avatarSweet Chaos

Summary

The web content provides a detailed guide on creating appetizing AI-generated food images using Midjourney, emphasizing the importance of precise prompts and iterative refinement to achieve the desired visual effects.

Abstract

The article, titled "AI Image How-To— #10," delves into the process of crafting enticing AI-generated images of food, such as cinnamon rolls, steak dinners, and ice cream sundaes, using the Midjourney platform. The author shares personal experiences and tips on how to fine-tune prompts to produce images that closely match the envisioned concepts, highlighting the significance of specific descriptors like "huge fluffy cinnamon roll" and the use of "+" to group food items. The article also touches on common challenges faced during the image generation process, such as AI adding unwanted elements like people or misinterpreting food items, and how to overcome them through persistent adjustments and inspiration from renowned photographers like Mario Testino and Joseph O'Toole. The author concludes by encouraging readers to experiment with the AI service ZAI.chat, positioned as a cost-effective alternative to ChatGPT Plus with similar capabilities.

Opinions

  • The author expresses a personal weakness for cinnamon rolls, indicating a strong emotional connection to the subject matter.
  • There is a clear appreciation for the AI's ability to create detailed and realistic images, particularly after refining prompts.
  • The author finds humor in the AI's unexpected additions, such as people appearing in images despite the "no people" tag.
  • Frustration is conveyed when the AI misinterprets prompts, such as creating a "weird, under-cooked bird" instead of a steak or a "fancy pizza pie" instead of lasagna.
  • The author values the work of famous photographers, using their styles to enhance the AI-generated images.
  • A sense of accomplishment is evident as the author successfully generates images that meet their expectations after multiple attempts.
  • The author endorses the use of Midjourney's V4 for image generation, suggesting a preference for this version.
  • There is an invitation for reader engagement, asking for feedback and suggestions for future topics.
  • The author promotes their other works and publications, indicating a desire to share their knowledge and experiences with a broader audience.
  • The author recommends ZAI.chat as a budget-friendly AI service, implying satisfaction with its performance and value for money.

AI-GENERATED IMAGES|ART|FICTION|NON-FICTION

AI Image How-To— #10

Delectable foods to tantalize your readers — secrets revealed!

Cinnamon roll — AI image using Midjourney

I originally created this image for my Jenna’s Big Day series I am re-writing…before changing all the images to publish the series on Tapas with the standard anime style (and Medium to match because it’s too much work creating separate images for both.)

In case you didn’t know, cinnamon rolls are my weakness. Fluffy, cinnamon-y goodness with just a little bit of glaze drizzled on. Mmmm.

Now that you are likely craving this delectable confection, I’ll share the creation journey.

I began my prompt with cinnamon rolls. AI obliged and created four (smaller) rolls on the plate.

I also began the prompt with a fork. AI creates forks as beautifully as it creates fingers. Enough said.

So, I took the fork out and changed cinnamon rolls to huge fluffy cinnamon roll and began getting the image I had in my mind. (This has been why I spend too much time creating AI images — trying to get what’s in my head to appear on the screen. LOL)

With those prompt alterations, I quickly received amazing results that would leave anyone salivating. I added glamour shot to the prompt to get different poses.

Image prompt for cinnamon roll:

Huge fluffy cinnamon roll, plate, hyperdetailed, cinematic still shot, glamour shot, 32k, by Mario Testino, by Albert Bierstad

Steak dinner — AI image using Midjourney

When I first was creating this image, AI was giving me very stuffy looking people (who appeared as though they were repulsed by the food set before them) despite listing no people in the prompt. I also began with steak dinner in the prompt.

After changing the wording to steak + loaded baked potato + salad, I received tempting images each time. I grouped each food item with + rather separated by commas.

I did not get the loaded baked potato, obviously. As I continued trying for it, AI thought it’d be funny to start throwing people into the images. At least they looked joyous to have a steak dinner.

Also, I searched food photographers and looked at several pictures before deciding to use Joseph O’Toole.

Image prompt for steak dinner:

steak + loaded baked potato + salad, no people, hyperdetailed, cinematic still shot, glamour shot, 32k, by Mario Testino, by Joseph O’Toole

Non-steak dinner — AI image using Midjourney

And because I’m sure you are curious how someone could look stuffy around a steak dinner, here you go.

That looks more like a weird, under-cooked bird — that is not a steak!

Image prompt for steak dinner:

steak dinner, no people, hyperdetailed, cinematic still shot, glamour shot, 32k, by Mario Testino, by Joseph O’Toole

I used the same prompt set-up for the lasagna meal as I did for the steak dinner. Surprisingly, I had so much trouble!

I had to generate a lot of images to get this one. The lasagna looks more like a fancy pizza pie (to me) than lasagna and AI didn’t seem to know what a bread stick is…many looked like a hard, rolled up cookie (Pirouette cookies.)

Most images had a person (regardless of the no people tag in the prompt) or a photograph of one next to the lasagna. Who eats with a photo?

After many determined tries, I began getting images without people and worked at generating those images to get the meal above.

Image prompt for lasagna dinner:

lasagna + breadsticks + salad, no people, hyperdetailed, cinematic still shot, glamour shot, 32k, by Mario Testino, by Joseph O'Toole

Ice cream sundae extraordinaire — AI image using Midjourney

For the finale of this feast is the iconic ice cream sundae!

This image was the easiest to create with only ice cream sundae and no people, giving AI freedom to run wild with our desires.

Some of the sundaes had ice cream dripping over the edges and all looked luscious!

Note: I’ve noticed, the more you work to generate an image, the more colorful and detailed they get. (Also, I have been using V4 for all of these images. To access V4 in Midjourney, type /settings then click enter to load the settings and choose V4.)

Image prompt for ice cream sundae:

ice cream sundae, no people, hyperdetailed, cinematic still shot, glamour shot, 32k, by Mario Testino, by Joseph O'Toole

Messy kitchen — AI image using Midjouney

My work here is finished. You all can clean up the kitchen. 🤣

Image prompt for messy kitchen:

messy kitchen and table, food spills, no people, hyperdetailed, cinematic still shot, glamour shot, 32k, by Mario Testino, by Joseph O'Toole

Let me know what you think and what you’d like to learn!

December 12, 2022 — Photographers and magazines to enhance your images!

AI Images — #8|Table of Contents|AI Images — #10

For more writing and AI tips, check out my link below!

Find me at Gone Rogue and The Infinity Library Collection on Tapas or The Infinity Library on Substack!

Sweet Chaos — Tickling your senses through fiction writing while shocking you with the outcome! Nothing is as it seems! Ready to read every addicting story leaving you on the edge of your seat?

Joining through my link gives me a smidge extra at no cost to you, and a treat for my dog!

Check out my publication — The Fictioneer’s Treasure Trove to raid and fill the library!

Artificial Intelligence
Tft Mj
Midjourney
Generative Art
Midjourney How To
Recommended from ReadMedium