How to properly write AI Art text prompts

Vladimir Chopine [GeekatPlay]
19 Oct 202227:45

TLDRThe video script discusses the do's and don'ts of crafting AI art prompts, specifically for the MidJourney AI model, emphasizing the importance of simplicity and understanding how prompt weights work. It advises against blindly copying and pasting prompts from the internet, as this can lead to unpredictable and often confusing results. Instead, the speaker recommends a structured approach: clearly defining the main subject, describing the environment, specifying colors, and adding technical details or artist references for more control over the AI's output. The video also highlights the power of mixing artist styles and the potential pitfalls of over-complicating prompts.

Takeaways

  • ๐ŸŽจ When using AI art prompts, aim for simplicity and avoid blindly copying and pasting prompts from the internet.
  • ๐ŸŒŸ Focus on the main subject and use descriptive words to emphasize important aspects, like 'very long hair' for an elf princess.
  • ๐Ÿ“ธ Understand how the AI model's weight system works, as it references image libraries based on the terms used in the prompt.
  • ๐Ÿ–Œ๏ธ Be specific with color grading; generic terms like 'color gradient' are less effective compared to 'pastel color grading'.
  • ๐Ÿ“ Pay attention to technical terms like 'depth of field' and 'shutter speed', as incorrect or conflicting usage can lead to unpredictable results.
  • ๐Ÿ‘ฉโ€๐ŸŽจ Reference specific artists or styles for a higher weight and more predictable outcomes in the AI-generated art.
  • ๐ŸŽˆ Experiment with different styles and artists, but avoid overwhelming the system with too many references, which can result in a 'muddy' output.
  • ๐Ÿ”„ Use the 'remix' feature to make variations of the AI-generated art, allowing for small adjustments to refine the result.
  • ๐ŸŒฒ When describing the environment, be clear and concise to avoid confusion and ensure the AI renders the scene as intended.
  • ๐Ÿ–ผ๏ธ For a more controlled and predictable outcome, provide clear and structured information, separating subject, environment, colors, and technical details.

Q & A

  • What is the main focus of the video?

    -The main focus of the video is to discuss the do's and don'ts of AI art prompts, specifically using the MidJourney model to convert text to image, and how to achieve predictable and desired results.

  • Why is it important to avoid blindly copying and pasting prompts from the internet when using AI art generators?

    -Blindly copying and pasting prompts can lead to confusing and inconsistent results, as the AI may not understand the combination of elements and weights specified in the prompt. It's better to use simplicity and understand what each element contributes to the final image.

  • What does the video suggest about the use of specific artist references in AI art prompts?

    -Referencing specific artists in AI art prompts can have a significant impact on the resulting image, as the AI will attempt to replicate the style of the mentioned artist, leading to more predictable and desired outcomes.

  • How can the weight of certain elements in an AI art prompt affect the final image?

    -The weight of elements in an AI art prompt determines the importance of those elements in the final image. Higher weight means the AI will prioritize including that element, ensuring it appears in the generated art.

  • What is the significance of using a simple and structured approach when creating AI art prompts?

    -A simple and structured approach helps to avoid overwhelming the AI with too much information, which can lead to random and uncontrolled results. It allows for better predictability and control over the final image.

  • Why is it recommended to specify colors in AI art prompts?

    -Specifying colors in AI art prompts helps the AI to understand the desired color palette, leading to more accurate and visually coherent results. General terms like 'color gradient' may not provide clear direction to the AI.

  • What does the video suggest about experimenting with AI art prompts?

    -The video encourages experimentation with AI art prompts, such as mixing different artist styles, but cautions against overcomplicating the process, as it can lead to muddled and unpredictable results.

  • How can the 'remix' feature in MidJourney be useful for refining AI-generated art?

    -The 'remix' feature allows users to make variations of an existing AI-generated image by adjusting small details, providing a way to fine-tune the image to better match the desired outcome.

  • What is the role of technical information in AI art prompts?

    -Technical information, such as specifying the desired resolution or lens type, helps the AI to generate images with the correct technical aspects, ensuring the final image meets the user's requirements.

  • What advice does the video give for achieving the best results with AI art generators?

    -The video advises to start with simplicity, understand the impact of each element in the prompt, reference specific artists for style consistency, and use structured and clear language to guide the AI in creating the desired image.

Outlines

00:00

๐ŸŽจ AI Art Prompts: Dos and Don'ts

The paragraph discusses the importance of crafting AI art prompts effectively. It emphasizes the need to avoid blindly copying and pasting prompts from the internet, as this can lead to unpredictable results. The speaker provides an example of a prompt taken from a community feed and explains the issues with it, such as conflicting terms and the misuse of technical terms like 'octane render' and 'Unreal Engine'. The focus is on understanding how to structure prompts to achieve expected outcomes, highlighting the significance of simplicity and clarity in the creative process.

05:01

๐Ÿ“ธ Understanding AI Art Prompts and Photography

This section delves into the specifics of constructing AI art prompts with a focus on photography. It critiques the misuse of photography-related terms in prompts, such as 'depth of field' and 'ultra-wide angle', which can lead to inconsistent results. The speaker explains the importance of understanding the weight and relevance of terms used in prompts, and how they correspond to the AI's reference libraries. The paragraph also touches on the concept of 'gamma' and 'white balance', emphasizing the need for specificity to guide the AI effectively.

10:02

๐Ÿ–Œ๏ธ Structuring Effective AI Art Prompts

The speaker shares personal recommendations for structuring AI art prompts. They suggest starting with the main subject, then describing the environment, colors, and technical aspects. The importance of referencing specific artists is highlighted, as it carries significant weight in influencing the AI's output. The paragraph also discusses the power of combining styles and the potential pitfalls of overloading the prompt with too many references, which can result in a 'muddy' outcome.

15:03

๐ŸŒŸ Experimenting with Styles and References

This paragraph explores the concept of experimenting with different styles and artist references in AI art prompts. The speaker demonstrates how mixing styles can lead to unique and creative outcomes, but warns against overcomplicating prompts. They also introduce the 'remix' feature, which allows for variations on a prompt to refine the AI's output. The importance of providing clear and concise technical information is reiterated to achieve predictable and desired results.

20:04

๐Ÿž๏ธ Community Feed: Learning from Others

The speaker discusses the value of the community feed in learning from others' AI art creations. They point out common mistakes in prompts, such as the misuse of 'HD' and '8K', and the overuse of references leading to generic outputs. The paragraph emphasizes the effectiveness of simple and well-structured prompts that reference specific artists or styles. The speaker also encourages viewers to engage with the community, share their prompts, and learn from the successes and failures of others.

25:05

๐ŸŽˆ Simplicity in AI Art Prompts

The final paragraph stresses the importance of simplicity in crafting AI art prompts. The speaker shares personal preferences for structuring prompts and the impact of overcomplicating them. They demonstrate how specifying colors in a prompt can lead to more controlled outcomes, as opposed to using vague terms like 'color palette'. The video concludes with a call to action for viewers to experiment with prompts, learn from their experiences, and share their findings with the community.

Mindmap

Keywords

๐Ÿ’กAI art prompts

AI art prompts refer to the text-based instructions or descriptions provided to AI algorithms to generate specific types of digital art. In the context of the video, these prompts are used to guide AI in creating images that match the user's vision, with the emphasis on understanding how to structure these prompts effectively for predictable results.

๐Ÿ’กMid Journey

Mid Journey is likely a reference to a specific AI art generation platform or tool that the speaker is using to demonstrate the process of creating AI art. It represents one of the tools available for converting text prompts into images, and the video provides insights on how to use it effectively.

๐Ÿ’กStable diffusion

Stable diffusion is a term that might refer to a specific AI model or technique used in the process of generating stable and consistent outputs from AI art rendering. It implies a method that minimizes variability in the results, ensuring that the AI's output matches the user's expectations more closely.

๐Ÿ’กCree model

The Cree model is not explicitly defined in the video script, but it likely refers to an AI model used for converting text to images. The model is part of the technology that allows users to create art by inputting descriptive text, and the video emphasizes the importance of understanding how this model interprets prompts to achieve desired results.

๐Ÿ’กWeight

In the context of AI art generation, 'weight' refers to the importance or emphasis assigned to certain elements within the text prompt. By adjusting the weight, users can influence the AI's focus on particular aspects of the image, ensuring that the final output aligns more closely with the user's vision.

๐Ÿ’กOctane render

Octane render is a term that refers to a specific rendering engine used in 3D graphics and visual effects. It is mentioned in the video as an example of a term that, when included in an AI art prompt, might lead the AI to reference images tagged with 'Octane render' from its libraries, potentially influencing the style and quality of the generated art.

๐Ÿ’กPhotorealistic

Photorealistic refers to the quality of an image or artwork that closely resembles real-life photographs in terms of detail and visual accuracy. In the context of AI art generation, achieving photorealistic results means creating images that look like they could have been captured by a camera, with a high level of detail and realism.

๐Ÿ’กDepth of field

Depth of field (DoF) is a photography term that describes the range of distance within a scene that appears acceptably sharp and in focus. In AI art generation, specifying depth of field can help create a more realistic and three-dimensional effect by mimicking the way a camera lens captures focus and blur.

๐Ÿ’กArtist reference

Referring to specific artists in AI art prompts can heavily influence the style and aesthetic of the generated image. By mentioning an artist's name, the AI is directed to search its libraries for images tagged with that artist's work, which can then be used as a reference for the style and mood of the AI-generated art.

๐Ÿ’กVariations

In the context of AI art generation, variations refer to the process of making slight modifications to an existing prompt or image to create new, related outputs. This allows users to explore different interpretations of their original idea, experimenting with small changes to see how they affect the final image.

๐Ÿ’กCommunity Feed

The Community Feed is a feature within AI art platforms where users can share their generated images and view the creations of others. It serves as a space for inspiration, learning, and feedback, allowing users to see how different prompts and styles are interpreted by the AI.

Highlights

The video discusses do's and don'ts with AI art prompts, specifically using the MidJourney model.

The importance of understanding how prompt weights work in AI art generation is emphasized.

Avoid blindly copying and pasting prompts from the internet to achieve desired results.

The video provides an example of a prompt taken from the community feed and analyzes its effectiveness.

The concept of 'predictability' in AI art generation is introduced, aiming for consistent results.

The video explains the significance of specifying colors in AI art prompts, such as 'pastel color grading'.

The importance of understanding photography terms when using them in prompts, like 'depth of field' and 'ultra-wide angle'.

The video demonstrates how mixing artist styles in prompts can lead to unique and powerful AI art outcomes.

The use of specific artist references in prompts can heavily influence the AI's output.

The video showcases the 'remix' feature in MidJourney, allowing for variations and adjustments to AI art.

The video emphasizes the value of simplicity in prompts, advising against overly complex or nonsensical combinations.

The video provides a personal recommendation for structuring prompts effectively, including subject, environment, and color description.

The video discusses the impact of technical information in prompts, such as specifying 'diorama isometric'.

The video highlights the importance of artist references in achieving desired styles, such as using HR Giger's style.

The video advises on the potential pitfalls of mixing too many artist styles, which can lead to 'muddy' or inconsistent results.

The video encourages experimentation with AI art prompts, but with a structured approach to avoid randomness.

The video concludes with a call to action for viewers to share their own effective prompts and experiences with AI art generation.