Уроки Midjourney с нуля / #6 – Продвинутые промты (Advanced prompts)

This paragraph discusses advanced techniques for creating prompts to generate images using a tool like 'PROMT'. It explains how to add images to prompts by finding and copying image links from Google, and emphasizes the importance of syntax where the image link should be the first element in the prompt to avoid errors. The paragraph also introduces the concept of using text guides or parameters alongside the image link to influence the composition, style, and colors of the generated image. A small guide is provided to demonstrate how to use an image of 'SpongeBob SquarePants' with a specific aspect ratio to generate an image. Furthermore, it talks about adding multiple image links and recommends limiting it to 2-3 images for predictable results. It introduces the 'Blend' mode for mixing images to create a single high-quality result, explaining the process of uploading up to 5 images and adjusting the output format using the 'dimension' parameter to avoid issues with different image formats.

The second paragraph delves into the nuances of fine-tuning image prompts to achieve desired results. It covers how to split prompts into key parts using double colons to assign relative importance and avoid misinterpretation by the image generation tool. An example is given where the term 'hot-dog' is corrected to 'hot dog' by splitting it, resulting in a dog with warm elements instead of a hot dog or a flying insect. The paragraph also explains how to emphasize a word in a prompt by adding a number after a double colon, demonstrating how this can increase the presence of fire elements in an image. It discusses the possibility of using both whole and fractional numbers to adjust the weight of words in the prompt. Additionally, it touches on the use of negative parameters to exclude certain elements, such as removing red flowers from a field, and clarifies that the sum of weights must be positive. The paragraph concludes with a reflection on the complexity of the syntax and invites viewers to share if they find practical applications for these advanced techniques.

In the final paragraph, the speaker wraps up the discussion on advanced prompting techniques and hints at the upcoming content in the next video, which will be the last in the series. The speaker mentions that they will explore where to find fronts and how 'Midjourney' can work with different styles, promising to generate several beautiful images in the process. The paragraph serves as a conclusion to the current topic and a teaser for what's to come in the next video, creating anticipation for the continuation of the course.