July went well.

We started the month with an ambitious list, built most of what we set out to build, changed a few things along the way, and left a few others for later.

More importantly, July gave us a format we want to keep. The RanGen Build Log is now a monthly series.

At the beginning of each month, we’ll share what we’re building, testing and improving inside RanGen Studio. Then, as the month gets closer to its end, we’ll come back to the same Build Log and update every feature with where it actually landed.

Live. Beta. Testing. Still building. Carried forward. Whatever the real answer is.

For August, a lot of the work is connected by one bigger goal:

Characters → Looks → Scenes.

We want the identity you establish for a character to survive further into the production process — through wardrobe, approved looks, casting and eventually the scenes you generate.

Here’s what we’re working on.

1. NEW CHARACTER REFERENCE SYSTEM · LIVE IN BETA

Characters are becoming more structured inside RanGen Studio.

Until now, having a character reference was useful, but we wanted considerably more control over which references actually define that character.

The new Character Reference System is our attempt to make that clearer. You can build a more structured character profile, manage multiple visual references and select the images that should matter most when the character moves further into the production workflow.

We’re also introducing the ability to lock a character once you’re happy with the identity.

The idea is simple: Once you decide who this person is, you shouldn’t have to keep rediscovering them every time you generate a new scene.

Where it is now: The new system is already available in beta.

We’re testing how reference selection, profile structure and character locking affect consistency further downstream. The bigger goal is stronger visual continuity from one generated scene to the next.

2. WARDROBE & OUTFIT MANAGEMENT · LIVE IN BETA

A consistent character is only part of the problem. What are they wearing? Which outfit belongs to which character? Which outfit was approved for a particular sequence? And how do you stop searching through old generations every time you need that same look again?

This month, we’re improving the wardrobe system around those questions. Outfits can be created, organised and associated directly with characters.

That gives us a clearer relationship between:

Character → Outfit → Approved Look

rather than treating every piece as an isolated image.

Where it is now: The improved wardrobe workflow is already being tested.

There’s still refinement to do, but the foundation is there: fewer disconnected references, clearer character relationships, and a more production-ready way to organise looks.

3. LOOK ROOMS & LOOK CARDS · ROLLING OUT

This is where the new character and wardrobe systems start coming together.

A Look combines an established character identity with an approved outfit to create something reusable. Instead of telling the system: "This is Ethan." And then separately: "This is the suit Ethan should wear." And then hoping those two things remain connected later...

...you create an approved production look.

One character.
One outfit.
One consistent visual reference.

That can then follow the character into scene generation.

Where it is now: Look Rooms and Look Cards are beginning to roll out as the surrounding character and wardrobe systems stabilise.

We’re treating this as an important bridge between character creation and actual scene production. Because ultimately, an approved character isn't very useful if they become somebody else the moment you put them into a scene.

4. MORE CONSISTENT SCENE GENERATION · IN TESTING

This is where a lot of the August work starts paying off.

Character profiles, references, wardrobe and Look Cards aren't meant to exist as isolated management tools. They need to affect generation. So we’re integrating the new character-appearance system more deeply into scene generation.

When preparing a scene, you’ll be able to select the characters involved and choose the appropriate approved reference or Look Card for each one. The goal is stronger continuity across shots. Not merely generating someone who loosely resembles the same person.

We want the same character to continue feeling like the same character.

Where it is now: The integration is currently being tested.

This is one of the parts we’re going to be watching closely through August because consistency doesn’t come from one setting. It comes from how several systems communicate with each other:

Character identity
References
Wardrobe
Looks
Casting
Scene generation

August is largely about connecting those pieces.

5. CHARACTER–BACKGROUND COMPOSITION · IN DEVELOPMENT

Consistency isn't only about what a character looks like. It also matters where they appear inside the frame.

A character can be perfectly recognisable and still produce a weak shot if their scale, placement, depth or relationship with the environment feels wrong. So we’re working on better character–background composition.

The goal is to make characters sit more naturally within generated environments while still respecting the intended story, scene context and composition.

Better scale.
Better placement.
Better depth.
Better interaction with the environment.

Where it is now: This one is still being built.

We’re experimenting with how much composition control should come from the filmmaker and how much the Studio can infer from the context of the scene. It needs to be controllable without turning scene generation into a control panel with fifty sliders. That balance is part of what we’re figuring out.

6. MICRODRAMA MODE · IN DEVELOPMENT

Vertical storytelling needs different thinking.

A composition designed for a widescreen cinematic frame doesn’t automatically become good just because you crop it to 9:16.

Characters sit differently.
Negative space works differently.
Depth feels different.
Close-ups behave differently.
Even the way attention moves through a frame changes.

So we’re continuing to develop a more robust Microdrama Mode specifically around vertical and short-form storytelling.

The goal isn't simply: Generate horizontally → crop vertically. The vertical frame should be intentional from the beginning.

Where it is now: We’re currently improving framing, character placement, composition and depth for vertical scenes.

There’s more work to do here, but we think vertical storytelling deserves its own production logic rather than being treated as an export setting.

7. MULTI-PASS LOOPING EDITS · PLANNED

Editing a generated result often becomes repetitive. Change the composition. Wait. Change the character. Wait. Fix the environment. Wait again.

But those instructions are usually connected. So we’re working toward Multi-Pass Looping Edits.

Instead of making each correction independently, you’ll be able to create a sequence of connected editing instructions. For example:

Pass 01: Adjust the composition.
Pass 02: Refine the character appearance.
Pass 03: Correct the environment.

One editing brief. Multiple connected refinements.

Where it is now: This hasn't reached the same stage as the character systems yet.

It's on the August build list and represents the direction we want editing to move toward: less repetitive prompting and more deliberate iterative refinement.

8. RANGEN MOBILE · IN PROGRESS

Mobile continues from the work we started in July.

RanGen already works through the mobile browser, and the Android APK has given us another way to test what the Studio feels like away from a desktop. But we still have more work to do before the dedicated mobile experience is where we want it.

The goal isn't to squeeze every desktop panel onto a smaller screen. Some RanGen workflows make much more sense on mobile than others:

Projects
Assets
Characters
Reviews
Tasks
Quick actions
And eventually some of the capture workflows we talked about in July.

Where it is now: The mobile apps are still being prepared.

We're continuing to figure out which parts of the Studio should travel with you and which parts are better left to a larger production workspace.

9. TEAM & ENTERPRISE ROLES · IN DEVELOPMENT

RanGen Studio is increasingly being used as more than a single-person creative workspace. And as soon as multiple people enter the same production, permissions become important.

A director doesn't necessarily need the same access as an accountant.
A producer isn't doing the same job as a designer.
A contributor shouldn't automatically have administrative control over an entire workspace.

So we're building a clearer role hierarchy for teams and production environments. Studio administrators. Creative directors. Producers. Designers. Coordinators. Contributors. Viewers. Guests. And more specific access around what each role can create, manage, review or see.

Where it is now: The role and permissions system is still being developed.

This becomes increasingly important as RanGen moves from being a collection of filmmaking tools toward becoming a workspace that an actual production team can share.

What August is really about

On the surface, this month's list contains a lot of separate features: Character references, Wardrobe, Look Rooms, Scene generation, Composition, Microdramas, Editing, Mobile, and Team permissions.

But most of them point toward the same thing.

A more structured production workflow.

A character shouldn't stop being a character once you generate a scene.
An outfit shouldn't disappear into an asset folder.
An approved look should remain useful later.
A scene should understand the creative decisions that came before it.
And the people working together on a project should be able to share that process without everyone having access to everything.

That's the direction we're building toward.

Characters → Looks → Scenes

One decision feeds the next.

We’ll update this Build Log before August ends.

This is the beginning-of-month version. As August gets closer to its end, we'll come back to this same post and update every section with where things actually landed.

Some of these may be live.
Some may still be in beta.
Some may need another month.
And something we're currently building may turn into something completely different along the way.

That's why we're keeping the Build Log. Not to show the roadmap as we imagined it, but to show what actually happened when we tried to build it.

Build. Test. Learn. Improve. Repeat.

Still making. Still shipping.

Keep following the build.

The August Build Log