works-films – Kai Tattersall

Liminal Places

Liminal Places

The feed is a place of friendship, love, community, and fairytale. Everyone is surrounded by an interested audience and stories curated just for you.

¸,ø¤º°`°º¤ø,¸¸,ø¤º°

Liminal Places utilizes artificial intelligence to create fully artificial voices. Manual editing allowed controlled intonations and pacing. Each voice relates to a person in my life.

Environments were constructed using photogrammetry. Augmenting the scans allowed for top-to-bottom production design and the visualization of impossible structures. Locations were collected from my hometown.

The social media feed combined fiction and reality. Sampled captions were modified to curate a realistic selection of diction and content. “User-strings” were created to fit each post.

Granular synthesis created the ambience and score. Live modulation of an original score married visuals and sound. The score was controlled through a rigged-joystick.

This is a USC junior thesis. A big thanks to Madison Kloeber, Valerie Jackson, and Jason O’Connor for their invaluable contributions. Also, many thanks to the amazing faculty: Amy Reynolds Reed, Brad Barnes, Dianne Farrington, John Rosenberg, Navid McIlhargey, and Rich Hyland.

Tell Me What You Think

Exploring a digital playground is a perfect way to meditate on your thoughts, free from the sensationalism of the real world, and curated just for you.

.・。.・゜✭・.・✫・゜・。.

Tell Me What You Think utilizes motion tracking for first-person point-of-view. The process involved some couple hundred sticky notes and detailed choreography. The cinematography was done by Anna Waldron. Valerie Jackson dressed the environmental elements.

Elements of sound were mapped to a keyboard and played live. The sample bank includes home videos, synthetic sounds, foley, and other sound effect libraries. Narration was done by Vivian Gray.

This is a USC project. A big thanks to my amazing producer, Partho Gupte, and my instructor, Brad Barnes.

ILY2222

The bedroom of a narcissist is the most dangerous place to be.

═════ஓ๑♡๑ஓ═════

ILY2222 utilizes AI to synthesize predictive text based on only one side of a conversation. The lines are delivered by Jason O’Connor, and shot by Max Von der Horst. The scene took 26 takes.

The end sequence was created using NVIDIA’s GAUGAN. Artificial intelligence translated stop motion animation into photorealistic landscapes. The result was manually stylized.

The score is played live. Additional sound design elements and vocal effects were added for aesthetic effect. Some instruments automatically react to the visuals.

This is a USC project. Thanks to my encouraging instructor, Jennifer Warren.