For my video project, I decided to continue Hwei’s story and give some context as to what drove him to absolute madness. After being fed up with Aggressive Technologies and their mass expansion and brainwashing of the world, Hwei decides that he must become an example of what humanity can achieve if they choose his route and embrace AI.
Like I said before in last week’s summary post, Hwei has never been animated to talk in front of a camera, so I had to find a mad genius look-alike in the League of Legends universe to fill in. I had to get clips from the movie Arcane, and there were plenty of clips of the mad scientist Viktor, who is actually the embodiment of Hwei’s character in its entirety (other than being European and not wearing robes / being artsy all the time).
Since the whole premise of my story was to show that initial human side of Hwei then transforming into a robot, I recorded my own voice with the script I wrote to really sell the idea of human becoming something different entirely.
I used Wav2Lip for the second time now, and it was just as disappointing as last time. Not sure if I’m doing anything wrong, but it seems as though the lip syncing is never up to my standards. There are a lot of parts where I am not speaking, yet the algorithm moved the model’s lips. I also couldn’t find a continuous 50 second clip of him just talking, so the 4 second clip I used kept looping over and over, which didn’t make it very realistic.
I got all of the video clips from YouTube, but I noticed that the YouTube to video file generator I used reduced the video quality tremendously. The entire video looks 480p, and I think I’ll try to find some ways to download crisper videos for the final project.
Just like I did in the Radio Show, I had a lot of fun syncing up the different music and voice line files to create dramatic effect. The black screen before the transformation cutscene signified a shift in atmosphere, along with the new music track being added in tandem. When Hwei touches the ominous purple device, jolts of lightning come into frame as the scene cuts again with the music at the perfect timing, and a picture of robot Hwei (Which is actually a portrait of Viktor) comes in. The voice lines were taken from YouTube, and since they were voice lines from the video game, I had to splice together a few lines in audacity to make a coherent “villain announcement” segment.
Luckily I had no issues with copyright music or clips like I was worried about last week, and I hope it stays that way so that everyone can still view the video! Overall this process took around 2 hours and I definitely have some takeaways for the final project moving forward.