Reflective Statement
While I have used RunwayML before, this was my first time using it with more detailed instruction and guidance. That said, it was still very difficult to get Runway to animate my images in the way I imagined in my artistic vision for this project. This made me rethink how some shots should look so that everything would be feasible for Runway while still maintaining a cohesive sequence, much like a movie trailer. Getting DALL-E to generate the images was easy enough, as I had been discussing my fictional world and strategizing with my custom GPT to come up with effective prompts for the images and scenes I had in mind. That said, I did forget to adjust the aspect ratio of my images to 16:9.
As my narrative within this world is very dystopian, I wanted to make that very clear in all the video, animations, and audio. Text prompts in RunwayML were unfortunately not as effective as I had hoped in animating clearly dystopian scenes, so I sought to make up for it by generating music in Suno.ai that reflected a repressive atmosphere. I then sliced it up and re-sequenced it to build up tension at critical moments. The music isn’t particularly sad or angry—it’s just tense and fluctuates in intensity. I thought minimal speech would be effective at maintaining this tension as well, so I used a soft but clear voice from ElevenLabs for the narrator to state just the major forces and conflicts that are at play in the story, with pauses for more dramatic effect. I also gave my main character just one line, in a Russian accent:
“They will….kill me…”