Didn't think this would work...

Directing Veo3 with diagrams.

(This is a little easier than JSON)

PROCESS:
01. Generate images + alt angles
02. Diagram your shot
03. Use Frames-to-vid in Veo3
04. Generate.

(More details in thread)

Nice unlock @Totemko

#veo3 #promptshare 1. Generate Image:

- Used Flux 1.1 Ultra in Weavy
- Created the base shot to play with

PROMPT:
A professional motorsport photograph of a Red Bull Dune Buggy captured during sunset in a vast beachside massive sand dune landscape with shrubs and layered terrain. The deep blue ocean is visible. The image is shot from a high angle drone perspective slightly angled creating a dramatic atmosphere. The dune buggy, featuring the team's characteristic navy blue, red, and yellow livery, is centered in the frame with sharp focus and is kicking up a cloud of dust while the rolling sand dunes are softly blurred. The white "Bull" logo is visible on the car's bodywork. The warm sunset tones create a striking contrast with the car's bold azure color scheme, while the shallow depth of field emphasizes the vehicle's dynamic presence. Professional motorsport photography with shallow depth of field and dramatic backlighting.

Midjourney Video → Keyframes Released.

It also works with "extend."

PROCESS:
01. Generate original image
02. Remove background.
03. Regenerate background.
04. Add both images as keyframes
05. Extend and continue.

(More Details in thread)

VIDEO PROMPT:
Side profile tracking shot of a matte green Land Rover Defender driving from a [environment1] into a [environment2]. Smooth cinematic transition. [ENV1] blurs into a [ENV2]. Motion blur on car wheels. [LIGHTING1] replaces [LIGHTING2]. Steady camera on vehicle, looping seamless movement --motion high

#midjourney #promptshare 1. Base Image.

- Create original image
- Side profile worked well for this one.

PROMPT:
Land Rover defender driving in a warehouse to the left, perfect side profile --ar 16:9 --profile wbkoiio zrmykh6 veclfmo

Come with me as we build...

A one-click AI-generated Photo Shoot Tool.

With Weavy.

(Full walkthrough in thread)

WORKFLOWS:
+ Model Designer
+ Clothing Inputs
+ Virtual Try On
+ Catalog Shots

TOOLS:
Workflow Build: Weavy
Nodes: Flux 1.1 Ultra, ChatGPT 4o, Gemini Flash 2.0, Magnific, ChatGPT Img

Pretty obsessed with this tool. 1. Model Designer

- First we need to build a model
- Create text nodes to control each model feature
- Run it to a "prompt concatenator"
- To combine prompt elements together
- Utilized Flux 1.1

Core Categories:
Gender, age, ethnicity, defining features, hair, body type, etc.

Ad Creative → GPT + Runway Ref + Flora.

Building usable workflows live.

Looked decent on the big screen.

PROCESS:
1. Reverse Engineer Images
2. Build System Prompt
3. Attach Ref image
4. Generate + Iterate.

(More details in thread)

#chatgpt #runwayml #promptshare 1. Start with the end.

First, we need to understand what we want.

WE WANT:
- To put the converse shoes
- In the composition of the black shoes
- So we can reconstruct the image.

Like the example below...

But how do we get there?

Runway References → Huge Upgrade.

Concepting meta ads in minutes.

Tools:
+ Iphone Photo: Reference image
+ Runway: Image gen
+ Figma: Ad composition

PROCESS:
+ Take/upload a photo to runway ref.
+ Write a text prompt
+ Generate + iterate
+ Create ad in figma

(More details in thread 🧵)

#runwayml #ecom @runwayml 1. Take / upload a photo

- I started with a photo from my iphone
- Uploaded into Runway
- Created a text prompt

(6) New Midjourney SREF Codes.

The mining addiction is back.

SREF CODES:
--sref 3848865494
--sref 917177894
--sref 1821837705
--sref 3068176932
--sref 2108485924
--sref 1484151111

Prompt:
{samurai, behance, mushroom, liquid 3d shapes, fire} --sref 3848865494 --ar 3:4

#midjourney #srefcodes #promptshare 2. --sref 917177894

Weavy → Aggressive YT Thumbnails.

Basically, AI Photoshop.

Love playing around in this tool.

BECAUSE I CAN:
+ Build in layers
+ Utilize like 100+ tools
+ Generate each element individually
+ See everything in one visual space
+ Generate and Iterate

(Check out the steps in thread 🧵)

#midjourney #promptshare @Weavy_ai 1. Base Characters:

- Generated base characters in MJ
- Utilizing oref and sref
- To make them weird and kind a awesome

Prompt:
pixar-style 3d caricature with exaggerated features, scared and screaming --oref [pic of person] --ar 1:1 --p veclfmo

Google Veo3 → I2V Prompting Guide.

Playing with "human expression"

Tips:
+ Guide the performance
+ Control facial expressions/gestures
+ It can handle finer details.
+ Don't overcomplicate the prompt

BASE PROMPT STRUCTURE:
[scene description], [expression/gesture description]

(More exploration in thread 🧵)

#google #veo3 #promptshare 1. Generate Your Character

- I used Google Imagen4 here
- Simple prompt, simple result

Prompt:
Medium shot of a 1980's character smoking a cigarette.

Google Veo3 → Extended Scene Guide.

Simpler than it seems.

PROCESS:
01. Gen initial video
02. Select "add-to-scene"
03. Select "extend"
04. Add text prompt + generate
05. Repeat

(More details in thread)

#google #veo3 #promptshare 1. Generate initial vid

- I used t2v for this gen
- You can use t2v or i2v
- Add your text prompt and run

PROMPT:
racing montage of a street race between a 1970's chevy camaro racing a 2025 lamborghini on a coastal road in calfornia. The scene is intense action from start to finish, with shot transitions showing the cars:

1. front tracking shot of both cars
2. transition to a closeup
3. transition to rear shot of the cars driving away

FLORA + CHATGPT → Product Dev. Concepts.

Palette to product variation.

PROCESS:
01. Upload product to Flora.
02. Upload color palettes
03. Connect via GPT image node
04. Add text prompt + generate.

Simple way to test.

(More details in thread 🧵🧵)

@florafaunaai #chatgpt #promptshare 1. Sneaker Gen.

- I'm using a MJ generated sneaker
- You can use w/e you want

PROMPT:
The sneaker features a blend of mesh and synthetic materials in its upper for breathability and comfort. The design incorporates strategic padding around the high-top ankle and a cushioned tongue for added support. A visible in the heel provides superior impact protection and cushioning, promoting a smooth ride. The lightweight foam midsole enhances cushioning while offering durability. In addition, the sneaker includes a rubber outsole designed with traction patterns for improved grip on various surfaces. A lace-up closure allows for a secure fit, and the heel counter is shaped to offer stability during movement. --no nike --chaos 5 --ar 1:1 --v 7

Google Veo3 T2V Prompting Guide V2.

Small changes for big iteration.

TIPS:
+ If you find a prompt that works...
+ Build a prompt structure
+ Play with small elements
+ Then iterate

BASE PROMPT STRUCTURE:
[establishing shot description], [subject details + environment], [energy/mood description], [environment details], [subject action description], [lighting], [style]

(Evolution in thread 🧵)

#google #veo3 #promptshare 1. Start with a structure

- Use a prompt structure
- Fill in the blanks + run
- So you can start to iterate

PROMPT STRUCTURE:
[Establishing shot description], [subject details + environment], [energy/mood description], [environment details], [subject action description], [lighting], [style]

VEO3 PROMPT:
The camera glides high above the sun-soaked cliffs of Monaco, nearly at helicopter-level. Below, a red-and-white 1980s Formula 1 car drives its way through the city circuit, a flash of polished paint and screaming power. Revealing a tightly woven tapestry of pastel architecture, narrow switchbacks, and the glittering blue of the Mediterranean beyond. The car is a blur of speed and precision—its silhouette dancing through hairpin turns and tunnel shadows, kissed intermittently by bright shafts of sunlight. The film look is raw and tangible: natural motion blur, gentle film grain, and sun-washed highlights mimicking vintage 35mm racing documentaries.

Google Veo3 T2V Prompting Guide.

For consistent iterations.

TIPS:
+ Start with a structure
+ Create a base prompt
+ Iterate small elements

BASE PROMPT STRUCTURE:
[PERSPECTIVE], [SHOT STYLE + DETAILS], [SUBJECT DETAILS + ENVIRONMENT], [SCENE DETAILS], [SOUND DESCRIPTION], [LIGHTING], [FILM STYLE]

(Evolution in thread)

#google #veo3 #promptshare 1. Base Generation

- Start with a shot you like
- Then begin to build off it
- So we'll use our prompt structure
- And create a prompt from it.

Ex Prompt:
A visceral close-up from a side-mounted camera on a 1980's red-and-white F1 car as it hurtles through the narrow, winding streets of Monaco. The shot hugs the car’s bodywork, capturing every tremor of the engine, every shimmer of the polished livery, and the blur of historic stone buildings flashing by just feet away. Tire sidewalls ripple with speed, while sun-glints bounce off chrome suspension arms. The sound is implied—intense and guttural. Light shifts rapidly as the car darts between sunlit straights and shaded corners, with brief flares from low-hanging balconies. The film look is raw and tangible: natural motion blur, gentle film grain, and sun-washed highlights mimicking vintage 35mm racing documentaries.

GOOGLE VEO3 → Wow.

Idk what I was expecting.

But I wasn't expecting sound + dialog.

Here's (10) examples + prompts 🧵.

Ex Prompt:
low angle selfie footage of a fat man running and laughing simultaneously. His laughs mix with his gasps for air. He has red hair, a mustache, a green sweat band sits on his head nestled next to his receding hair line with unkempt curly hair. The camera shakes as he runs with vigilance, but his laughter makes the scene light and comical

#google #veo3 #promptshare 2. Docu-footage.

Prompt:
A british rockstar with wild 80s-style hair and a studded leather jacket fumbles with a camera backstage at a rock concert, holding it in close like a snorricam rig. He looks directly into the camera, adjusting the focus, then smirks and says [Alright, is this thing on...let’s do it]. The image has a raw, handheld feel—his face inches from the lens, eyes wide with anticipation. Around him, the dimly lit stage prep area buzzes with scattered instruments and cables. Just before stepping into the light. The lighting is natural and moody, with a faint blue cast from the stage lights behind him. Texture-rich details in his jacket and the background add realism and grit amongst the backstage hum of the people around him.

RUNWAY REFERENCES → Template-to-Image.

Love how this works.

With a little help from Midjourney.

(More details in thread 🧵)

PROCESS:
01. Create Composition
02. Gen Reference Images
03. Upload Ref images
04. Add text prompt + iterate

So much you can do with this.

#midjourney #runwayml 1. Create Template

- I used figma for a simple composition
- Added some pose direction
- Uploaded into Runway

FLORA + Gemini → Multi-Character Scenes.

(Super basic workflow 🧵)

PROCESS:
01. Generate (2) Characters
02. Upload into Flora
03. Use Gemini to place characters
04. Iterate shots

This isn't the most exciting scene.

But the process is replicable.

#google #promptshare 1. Generate Characters

- I used Midjourney
- Pulled images from my library
- Nothing Crazy

ChatGPT 4o → Creating Fonts.

I gave it my logo...it created a font.

(More details in thread)

GPT PROCESS:
1. Provide ChatGPT reference image.
2. Ask for it to create a font
3. Upload image.
4. Ask to test font on image.

FONT CREATION PROCESS:
+ Vectorize + trace this into Illustrator
+ Build font file (.OTF/.TTF)
+ Then start testing weights

#openai #promptshare 1. Created Ref Image

- Removed logo color
- Aligned it for easier readablity

HOT TAKE...😱

Image Curation > Prompting...in Runway.

Using the wrong images = headaches.
Using the right images = saved credits.

🧵 Here’s a few tips ↓

#runwayml #runwaygen3 1. For Starters:

+ Reference Image = Huge output influence
+ Poor image selection = Poor vid generation.
+ Static images + motion = Better vid motion
+ Higher Res images = Sharper definition