đī¸ Tip 1: Imitating
I think the best way to learn image prompts, similar to learning to draw, is not to use templates directly. Instead, use real images or images generated by others to imitate. If your English is not good, you can first write in Chinese and then have ChatGPT translate it. After imitating several images, you will gradually understand how to create similar images.
đī¸ Tip 2: Experimenting
The above example also demonstrates a second technique - experimenting.
đī¸ Tip 3: Using Image2Image
For stock photos, there is a very powerful technique. At first I didn't want to teach this because it has a big impact on stock photo sites đ
đī¸ Tip 4: Adding Style - Art Movements
For length I won't cover many styles here, but I'll share more on my Twitter - follow me there for updates.
đī¸ Tip 5: Adding Style - Artists
For length I won't cover many styles here, but I'll share more on my Twitter - follow me there for updates.
đī¸ Tip 6: Use "no" to remove unwanted elements
In the badge logo examples, you may have also seen single-color emblems like school crests. Midjourney tends to generate complex outputs, so for simplicity you can add color, background, etc. to the prompt.
đī¸ Tip 7: Using Multiple Parameters
When generating avatars with img2img, I found the issue was "text weight is higher than image weight", so the outputs didn't match the original. With iw, V5 caps image weight at 2. So I tried using the s parameter, and it improved results a lot.
đī¸ Tip 8: Modifying Images Using Seed
Note: I think this technique has potential but is currently unreliable in Midjourney. The official help docs also mention seeds are very unstable in V5. See my Midjourney FAQ chapter.
đī¸ Tip 9: The Mysterious Blend Feature
To be honest, I hesitate to call this a technique - it's a very unstable Midjourney feature. But it's quite important so I'll introduce it.
đī¸ Tip 10: Iteratively Improve via Controlled Variable Changes
Many compare AI image generation to alchemy. It's true that small mystery tweaks in prompts can greatly change the output.
đī¸ Tip 11: Add Styles - Countries
Speaking of figures, Japan likely comes to mind first. Here's a technique - add "Japanese style" to get:
đī¸ Tip 12: Increase Weight
Notice how in the Chinese figures, only the 3rd looks Chinese? That's because the low weight of "Chinese style" resulted in only one output. You'll encounter ignored prompt words too, like mentioning a bird but none in the image.
đī¸ Tip 13: Use Lighting
To be upfront, I'm neither a designer nor photographer (and very bad at it), so can only share basics here. The details are too applied - hard to grasp without hands-on experience.
đī¸ Tip 14: Add Styles - Time Periods
The old photos used 1990s style - strictly speaking, decades fit better in subject or environment, but works in style too since we want both 90s subjects and a 90s style.
đī¸ Tip 15: How to Generate More Diverse People
Diversity here refers to people's height, weight, body types, disabilities etc.
đī¸ Tip 16: Change Camera & Lens
Covered lighting the past few chapters - now let's look at cameras and lenses.
đī¸ Tip 17: Reconstructing Prompts
Simplest way is just asking the creator.