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

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

10.25

Tencent released Hy-Nomas, an open-source video world model that builds a scene one video segment at a time while you direct each new segment with text instructions, camera trajectories, or both…

AI in film Generative AI

Read at VP Land ↗

One letter a week, starting soon.

Latest

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

Related

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

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

Runway Labs Tests Generative Video Interfaces That Recompose Scenes as Windows Resize VP Land

We Tested ChatGPT Image 2.5 While TIFF Filmmakers Wrestled With AI Slop VP Land