summary
A Reddit user shared some convincingly bad food photos, all of which were generated by AI. The Reddit user shared the prompts they used to generate these photos in Midjourney, and you can easily recreate the same effect to create impressive iPhone-like AI food photos.
AI image generators are one of the most prominent ways we can leverage AI in our lives. Enter a text prompt and an image will be generated to meet your needs. However, generating a good AI image from an AI image generator requires a fair amount of skill, and it can take quite a few tries to create a good prompt that works for you. If you want to create an amazingly deceptive AI image, Midjourney has prompts that can get you some very convincing iPhone-like results.
Reddit user MagicJourknees posted the series of images to Reddit with the caption, “My girlfriend puts raisins in everything! Is this normal?!”
These photos left many astonished and confused at this person’s ability to enjoy any meal with raisins in it, but many also noticed that these photos were posted to the Midjourney subreddit, which means that none of these photos are real.
Yes, these images were all AI-generated using Midjourney, but the photography style and output make them look incredibly realistic – if you showed these photos to someone without context, they’d be hard-pressed to tell they were AI-generated.
Thankfully, Reddit user MagicJourknees was kind enough to share the prompts he used to create these images. Below are the prompts used for Midjourney:
code
Copy text iPhone photo of (food name) with lots of raisins on top, taken at (kind of) restaurant (or other place). –ar 3:4 –style raw –s 75 PLUS –sref (Image URL of bad food photo found on Yelp or elsewhere)
The first part of the prompt is self-explanatory and requires you to enter variables in () brackets. The -ar parameter defines the aspect ratio Midjourney will use (3:4 in this prompt), while the -style raw parameter and argument will generate a less “beautified” and more “correct” image. The -s parameter is a styling parameter that can take a value between 0 and 1000, with a value of 75, defaulting to 100, so the generated image will have less artistic freedom and will be closer to the prompt. Finally, the -sref parameter allows you to use an existing image as a visual style reference, while entering a direct image URL will provide a solid reference point for the generated image.
Give the prompts above a try to create some seriously convincing bad food photos on Midjourney and let us know about your experience in the comments below.
Have a tip? Talk to us! Email our staff at news@androidauthority.com. It’s up to you to remain anonymous or get credited for your information. Comments
Source link
