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
Chase Lean Profile picture
@chaseleantj
Oct 11, 2023 • 12 tweets • 4 min read • Read on X
Scrolly
A few hours ago, Adobe released Firefly 2, their new AI image generator.

But how good is it?

I compared it to Midjourney & DALL-E 3 across 10 very different image categories so you don't have to.

Here are the results:
1/10 Realistic photos

Firefly 2 is super good at generating photos.

The shots are more realistic and have higher resolution than MJ, but both are good.

Meanwhile, DALL-E 3 lags behind. Image
2/10 Product photos

I'm pretty satisfied with all three.

I might prefer the Firefly one a bit more, because of the higher contrast making the perfume bottle stand out. Image
3/10 Interior designs

Again, all 3 generators did quite a good job here.

The Midjourney image seems to have better lighting, while Firefly made it more luxurious, following the prompt closely. Image
4/10 Text generation

DALL-E 3 wins.

End of story. Image
5/10 Landscape photos

The lighting on the Midjourney image is epic.

It wins this one.

Firefly is a close second, while DALL-E 3's image... would have been impressive last year. Image
6/10 Closeup shots

Firefly's image has incredible detail.

It really shines on photos, especially those of people.

MJ is good as well, although the style is less realistic and more suited for a fantasy film. Image
7/10 Vector graphics

This one is interesting.

All three have their unique styles and strengths.

I'd say DALL-E 3's result looks more like conventional vector illustrations. Image
8/10 Pixel art

This one is highly inspired by @ProperPrompter.

DALL-E 3 is by far the best, listening to the prompt and making chibi characters, pixel art and item sets at the same time.

Firefly managed to get the pixel art right, while MJ's one isn't even pixelated. Image
9/10 Wildlife photos

Unsurprisingly, Firefly does extremely well.

So does Midjourney.

I'd say they're evenly matched. Image
10/10 Coloring pages

Both Firefly and DALL-E 3 are good choices for making coloring pages for very young children.

The ones by Midjourney are more beautiful and artistic - suitable for slightly older children. Image
I hope you found these examples helpful!

I'll be sharing more insights on Firefly, DALL-E and Midjourney very soon.

1. Follow me at @chaseleantj so that you won't miss it.

2. Like and repost the first post below to share it with others.

• • •

Missing some Tweet in this thread? You can try to force a refresh
　

Keep Current with Chase Lean

Chase Lean Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @chaseleantj

Chase Lean Profile picture
Jun 11, 2024
PixVerse just released a powerful new AI tool called Magic Brush!

It allows you to control how objects move in videos, and has some very interesting features.

Let me demonstrate with 7 examples + how to try it for free 🧵
1. Cars racing

Notice how you can make the car on the left move faster by drawing a longer arrow.
2. Colored balls

You can specify the objects to move in up to 6 different directions.

Each brush corresponds with one arrow.
Read 9 tweets
Chase Lean Profile picture
May 14, 2024
A few hours ago, OpenAI released their latest model, GPT-4o.

It's by far the best model we've seen so far, and it's available for free.

Here are 10 incredible things it can do:
1. It can accept text, audio, and images as input, and it is FAST.

The voice mode has a latency of ~300ms, 10x faster than GPT-4.

This means it can translate in real time:
2. You can get it to participate in a meeting.

It can listen to the conversation, and even speak for itself.
Read 12 tweets
Chase Lean Profile picture
May 10, 2024
Coloring book images generated with AI usually have 3 problems:

1. They are too detailed for young children
2. They have shading
3. It's unclear how to make a colored version as a reference

Here's a solution I found: Image
To solve problems 1 and 2, try the AAM XL model on @getremixai.

This model usually gives simpler images vs other generators.

For example, left: AAM, right: Midjourney.

It is also an anime model, which means that there is no shading.
children's coloring book image of a girl riding a unicorn in space, beautifully colored, thick lines, simple drawing, plain, minimalistic
children's coloring book image of a girl riding a unicorn in space, beautifully colored, thick lines, simple drawing, plain, minimalistic
To solve problem 3, we'll use a trick.

First, generate an image WITH color.

Here's a prompt you can use:

children's coloring book image of a [object], beautifully colored, thick lines, simple drawing, plain, minimalistic

Negative prompt:

shading, intricate details, bad, ugly, distortedImage
Read 5 tweets
Chase Lean Profile picture
May 4, 2024
ChatGPT can now be used inside Microsoft Excel or Google Sheets.

No more memorizing formulas or functions.

Here’s how to use it to reduce work hours: Image
Numerous AI lets you create ChatGPT prompts in Microsoft Excel and Google Sheets.

It's an AI add-on and you can check it out at .

Here's how it works: Numerous.ai
AI is changing the game when it comes to improving productivity.

Here’s how you can use ChatGPT in Excel and Google Sheets with @NumerousAI:
Read 7 tweets
Chase Lean Profile picture
Apr 9, 2024
I’ve shown you guys how to make flat vectors with AI.

But what if you already have a vector you like, but it doesn’t suit the theme of your website?

Here’s a trick to change a vector to any theme you want: Image
To demonstrate, I’ll be converting these two vector illustrations into a common theme.
Image
Image
First, upload the image to @rendernet_ai and turn on ControlNet.

This helps keep the subject and the pose consistent, while allowing you to change the style. Image
Read 7 tweets
Chase Lean Profile picture
Mar 16, 2024
Midjourney's new character reference feature is also great for making product photoshoots.

Now you can take different photos of the same product.

Here's a guide to get the best results:
Image
Image
First, a quick overview:

To use the character reference feature, add --cref at the end of your prompt, followed by the image link.
For this demo, let's generate a few different product photoshoots.

I'll be using this (real) image below as a reference product. Image
Read 11 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

Send Email!

:(