Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Umesh Profile picture
Umesh
@umesh_ai
Creator, Prompter, Developer | Built https://t.co/UbiUnFZVjS
Sep 7 • 8 tweets • 2 min read
Camera movements are an essential part of any video!

Motion 2.0 from @LeonardoAi_ offers more than 20 motion controls.

Here are some that come in handy most often. ⤵️

1. Crane Overhead 2. Super Dolly In
Aug 10 • 7 tweets • 6 min read
Seeing is believing!

But these images aren’t real

Lucid Origin by Leonardo AI is incredible with light, colours, and textures.

Here’s a mega thread with 24 examples and prompts. Arancini crack & melty core 100–105mm macro, f/7.1, ISO 100, 1/200s. Hard side key at 35° to dramatize breadcrumb crust; warm-gelled strip rim from rear to kiss mozzarella stretch; silver bounce low-front; negative fill opposite. CPL off for full specular bite. Pea flecks and saffron grains visible in rice matrix. Focus stack 8 frames from fracture edge to cheese ribbon. Matte charcoal plate, faint sauce smear as controlled accent. Palette: amber crumb, ivory melt, emerald peas, onyx. No hands, no text.
Charred street-corn kernels with chili-lime crema ribbons and cotija crumble, macro on matte charcoal plate. 100mm macro at f/7.1, ISO 100, 1/200s. Overhead diffused key; back rim with warm gel to kiss char blisters; white card front for lift. Sprinkle cilantro micro-leaves as directional flow; lime zest dust for sparkle. CPL adjusted to balance crema glare. Focus stack 6 frames across leading kernels. Palette: maize yellow, chili red, verdant green, chalk white against deep gray. Gentle clarity and micro-contrast; avoid heavy sharpening halos. No hands, no text, no messy drips beyond design
Honeycomb dripping onto black marble, extreme macro of hex cells. 90–105mm macro at f/9 to hold structure; ISO 100, 1/200s. Backlit stripbox for amber glow, small snooted kicker to ignite edges; white card below for internal bounce. Slow pour captured mid-thread; shoot multiple to freeze a perfect strand. CPL partially engaged for balanced reflections. Warm color palette: amber, gold, espresso. Focus stack 10 frames from foremost cell to meniscus. Keep background minimal gradient to near-black. Tiny pollen specks visible. No bees, no labels. Surfaces spotless, high-contrast, editorial polish.
Pretzel salt & lye gloss 100mm macro, f/8, ISO 100, 1/200s. Hard specular: snooted strobe from 25° left to punch crystalline salt; second strip rim warm from rear; minimal top diffusion to preserve crisp edges; black flags to deepen valleys. CPL at 30° to control lye sheen while keeping salt sparkle. Focus stack 7 frames from leading salt cluster into crust creases. Palette: mahogany crumb, alabaster salt, matte tin sheet. Mustard whisper blurred behind. No hands, no text, controlled crumb fallout. All prompts are in ALT of each image.
[5-8] Baklava filo strata & syrup threads 90–105mm macro, f/8, ISO 100, 1/200s. Hard backlight stripbox to ignite filo translucency; overhead octa double-diffused for soft top sheen; black cutters left/right for depth; white bounce low-front. Capture pistachio grit, toasted edges, micro syrup filament mid-drip (high-speed pop). Cross-polarization half-on to tame glare without killing sparkle. Focus stack 10 frames from top shard to nut bed. Palette: honey gold, pistachio green, warm chestnut. Matte obsidian plate, crumb scatter controlled. No hands, no text, no powdered sugar fog.
Berry medley with mint granita shards 100mm macro at f/7.1, ISO 100, 1/200s. Mixed-source look: cool overhead diffused key + warm gridded rim to separate frost from berry gloss; silver bounce low-front; negative fill left. CPL off to keep crystalline sparkle on granita. Glycerin mist for controlled dew. Place single powdered-sugar kiss as focal anchor. Focus stack 8 frames from front strawberry seed crown to blueberry bloom. Matte basalt plate, deep navy backdrop gradient. Palette: carmine, indigo, emerald, snow. Clean, editorial finish. No hands, no text.
Burrata torn open with slow-oozing cream, cherry tomatoes confit, basil oil dots. 100mm macro at f/6.3, ISO 100, 1/200s. Key: top-left 60cm octa; bounce: white foam core front; flag: black card right for depth. Slate blue backdrop for complementary palette. Oil droplets arranged as leading line; salt flakes and cracked pepper precisely placed. Capture micro-texture of stracciatella strands. Focus stack 7 frames across tear line. CPL for porcelain highlights, dialed to retain tomato gloss. Minimal prop: thin antique knife tip, out of focus. True whites, gentle contrast curve. No hands, no te...
Ceviche spoon: sea bass cubes, red onion curls, cilantro, leche de tigre 90–105mm macro, f/7.1, ISO 100, 1/200s. High-key overhead through 2-stop diffusion; narrow snoot behind to ignite citrus translucency; white cards all sides; subtle negative fill under spoon bowl. Cross-polarization half-on to balance wet glare. Chili oil micro-dots as color pops. Focus stack 6 frames from nearest cube to onion arc. Stainless spoon on matte charcoal tile. Palette: pearl, lime,
Jul 21 • 7 tweets • 3 min read
A simple guide to creating stunning phone wallpapers with Lucid Realism model on @LeonardoAi_

Every word in a prompt contributes to the visual outcome. Choosing the right words is the key.

Here are 7 examples with a generic prompt and tips to create countless images. 🧵 Frame a quaint stone cottage encircled by blooming lilacs in soft morning sunlight. Lilac petals vary from delicate mauve to creamy lavender, each meticulously textured. Warm sunrays softly penetrate through flower clusters, illuminating stone walls etched with ivy. Foreground petals blur gently into dew-speckled bokeh, establishing subtle icon placement zones in lower-middle frame. Behind the cottage, blurred verdant meadows and softly defined treelines fade into pastel dawn skies, accentuating graceful tranquility. Palette: lavender hues, stone grays, leafy greens, and faint gold, effortl... Prompts for each image are provided in the ALT text.

While each LLM excels in its unique way, I used ChatGPT 4.5 to generate prompts due to its exceptional writing capabilities.

The following generic prompt is designed for wallpapers inspired by 18th-century countryside life. Capture late‑summer afternoon when a wooden cart piled high with sheaves lumbers across amber stubble. Sunlight grazes the hay, throwing specular flecks; long cart shadows form a natural vignette that darkens icon territory. The horizon sits high, letting half‑empty foreground breathe. Leather reins, iron‑rim wheels, and sweat‑glistened draft horse musculature are rendered with painterly realism. Palette: warm ochres, dusty siennas, powder‑blue sky. Leave dust motes drifting lazily to suggest motion without visual fatigue. Unlocking the phone feels like revisiting a Turner‑esque harvest scene
Apr 9 • 11 tweets • 3 min read
ChatGPT-4o is a phenomenal model from OpenAI.

I've been experimenting with it and discovered many amazing uses.

Here are the best ones.

Bookmark this for future reference! 🔖 Image 1. Turn yourself into a sculpture image.

Nov 12, 2024 • 6 tweets • 2 min read
It's easy for anyone to make simple animated videos!

I created this in around an hour.

Tools used :
Images : Story Illustrator app
Videos : Hailuo AI
Audio : ElevenLabs
Editing : InShot

Detailed workflow ⤵️ The Story Illustrator app (makemycomics) helps create images with multiple characters that stay consistent.

You can try it out by visiting my profile.
Nov 6, 2024 • 7 tweets • 3 min read
You can create short animated films with consistent characters in no time!

I created this in less than an hour!

The best part: everything is done on my smartphone!

Tools used:

Images: Story Illustrator App (MakeMyComics)
Video: Hailuo AI
Audio: ElevenLabs
Editing: InShot

Detailed workflow ⤵️ The Story Illustrator app (created by me) can create stories with fairly consistent characters, generating both story and visuals.

You can access it from the link below:



Let’s dive into the steps ⤵️:makemycomics.com
Jul 5, 2024 • 5 tweets • 2 min read
This is amazing!

I created this image without worrying about how to prompt!

Here’s how! ⤵️ Image This was done using “getimg AI’s” brand new ‘Presets’ feature!

Check out this walkthrough video!
Mar 30, 2024 • 6 tweets • 3 min read
Ok, This was unexpected!

I discovered a way to duplicate characters!

You can take any image (left) and generate images like these on the right!

Here is how ⤵️ Image While playing with Midjourney, one of my prompt produced multiple identical characters with --cref as one of the parameter.

So I refined the prompt to get a group of Monalisa sculptures. Image
Feb 20, 2024 • 5 tweets • 2 min read
Given a plot (or image), it is possible to generate an entire story with illustrated images that are fairly consistent with characters and visuals!

✨ Presenting my new GPT: Story Illustrator! 🎨

Features:

1. Describe a plot for your story in a few words or upload any image, and the GPT will construct the story and create consistent visuals.

2. Choose the artistic style for the illustrations or specify your own!

3. You can specify the number of images to generate.

All images will have a consistent style, characters, settings, and a brief narration of the story.

This would be a great fit for kids, parents, teachers, and anyone who likes to have a fun story made with illustrations.

Some examples and a link to the GPT below!

❤️🤖🎨Plot: A mother helps her baby learn to walk. Style : Watercolor painting style Story : Summary: A mother tenderly supports her baby's first hesitant steps, encouraging each small achievement with love and pride, culminating in the baby's joyful first independent steps. First Stand: The baby stands with the mother's support, looking curious. Speech bubble: "Up we go!" Tiny Steps: The baby takes cautious steps, guided by the mother's hands. Speech bubble: "You got this!" Faltering Move: The baby wobbles slightly, the mother ready to catch. Speech bubble: "I'm here!... Link to GPT.

chat.openai.com/g/g-BnUlAisOD-…
Nov 4, 2023 • 6 tweets • 2 min read
🚨 DALL-E 3 has released a new update! 🚨

Now, you can create an image that follows the style of any previous image!

🔖 Bookmark for later use.

Here's are the steps I followed: Image 1. I started by requesting GPT-4 to create a pencil sketch of a house.

2. It responded with two images, and I noted down the gen_id of each image.

Exact request and response provided in ALT
{ "request": { "size": "1024x1024", "prompt": "Create a detailed pencil sketch of an old Victorian-style house with a steep gabled roof, a prominent front porch with ornate wooden trim, and a towering chimney. Include surrounding details like a wrought iron fence, a cobblestone pathway leading to the house, and a few mature trees with intricate branch patterns. The sketch should have a haunted, eerie feel, with shadows and light creating a dramatic contrast, hinting at the house's long history and the many stories it might tell." ...
{ "gen_id": "7N58tmp5GF2aobK7", "prompt": "Create a detailed pencil sketch of an old Victorian-style house with a steep gabled roof, a prominent front porch with ornate wooden trim, and a towering chimney. Include surrounding details like a wrought iron fence, a cobblestone pathway leading to the house, and a few mature trees with intricate branch patterns. The sketch should have a haunted, eerie feel, with shadows and light creating a dramatic contrast, hinting at the house's long history and the many stories it might tell.", "...
Aug 25, 2023 • 4 tweets • 3 min read
📚 Midjourney Guide : Generating Consistent Characters with Different Expressions 🎭

🎨 This guide explains two methods to generate consistent characters with different expressions.

1️⃣ Using Vary(Strong)

2️⃣ Using Inpainting with Split Images
(Inspired from @chaseleantj ) Image Part 1: Using Vary(Strong)
1️⃣ Write a prompt to generate the desired image of the character.
2️⃣ Upscale one of the images.
3️⃣ Utilize the Vary(Strong) feature to modify the character's expression.


Image
Image
Image
Image