The “Image Prompt URL” is easy to explain. There are four input fields where you can enter up to four URLs. You can either insert any image URL to let Midjourney know what you have in mind, or you can upload a photo of yourself or a friend to create a new image.

In my experience, if you use a close-up image, such as a face shot, and input a full body description in the text, Midjourney will still create a close-up of the face. The URL comes before the text input. However, it’s important to note that Midjourney’s AI models are not perfect and may not always provide the desired results. It’s also possible that Midjourney may emphasize certain features depending on what’s visible in the image URL. So, it may be helpful to carefully select the images to use as inspiration and try different options to achieve the desired results.

I would recommend uploading the images directly to Midjourney. While there is an option to upload your own images, I’m from Germany and dealing with data privacy here can be a hassle. If I were to do that, you would have to dismiss a popup every time you open this page, and I don’t want to put you through that.

Describe your Picture” is also self-explanatory. Here, you should type in your idea of the image. For example: “A young girl dances barefoot on a pond over the water and holds balloons in her hands.”

If you ever can’t think of anything, there is now a new “Random” button above the input field 🙂 This generates sentences made up of subjects, verbs, adjectives, objects, endings, and artists. Sometimes, the sentences don’t make sense at all, but the results are still amazing! Feel free to try it out 🙂

Photorealism for V5″ is divided into Portrait Styles, Color Hints, Lighting Conditions, and Misc Styles. These are further divided, and you can select what you want by checking the appropriate boxes. It’s worth taking a look at everything once, and then you’ll work faster the next time. There is an infinite amount of choices available!

More Extras” is also divided. Here, you can select checkboxes for Fashion, Resolution, Light Leaks, etc. There are endless options available!

Backgrounds ist unterteilt in Lighting Conditions, Colors, Styles, Composition und Special Effects

If you don’t want to select a background from the options above and know exactly what kind of background you want, you can enter what kind of background you are looking for here.

Here, you can select the camera quality of some cameras, meaning that you directly choose the camera. Only very expensive, some almost unaffordable, and special cameras are selected, otherwise the list would be too long.

You can select the appropriate lens for the camera here, and only the top brands are included. Also note that, regardless of the camera you choose, the lens does not have to match the camera in real life. The system doesn’t care about that. What doesn’t work in reality works virtually. So, you can connect Sony cameras with Canon lenses, no problem!

Here, we have Midjourney’s own parameters!

Aspect Ratio: The pixel value doesn’t really affect the size of the image, but is only used to help you visualize how the aspect ratio should look. From top to bottom, the horizontal size decreases until it becomes a square, then it becomes larger again, but horizontally.

Style: I can’t say much about this, please check the official website.

Quality: The quality of the generated image. If the quality doesn’t matter to you, choose q .25, but if it does, go up to –q 2.

Image Weight: I’m not sure about -iw myself as I’m still in the testing phase.

Version: Newly added is version 5.1, so use –v 5.1. I highly recommend this version as the results are amazing!

Niji: Here, you can choose between Niji and Niji 5. I’m not sure what the difference is between the two, but when you choose Niji, Midjourney focuses on creating an anime/drawing-style image. The results are fantastic!

Uplight: No idea yet.

Upbeta: No idea yet.