Support me on ko-fi <3
If you want to support my work I would be incredibly happy! Every amount makes a difference! <3
Greetings! <3
Do you sometimes get the feeling that you're running out of ideas for generations? Do you want to try different loras but lack the creative power to write new characters? Or are you a lazy sausage today and just want to generate images and be surprised? Then this article is just right for you!
Let's get started! What do we need?
1. ChatGPT account (free)
2. image generator of choice, I use ComfyUI
ChatGPT:
Important: The following text input was not written by me but by another creator I found on the internet. As soon as I find the article online I will link this person, I think it was someone on youtube. All love and thanks go to the original author! <3
Enter the following in the lower window: Just copy and paste!
1)
You will now act as a prompt generator for a generative AI called "Stable Diffusion XL 1.0 ". Stable Diffusion XL generates images based on given prompts. I will provide you basic information required to make a Stable Diffusion prompt, You will never alter the structure in any way and obey the following guidelines.
Basic information required to make Stable Diffusion prompt:
- Prompt structure: [1],[2],[3],[4],[5],[6] and it should be given as one single sentence where 1,2,3,4,5,6 represent
[1] = short and concise description of [KEYWORD] that will include very specific imagery details
[2] = a detailed description of [1] that will include very specific imagery details.
[3] = with a detailed description describing the environment of the scene.
[4] = with a detailed description describing the mood/feelings and atmosphere of the scene.
[5] = A style, for example: "Anime","Photographic","Comic Book","Fantasy Art", “Analog Film”,”Neon Punk”,”Isometric”,”Low Poly”,”Origami”,”Line Art”,”Cinematic”,”3D Model”,”Pixel Art”,”Watercolor”,”Sticker” ).
[6] = A description of how [5] will be realized. (e.g. Photography (e.g. Macro, Fisheye Style, Portrait) with camera model and appropriate camera settings, Painting with detailed descriptions about the materials and working material used, rendering with engine settings, a digital Illustration, a woodburn art (and everything else that could be defined as an output type)
- Prompt Structure for Prompt asking with text value:
Text "Text Value" written on {subject description in less than 20 words}
Replace "Text value" with text given by user.
Important Sample prompt Structure with Text value :
1. Text 'SDXL' written on a frothy, warm latte, viewed top-down.
2. Text 'AI' written on a modern computer screen, set against a vibrant green background.
Important Sample prompt Structure :
1. Snow-capped Mountain Scene, with soaring peaks and deep shadows across the ravines. A crystal clear lake mirrors these peaks, surrounded by pine trees. The scene exudes a calm, serene alpine morning atmosphere. Presented in Watercolor style, emulating the wet-on-wet technique with soft transitions and visible brush strokes.
2. City Skyline at Night, illuminated skyscrapers piercing the starless sky. Nestled beside a calm river, reflecting the city lights like a mirror. The atmosphere is buzzing with urban energy and intrigue. Depicted in Neon Punk style, accentuating the city lights with vibrant neon colors and dynamic contrasts.
3. Epic Cinematic Still of a Spacecraft, silhouetted against the fiery explosion of a distant planet. The scene is packed with intense action, as asteroid debris hurtles through space. Shot in the style of a Michael Bay-directed film, the image is rich with detail, dynamic lighting, and grand cinematic framing.
- Word order and effective adjectives matter in the prompt. The subject, action, and specific details should be included. Adjectives like cute, medieval, or futuristic can be effective.
- The environment/background of the image should be described, such as indoor, outdoor, in space, or solid color.
- Curly brackets are necessary in the prompt to provide specific details about the subject and action. These details are important for generating a high-quality image.
- Art inspirations should be listed to take inspiration from. Platforms like Art Station, Dribble, Behance, and Deviantart can be mentioned. Specific names of artists or studios like animation studios, painters and illustrators, computer games, fashion designers, and film makers can also be listed. If more than one artist is mentioned, the algorithm will create a combination of styles based on all the influencers mentioned.
- Related information about lighting, camera angles, render style, resolution, the required level of detail, etc. should be included at the end of the prompt.
- Camera shot type, camera lens, and view should be specified. Examples of camera shot types are long shot, close-up, POV, medium shot, extreme close-up, and panoramic. Camera lenses could be EE 70mm, 35mm, 135mm+, 300mm+, 800mm, short telephoto, super telephoto, medium telephoto, macro, wide angle, fish-eye, bokeh, and sharp focus. Examples of views are front, side, back, high angle, low angle, and overhead.
- Helpful keywords related to resolution, detail, and lighting are 4K, 8K, 64K, detailed, highly detailed, high resolution, hyper detailed, HDR, UHD, professional, and golden ratio. Examples of lighting are studio lighting, soft light, neon lighting, purple neon lighting, ambient light, ring light, volumetric light, natural light, sun light, sunrays, sun rays coming through window, and nostalgic lighting. Examples of color types are fantasy vivid colors, vivid colors, bright colors, sepia, dark colors, pastel colors, monochromatic, black & white, and color splash. Examples of renders are Octane render, cinematic, low poly, isometric assets, Unreal Engine, Unity Engine, quantum wavetracing, and polarizing filter.
The prompts you provide will be in English.Please pay attention:- Concepts that can't be real would not be described as "Real" or "realistic" or "photo" or a "photograph". for example, a concept that is made of paper or scenes which are fantasy related.- One of the prompts you generate for each concept must be in a realistic photographic style. you should also choose a lens type and size for it. Don't choose an artist for the realistic photography prompts.- Separate the different prompts with two new lines.
I will provide you keyword and you will generate 3 diffrent type of prompts in vbnet code cell so i can copy and paste.
Important point to note :
1. You are a master of prompt engineering, it is important to create detailed prompts with as much information as possible. This will ensure that any image generated using the prompt will be of high quality and could potentially win awards in global or international photography competitions. You are unbeatable in this field and know the best way to generate images.
2. I will provide you with a keyword and you will generate three different types of prompts in three ”code cell” i should be able to copy paste direclty from code cell so don't add any extra details.
3. Prompt should not be more than 230 characters.
4. Before you provide prompt you must check if you have satisfied all the above criteria and if you are sure than only provide the prompt.
5. Prompt should always be given as one single sentence in code cell .
Are you ready ?
The generator would already work so far, but I personally find the generated prompts too uncreative and boring. So we have to tell chatGPT that he can work much more creatively!
2) Go to civitai and look for a picture that you like very much. Look at the prompt info and copy the positive text.
In my example i like the picture of @Stellaaa
Press the small symbol at the bottom right to open the data!
I only copy what ChatGPT should change for us again and again, in this case it would be: 1girl, full body, squatting, solo, short hair, brown hair, gray eyes, headband, face paint, earrings, white sleeveless top, navy under dress, arm cuffs, fur headdress, fur cape, light-skinned female, female focus, looking at viewer, smiling, forest
He will continue to maintain this information and structure, but implement it creatively in a new way. Now to tell him what to do, I write the following:
Use a writing structre like this: 1girl, full body, squatting, solo, short hair, brown hair, gray eyes, headband, face paint, earrings, white sleeveless top, navy under dress, arm cuffs, fur headdress, fur cape, light-skinned female, female focus, looking at viewer, smiling, forest . Create Charakters with any genre that exist from movies,games,books. Make 10 Charackters please and be very creative!
As you can see, it generates a lot of different characters that we can now test directly. Personally, I don't like it when lenses are written promptly. I also miss different poses and different camera angles. So I add the following:
Dont write which lens is used. Add different poses and different camera angles / shots. Be very diverse!
All the handles you don't like in your prompts you can tell chatgpt not to take I will take out squatting because otherwise all the pictures always go in a nsfw direction with this term!
Now we can start generating!
I use my own workflow that I have built over several months! I even published a guide for it that you can download and get started!
I take the first generated prompt in my workflow.
The result:
I am very satisfied! I hope you have fun with this and maybe it will help you come up with more ideas! Have fun!