US tech incubator, Rearden LLC, has been engaged in a drawn-out battle over its facial motion-capture technology, MOVA Contour. The disputed software was used to create the famous faces of Beast from Disney’s Beauty and the Beast and Thanos from Marvel’s Guardians of the Galaxy, to name just a few. But little did these movie studios know that they would be involved in a legal dispute over who owns the copyright in their computer-generated (CG) characters.
MOVA was first developed in the early 2000’s by Rearden. The head of the incubator, Steve Perlman, put his long-time friend Greg LaSalle in charge of the project. LaSalle formed his own company, MO2, and transferred MOVA to it before selling the software to Digital Domain Holdings (DDH), a US visual effects studio. MOVA was transferred to DDH’s Chinese subsidiary, Shenzhenshi Haitiecheng Science and Technology (SHST).
Rearden’s first legal battle over the ownership of MOVA was with SHST in 2016. SHST initiated the proceedings, but Rearden argued that LaSalle had no right to sell MOVA to SHST in the first place. A federal court in San Francisco issued an injunction against SHST and all of their licencees (including DDH’s US subsidiary, Digital Domain 3.0 or DD3) for distributing and using MOVA.
In August 2017, the US District Court of Northern California ruled that Rearden officially owned MOVA, not DDH. Rearden is now going after the movie studios Disney, Fox, and Paramount for using the copies of MOVA that were illegitimately licensed to DD3. The studios have filed a motion to dismiss Rearden’s suit, which was subsequently opposed by Rearden.
One of Rearden’s arguments is that they own the copyright in the studio’s CG characters because 1) they own the output of their software, and 2) this output was incorporated into the films, thus making the films derivative works.
For output ownership, Rearden relies on two video game cases (Atari Games Corp v Nintendo of America, Inc and MDY Industries, LLC v Bilzzard Entertainment, Inc) where the courts ruled that the audiovisual output of computer programs is afforded copyright protection, even if the end user contributes to this output.
This is a questionable comparison to the current case because the nature of audiovisual output in a video game is significantly different from the output of a graphics design program. With interactive entertainment, the output is not intended to have any further use. Even when video game footage is used in subsequent works (like Let’s Play videos), it is done so under licence. The content creator only owns the copyright in their footage and commentary, not the game. This is unlike the output of MOVA, where the 3D render output is intended to be used in further works.
Rearden also claims that they own MOVA’s output because the software does a “lion’s share” of the work required to produce the output. They rely on a US case, Torah Soft, Ltd v Drosnin, in which the court ruled that the plaintiff software maker owned the output of their software because the end user only contributed single words or phrases, and this was considered “marginal” input.
Deciding who does the “lion’s share” of the work when producing a CG character is a difficult issue. To briefly explain the process of creating a CG character using MOVA, an actor wearing special makeup is filmed. This 2D recording is passed through MOVA, which synthesizes a 3D rendering of the face and head. This 3D model is then used as a template onto which animators add features like hair, colour, and wrinkles. Once this is complete, the 3D model is removed, leaving just the animated shell that we see in the movies.
So, it becomes a chicken-and-egg problem: without the software, there is no output, but without the acting and direction, there is no input to make the output. The analogy that the movie studios draw to Adobe Photoshop is quite apt because there is a clear distinction between the contribution of the artist and the contribution of the software, and in that case, it can hardly be said that the artist contributes “marginal” input. The software used in Torah was much simpler and may not offer the best comparison to the present case.
It will be interesting to see how the court decides the issue of derivative works. The movie studios argue that their CG characters and movies are not derivative works because they do not “substantially incorporate protected material from [a] pre-existing work” (quoting Micro Star v Formgen, Inc). Rearden, on the other hand, argues that the CG characters and movies do incorporate their works (i.e., the output of their software). Even though the 3D template and related coding are taken out during the production process and do not appear in the movie, Rearden argues that because the facial motion is transferred from the 3D template to the CG character with “sub-millimeter precision,” the CG character is similar enough to constitute a derivative work. But how would that weigh against the fact that the CG character is also “substantially similar” to the original actor’s face and performance, which the studio owns?
Another important point is the substantial relief that Rearden is requesting. They want all infringing copies of MOVA output and the movies that contain them to be destroyed. Logistically, it is difficult to imagine what this would entail. There is no a guarantee that every copy of Guardians of the Galaxy can or will be destroyed. There is also the question of what will happen to clips of the movies that are being used by third parties. Would those have to be destroyed as well?
The Rearden case brings up some complex copyright issues, and it will be interesting to see whether the courts decide in favour of Rearden or the movie studios. A ruling for Rearden could have major implications on the way end users manage their copyright when using computer programs.
Alexandria Chun is a JD Candidate at Osgoode Hall Law School and was enrolled in Osgoode’s Intellectual Property Law Intensive Program. As part of the program requirements, students were asked to write a blog on a topic of their choice.