The strong digital acceleration we are experiencing has generated an unprecedented demand for software, to the point of making a business based on the exclusive production of professional developers is hardly sustainable.
To them are added the citizen developers. The spread of low code / no code methods and AI tools also allows simple software users to customize and automate their applications, in an increasingly simple and intuitive way, leaving the burden of dealing with development projects above all to the pro teams more important ones, which have more complex functions and long life cycles.
Software engineering today also has a new goal, which for many represents the future of industry: to make software available to everyone.
The democratization of software it is a phenomenon to be interpreted on various levels, ranging from the initial development to the use of a commercial product by end users.
To understand what are the main trends that characterize the development, we met Paolo Emilio Selva, Principal Engineer of Weta FXa great expert in software engineering and fresh off the extraordinary success of Avatar 2, which received the Academy Award for Best Visual Effects at the most recent edition of the Academy Awards.
After his first experiences at Cinecittà, Paolo Emilio Selva flew first to London and then to Wellington, to the court of Peter Jackson, where he contributed to the development of the tools with which many of Weta Digital’s masterpieces were created.
The Software Engineering department, headed by Paolo Emilio Selva, has developed many of Weta Digital’s key technologies, used in films such as Avatar, TinTin, The BFG, The Guardians of the Galaxy and many others.
VFX productions have always been a forge of new technologies in the software field, where products are born that often find subsequent adaptations and applications also in the enterprise field, especially in the field of computer graphics.
How does a company like Weta approach software development?
Selva: Weta digital boasts about 20 years of internal development, in which we have witnessed various stages of evolution of the industry. We are in a very technical context, traditionally oriented towards very competent technicians and artists.
From these experiences, important contributions have come for the development of commercial software and now, after the acquisition by Unity, a new season is starting, which will lead us to make our software available in the cloud to everyone. I will return to this topic later.
This aspect marks a turning point in the field of VFX, where large companies such as Weta and ILM have so far not marketed their software directly. Pixar, with Renderman, is probably one of the few exceptions on the market in this sense.
From a development point of view, what is changing?
Selva: You start to look more and more at what happens outside the VFX studios, encouraging the adoption of an open standard that allows you to absorb the products of the community. There is a progressive move from a very specific pipeline, written ad hoc by internal developers to the integration of parts developed by students and artists. For this to happen, it is necessary to define open standards that allow all levels of development to speak the same language.
The VFX industry has always been a source of applications that subsequently see the adoption also in the enterprise context. But what does the life cycle of a software really consist of in a reality like Weta? How was the software born?
Selva: In a visual effects studio software is developed to support artists in their productions. For example, our tools automate tasks that are repeated many times, aim to increase the efficiency of processes, or solve very specific situations that are manually extremely complex or time-consuming.
I could cite many examples. For Guardians of the Galaxy the scenes included a large use of fractal systems, for which we developed a tool capable of creating them in a parametric way, avoiding the artists the burden of having to model them one by one. A similar circumstance occurred again for Avatar 2.
Each film generates new reusable assets, which help grow our portfolio to support artists.
The software you develop at Weta is also well known for the frequency with which it wins major international awards, starting with the technical Oscars. Could you tell us some development stories?
Selva: When a production requires particular solutions, there is a tendency to industrialize the software in order to make it available in various circumstances. For example, Lumberjack was created for the Hobbit trilogy, a software for generating entire vegetation systems in a realistic way, which we subsequently also used in other productions.
For the last chapter of Planet of the Apes, there is a pivotal scene involving a mountain with a snowy forest. Lumberjack wasn’t enough anymore, as we soon realized that we needed a forest generator, not just trees. Each tree had to grow on the basis of certain climatic conditions, with particular distribution logics, without crossing each other.
We worked to achieve a realistic physics simulation, where the branches could believably give way under the weight of the snow, which was also procedurally generated. From this production was born Totara, our forest generation software.
Production is distinguished by the need to customize every aspect, in the creation and management of assets. How do you evaluate the impact of the new AI Tools?
Selva: They are not ready to be implemented in a production pipeline. Their potential is incredible and they will certainly be more and more protagonists in the creative industry in the future. We follow their evolution in research and development to test their applications, but their use in projects is not currently envisaged.
Today these are systems that do not allow you to manage changes by controlling every aspect of the visual. A similar condition is not acceptable in a commercial production, where you cannot afford to obtain a random result, however appreciable from an aesthetic point of view.
We have totally different needs. We need to standardize processes to have absolute certainty that each asset is saved and managed in the same way, to be perfectly interoperable with the pipeline tools.
All artists must work under the same conditions. As for the creation, the use of various tools is allowed, there is room for customization. Assets are managed through a unified workflow.
When the artist saves a scene, all the assets that compose it are automatically updated in their respective archive locations. The result must be an absolutely “error free” system. The artist doesn’t even have to consider an IT problem, we have to put him in a position to think only of doing his job.
As for the no code approaches, how do you evaluate the implementation in the computer graphics field?
Selva: Il VPL (Visual Programming Language) and in general, nodal editors have been very popular for years now. Today even a fifteen year old is able to make simple customizations on Blender. I don’t think this is the central aspect of the matter.
In production, as I mentioned, other needs emerge, which are related to the ability to customize some functions without generating exceptions in the pipeline standards. Commercial software like Maya, the most widespread DCC (digital content creator) in the VFX field, in its default configuration is not able to do everything we need.
Our job is to understand what artists really need and develop the tools necessary to allow everyone to work in the same conditions, producing a result that can be used in production.
What did the acquisition of your unit by Unity entail at a strategic level?
Selva: Unity is a company that develops and distributes commercial software, it is not a production company like Weta Digital. The goal is to make the solutions that we have used internally for many years available to everyone.
Unity intends to democratize software in the expectation that in the coming years the market will require more and more applications capable of creating 3D virtual worlds. For this to happen, a very important work is needed.
Now it is necessary to define many tools starting from an overall pipeline and make them available in the cloud, on the AWS platform, where users can use them through a simple web interface. We have 20 years of code available to reorder and make usable again, this time to all 3D content creators in the world. This is the democratization challenge that awaits us in the future.
breaking latest news © (Article protected by copyright)