script writing or editing tasks, for example, there are ways to usher in the new era – and leverage the power of AI purposefully and successfully.
While AI is still in its infancy, and we expect many more purposeful applications to come around over time, AI can already make a big difference today – especially when it comes to media asset management. In a fast-pace environment such as post-production, human beings don’t have to bother with crunching the data anymore – but can use their precious time elsewhere. Designed to perform extremely tedious and time-consuming tasks, AI gets the job done faster and probably better than any human, increasing efficiency and usability significantly.
Adding “intelligent” metadata to the footage used to require hours and hours of manual work, and at the end of the day you still couldn’t be certain that all relevant descriptive data was attached properly.
AI can easily execute this task, probably in a much more diversified fashion than any human being, but definitely a lot faster than any human being could ever do it. If analyzing and tagging of an entire day of footage takes only minutes instead of multiple hours, while providing even more valuable in-depth information compared to the human attempt, both efficiency and productivity are increased significantly. This boost in capability will have a positive impact on the post-production of the entire project, as the enhanced searchability of the media assets.
Keep in mind though: When deploying AI for your media asset management be prepared to spend some time “teaching” your AI application what you need. Because that’s where the ML (machine leaning) comes into play.
Another daunting task is the preparation of transcripts. With the right AI engine, transcripts can be performed automatically – and even be translated into a multitude of other languages, almost simultaneously. The benefits are obvious: Subtitling becomes a breeze with AI, even in different languages.
On the flip side, AI requires training to be able to replicate human problem-solving capabilities. Fortunately, video editing software trains the AI to understand some of the basic problems editors face and helps editors to resolve it automatically.
Not only does AI help with the video access, but it can also automatically suggest edits for mismatch in video colourations during the production process.
Artificial intelligence is transforming the way we handle post-production tasks is an accepted fact.