Our mark is 銀幕, ginmaku — the silver screen, and the same two characters in Japanese, Chinese and Korean.

I came to Tokyo at twenty-three for a summer job at a Shibuya mini-theatre and the job never quite ended. Fifteen years of festivals later I still think the only review that counts is the one the room gives you: where the laugh lands, the moment three hundred people stop moving at once, whether anybody stays for the credits.

Cinepuuulse follows Asian cinema — Japan, Korea, China, Taiwan, Hong Kong, Vietnam, Thailand — across about a hundred and twenty sources. A good number of them are Spanish, French and Portuguese, because those countries programme this work seriously and write about it better than most of the English-language trade. Genre and art-house get the same seat here. I am not ranking them, and nothing is scored out of ten.

Each story is a headline, a short excerpt, one still and a link to the outlet that did the reporting. We never republish the piece, and the still belongs to its studio or its photographer. A Korean trade paper appears in Korean, a French review in French. We point at the work; we do not put words in anyone's mouth.

— Daren Wise, Tokyo

← Cinepuuulse Archive

VP Land

fal Launches FLUX 3 Max, a Post-Trained Video Model Benchmarked at About 12 Seconds per 10-Second Clip

10.26

fal launched FLUX 3 Max, a post-trained version of Black Forest Labs' FLUX 3 that generates a 10-second video in about 12 seconds, according to fal's FLUX 3 Max page.

Streaming AI in film Generative AI

Read at VP Land ↗

One letter a week, starting soon.

Latest

fal Launches FLUX 3 Max, a Post-Trained Video Model Benchmarked at About 12 Seconds per 10-Second Clip

Related

Hy-Nomas Is Tencent's New Open-Source Video World Model You Steer With Text and Camera Paths VP Land

BytePlus Introduces Seedance 2.5 4K With Lower-Cost 480p Draft Previews VP Land

OmniHuman 2.0 Is BytePlus' New Avatar Model for Five-Minute, Audio-Driven Performances VP Land

Prime 360 Is Prime Video's New Virtual Camera System for 3D-Reconstructed NBA Replays VP Land

Viggle Launches JST-2, a World Model That Generates 3D Scenes, Animation-Ready Characters, and Full-Body Motion VP Land

We Question Jeffrey Katzenberg's Plan for a Pixar of AI and Weigh Opus 5.5's Blender Results VP Land