🚀 Midjourney Office Hours Highlights: March 19th

We're halfway through March and still no V7. I suppose we'll have to wait a bit longer. However, the office hours have been enjoyable, offering some great insights.

The internal testing of the stable V7 build is scheduled for this week, with a potential public release as early as late next week, though a backup V6.2 release is prepared in case of delays. Key features of V7 include optimization of memory usage and processing speeds, the introduction of the omni-reference feature to enhance prompt versatility, and considerations for a "mega" model with enhanced capabilities. Additionally, the team is progressing on video model development, seeking community feedback, and working on aesthetic improvements, with plans to revisit 3D model development and storytelling enhancements in the future.

For a quick overview, head straight to the TL;DR; section otherwise keep on reading.

Extended Recap

⏰ V7 Release Timeline: What's Next?

The internal testing of a stable V7 build is slated for this week, involving guides and moderators. If all goes well, the public release could happen as early as late next week. However, potential delays might occur, and if they do, a backup V6.2 release featuring omni-reference and aesthetic updates is on the cards.

⚙️ Optimization and Performance Enhancements

Optimizing memory usage and processing speeds is a top priority to boost V7's performance. Recent trials have revealed some trade-offs in image quality, but a new approach has proven effective. The final optimizations and omni-reference tuning are expected to be completed in the coming week.

🔄 Omni-reference Feature: A Game Changer

The omni-reference feature is in its final tuning stages and promises to significantly enhance prompt versatility. Testing is scheduled soon, with potential rollout to guides and moderators within days.

💡 V7 Premium (“Mega” Model) Considerations

The team is exploring the possibility of a "chunky" or "mega" V7 model with enhanced capabilities at different pricing tiers. Initial attempts faced data contamination challenges, leading to a delay in further development. However, future training for higher-performance, data-rich models is planned for later this year.

🎥 Video Model Progress and Community Engagement

The development of a small-sized video model is complete, and training for a medium-sized model is about to commence. If successful, a release could be expected in 1–2 months. The team is also actively seeking community feedback through a survey on Discord to gauge interest and priorities, which will inform future video feature releases.

🎨 Aesthetic and Editing Improvements

Efforts are underway to calibrate aesthetics, focusing on age- and gender-neutral outcomes. Recent enhancements to the external image editor, including improved selection tools, are now available for users.

🖼️ 3D Model Developments on the Horizon

Following the release of either the image or video models, the team plans to return to 3D development. New dataset preparation is in progress, although timelines and priorities for 3D features are yet to be finalized.

📚 Storytelling and Interface Innovations

Experimental storytelling enhancements are in the pipeline to boost user creativity and content generation. Potential interface upgrades are also being discussed to leverage the expected speed improvements with V7.

TL;DR

🚀 V7 Development: Features nearly finalized, focusing on image quality and prompt understanding. Tackling data contamination issues.

⏰ Release Timeline: Internal testing this week; public release possibly late next week. Backup V6.2 if delayed.

⚙️ Optimization: Enhancing memory and speed; final tuning next week.

🔄 Omni-reference: Final tuning; rollout to guides/mods soon.

💡 Premium Model: Exploring advanced "mega" model; future training planned.

🎥 Video Model: Medium-sized training soon; community feedback sought.

🎨 Aesthetic Improvements: Age/gender-neutral focus; enhanced editor tools available.

🖼️ 3D Developments: Post-image/video release; new dataset prep underway.

📚 Storytelling/Interface: Enhancements planned; discussions ongoing.

Stay tuned for more updates as Midjourney continues to innovate and enhance its platform!

If you want to support me, feel free to buy me a coffee ☕️ 

If you’re not subscribed yet to my newsletter Imagine Weekly, I’d be thrilled to welcome you on board!