Big Week For AI Image Generation

In partnership with

Anyone else’s social media feed suddenly shift from Saratoga Springs bottled water and bananas to Studio Ghibli everything?

It was a wild week for AI. Beneath the memes, there are some really powerful implications for pre-viz, like this hand sketch to photoreal image demo from John Finger.

Also, one week to NAB! Gearing up for all our video coverage, so make sure you’re subscribed to the YouTube channel.

Joey

Big Week for AI Image Generation (And Studio Ghiblifying Everything)

OpenAI ChatGPT Unveils Advanced Image Creation

The biggest update this week: OpenAI has launched native image generation in ChatGPT 4o.

The system can generate images from text prompts with precise text rendering – solving a persistent problem in previous image generators
You can modify existing images by giving the AI simple instructions, enabling workflows like turning selfies into anime or creating themed visual content
The technology understands visual context across multiple exchanges, allowing for iterative refinement of images through conversation
Unlike previous models, it can blend multiple image inputs with text instructions to create new composite visuals

This also led to the Studio Ghiblifying every image known to humanity, mainly on X. It’s a bit of a poke in the face to Hayao Miyazaki, the co-founder of the studio, who’s reaction to an AI demo from a few years ago went viral when he said, “I strongly feel that this is an insult to life itself.”

We get more into the weeds of this and the implications when well-known creative styles are blatantly ripped in today’s Denoised.

Ideogram 3.0: Better Style References and Text Rendering

Ideogram has released version 3.0 of its text-to-image AI platform

Style References feature allows uploading up to 3 reference images to maintain consistent visual aesthetics throughout a project
Enhanced text rendering capabilities generate precise, stylized text for graphics
Improved photorealism delivers sophisticated spatial compositions with detailed lighting, environments, and colors

New Image Model: Reve Image 1.0

Reve Image 1.0 has launched as a free preview, offering superior prompt adherence, text rendering capabilities, and aesthetic quality that's already outranking established players like Midjourney and Google's Imagen 3.

Halfmoon is Reve Image — and it’s the best image model in the world 🥇
(🔊)
— #Reve (#@reveimage)
4:37 PM • Mar 24, 2025

Read more to learn how production teams are already integrating Reve Image into their workflows and what the company's "semantic intermediate representation" approach means for the future of AI-assisted creative production.

SPONSOR MESSAGE

Your job called—it wants better business news

Welcome to Morning Brew—the world’s most engaging business newsletter. Seriously, we mean it.

Morning Brew’s daily email keeps professionals informed on the business news that matters, but with a twist—think jokes, pop culture, quick writeups, and anything that makes traditionally dull news actually enjoyable.

It’s 100% free—so why not give it a shot? And if you decide you’d rather stick with dry, long-winded business news, you can always unsubscribe.

Check it out

Chaos Arena Launches for Virtual Production

Chaos Arena, the advanced path tracer formerly known as Project Arena, has officially launched for virtual production professionals. The technology delivers real-time ray tracing directly to LED walls without game engines, enables instant creative changes on set, and supports NVIDIA DLSS 4 for significant performance improvements.

Key details about this new virtual production solution:

Works directly from 3D applications to LED walls with no baking, rasterization, or game engine required
Eliminates data conversion, maintaining the same assets throughout your production pipeline
Allows instant creative changes with physically accurate lighting rendered in real-time
Currently supports V-Ray pipelines with USD and MaterialX support coming soon
Scales to projects of all sizes, making virtual production more accessible than ever

Read more to discover how Chaos Arena compares to traditional game engine workflows and why productions are seeing it as a cost-effective alternative that might reduce your dependence on extensive post-production fixes.

Severance director Ben Stiller and editor Geoffrey Richman reveal how they shaped the gripping season finale. Really gets into the weeds and decision making behind editing (with a shoutout for the Kuleshov Effect).

Stories, projects, and links that caught our attention from around the web:

👨🏻‍💻 MAGIX, which owns VEGAS Pro (formerly Sony Vegas), has been acquired by RM Equity Partners after a successful restructuring, paving the way for strategic growth in digital content creation

🏟️ SP Grid from Stage Precision allows users to visualize system connections, automate workflows, and deploy custom interfaces, making it easier to manage live and virtual productions.

🎙️ Zero Density is launching a template-based workflow for broadcast graphics, enabling producers to efficiently create and manage on-air graphics, video walls, and virtual studios using Unreal Engine 5.

🗄️ Axle AI and Archiware have announced a collaboration aimed at enhancing media management and archiving solutions on M4 Mac Mini and Mac Studio.

📲 Creamsource announced the Slyyd app, where film professionals can streamline their lighting workflow by organizing and controlling LED fixtures, dimmers, and DMX rigs through a unified interface, simplifying complex lighting setups.

🎥 Sony's new VENICE Extension System Mini is significantly more compact and lightweight than its predecessor, allowing for enhanced mobility in a variety of filming scenarios without compromising the 8.6K image quality.