Taking Control of Image Angles: Tips for Creative AI Images

In this newsletter, read about:

• 🕵️‍♀️ Tips For Controlling Image Angles

• 🗞 News and Top Reads

• 📌 AI Art Tutorial: New Midjourney Features

• 🎨 Featured Artist: Fabio Comparelli

• 🖼 AI-Assisted Artwork of the Week

• 🤓 How to Get Started with Generative AI?

🕵️‍♀️ Tips For Controlling Image Angles

The latest AI image generation tools are capable of creating stunning photorealistic images, but when you have a very specific image in mind, controlling the output is still very challenging. One of the aspects that creators often want to manage is the camera angle for a generated image.

In this piece, I want to focus on controlling the image angle using your text prompt only. Of course, there are ways to guide the image generation via uploaded reference images or using the plethora of tools available in the Stable Diffusion ecosystem, but actually, it’s also possible to achieve a certain level of control relying exclusively on your prompt. Let’s explore this possibility together!

Camera angles

To capture truly stunning images, it's crucial to understand the impact of camera angles in photos. Whether you're generating landscapes, still images, or portraits, angles can make all the difference between a mediocre and a masterpiece shot.

ChatGPT provided me with the following descriptions of the most common camera angles:

1. Eye-level: A straight-on angle, shot at the same level as the subject, creating a natural, relatable perspective.

2. Low angle: A view from below the subject, looking up, which can make the subject appear more prominent or dominant in the frame.

3. High angle: A view from above the subject, looking down, which can make the subject appear smaller or less significant in the frame.

4. Bird's-eye view: An overhead view, shot directly from above the subject, providing a unique perspective and often showcasing patterns or symmetry.

5. Worm's-eye view: A view from the ground, looking up at the subject, creating a dramatic and unconventional perspective that emphasizes height and depth.

6. Wide-angle: A shot taken with a wide-angle lens, capturing a larger field of view and creating a sense of space and depth in the image.

It’s time to see how responsive Midjourney is to the corresponding keywords.

Experiments

For the experiments, I chose a minimalist still image that showcased a vintage camera and a pineapple. To my surprise, it proved to be a challenging task for the AI model, as it struggled with getting the relative sizes right and occasionally added some unexpected and unidentifiable elements to the composition. Nonetheless, after several attempts, I managed to get a few good images. Let’s see!

I’ve started with an eye-level shot, which is pretty straightforward and easy to get.

I love the colors in this one!

Next, I moved to low-angle shots, and this was not that easy. As you can see below, I’ve added an extra weight for a “low-angle” keyword (with ::2), because usually this part was ignored by Midjourney. Another tip is to use “from below” in a text prompt, in addition to “low-angle”. This often helps, but still, not always.

It looks like a low angle, but the camera seems to be very unusual. I doubt such one exists 😃 

High-angle images turned out to be similarly challenging as low-angle ones. I was able to get one image with a high-angle view, but it took me several attempts.

Here again, the recommendation is to experiment with adding “from above” to the text prompt, but from my experience, this combination tends to generate bird’s-eye view shots, which I wanted to explore separately.

The bird’s-eye view images turned out to be the easiest to get from Midjourney. The tool generates corresponding images without any additional tricks.

In my opinion, the bird’s-eye view images tend to be the most eye-catching, and this one is not an exception.

But as we learned from ChatGPT, there is not only a bird’s-eye view but also a worms’s-eye view. Does it work as well as a bird’s-eye view? Unfortunately, not.

After several attempts, I was able to get this image, which looks like a worm’s-eye view to me, but it also has a very strange mini camera lens in the composition. Variations of these images resulted in other unidentifiable objects. So let it be! 😃 

Finally, I generated a wide-angle image. From a dozen of generated images, only the one below looked like a wide-angle shot to me. So, this is also a challenging case for Midjourney.

Seed parameter

In the last part of this series, I’ve been experimenting with a seed parameter. This parameter defines a starting point to generate the initial image grids.

Let’s say you’ve generated an image that you like in general, but want to change some details. You can check the seed parameter of the corresponding image set by reacting with an envelope ✉️ emoji to the output image grid or one of the images after upscaling.

So first, I’ve generated the following low-angle image by adding “bright colors” to the prompt.

I liked the image and requested a seed parameter using the envelope emoji. Below is the message I’ve got from the Midjourney bot.

Then, I used the seed parameter from this message to create similar images, but with an eye-level angle. Here’s the result.

The angle is slightly different, but the overall composition is very similar.

You can experiment with the seed parameter to change the details of AI-generated images. Note that the output images would be exactly the same if you use the same prompt and the same seed parameter. If you introduce changes to the prompt, the new images will also change, but you’ll most likely be able to keep the overall composition and some color solutions.

To sum up

It’s possible to control the camera angle in AI-generated images by tweaking a text prompt. However, Midjourney will not always be responsive to the corresponding text prompts. To improve the chances of getting what you’re looking for:

1. Use photography language to specify the camera angle of a photo.

2. Consider using text weights to emphasize the angle of an image.

3. Add more words to support the request for a specific angle (e.g., “from below” for low-angle images).

4. Try many times! 😄 

Happy prompting!

🗞 News and Top Reads

• Microsoft recently released significant updates to Bing Chat, introducing new features such as image and video answers, plugins (similar to ChatGPT), chat history, and better integration with Edge.

  • The icing on the cake is that they have removed the waitlist, meaning all Microsoft account holders can access these features for free.

  • You can access Bing Chat through Edge or download the mobile app on iOS or Android.

• OpenAI unveiled a groundbreaking project called Shap-E, which uses text-to-3D technology to create complex and diverse 3D assets.

  • The current version is limited to single-object prompts with simple attributes, but with the rapid advancements in AI, it might not be long before text-to-3D printing becomes a reality.

📌 AI Art Tutorial: New Midjourney Features

In this tutorial, Matt Wolfe covers the latest updates to Midjourney, including the v5.1 version, RAW mode, permutations, the Niji model, and improved AI moderation. He is also talking about some of the things that will be released in the near future, according to the Midjourney developers.

🎨 Featured Artist: Fabio Comparelli

Fabio Comparelli is a digital artist based in Switzerland who is driven by a passion for creating beautiful and inspiring visuals. As a self-taught artist, he has always been fascinated by the endless possibilities that technology offers for creative expression.

Check out @fabdream.ai for remarkable evolution videos that made Fabio known worldwide.

If you want to recommend an artist to be featured in this newsletter, feel free to respond to this email with the links to the artist’s website or Instagram profile. Self-promotion is also allowed 😃 

🖼 AI-Assisted Artwork of the Week

🤓 How to Get Started with AI Art?

1. DALL-E: Creating Images from Text – introduction to text-to-image generation.

2. The DALL-E 2 Prompt Book – a guidebook by OpenAI that explains how to effectively right prompts to generate images across different domains (e.g., photography, illustration, art history, 3D artwork).

3. Best Midjourney Prompts – a guide that covers the basics of Midjourney prompts (e.g., which keywords to use to create abstract art, surreal art, minimalism, etc) as well as some more advanced options (e.g., keywords related to camera lenses and filters, imitating certain artists and photographers without using their names). Finally, they provide a list of 600+ creative text prompts for image generation.

4. Stable Diffusion Prompt Book – a prompt book prepared by OpenArt. The book discusses ideal prompt format, using modifiers to change the style, format, or perspective of the image, applying ”magic words” to improve image quality, adding negative prompts, and adjusting Stable Diffusion parameters.

Share Kiki and Mozart

If you like this newsletter and know somebody who might also like it, feel free to share this newsletter. Let’s have more people learn about AI art!

If you have been forwarded this email and you like it, please subscribe below. And welcome to the world of AI art!