AI art "非実在絶景:幻想的な「空中浮遊」系"

非実在絶景:幻想的な「空中浮遊」系

0

投稿のデイリークエスト用に。提案された特徴全部載せたかったけどさすがに以下のものは除外 CONTEST「わくわくアイス祭り」

national geographic photography, breathtakingly beautiful world scenery, emerald green lake, countless giant moss-covered rocks and islands floating in midair defying gravity, waterfalls gently flowing from rocks into the lake, golden twilight light enveloping the scene, perfect reflection of floating islands on water surface, ultra high quality, photorealistic, fantastical landscape photography, (fantasy "floating islands" theme:1.3), masterpiece, best quality, ultra detailed
ดูพารามิเตอร์
พรอมต์ต้นฉบับ
プロンプト: ナショナルジオグラフィックが撮影したような、息をのむほど美しい世界の絶景。エメラルドグリーンの湖の上に、無数の巨大な苔むした岩や島が、重力を無視して宙に浮かんでいる。岩からは滝が静かに湖へと流れ落ちている。夕暮れ時の黄金色の光が全体を包み込み、水面には浮遊する島々が完璧に反射している。超高画質、写実的、幻想的な風景写真。 # 幻想的な「空中浮遊」系 \ negaPrompt_copy ,OriginalPrompt_gemini ,PromptHelper_on # Model_SD ,Credit_Low ,Resolution_4x2_bottom ,SampMethod_EulerA ,Steps_def ,LoRA_rear ,Priority_low
ขนาด
1024 x 512
ขั้นตอนการสุ่ม
20
วิธีการสุ่มตัวอย่าง
Euler a
ค่า CFG
6
ค่า Seed
131
ไม่ต้องการ
human, people, girl, character, text, error, blurry, bad quality
ตัวช่วยพรอมต์

โมเดลและ LoRA ที่ใช้

รูปภาพคล้ายกัน

ดูเพิ่มเติม
AI artwork
bs
``` <artist tag>, <character tag>, <natural language description> ``` Since the user input consists entirely of asterisks (placeholders or noise) and no specific character, artist, or style is mentioned in the `user_input` string itself (the `recall` list is just history, not current input), and the rules state "If the user input doesn't mention characters, don't add characters", the prompt will focus on the default style (anime) and general atmosphere, lacking specific subjects. However, to make a valid image prompt, a generic subject or abstract composition is usually implied by asterisks in some contexts, but strictly following "no characters mentioned -> no characters", we describe a scene. Given the lack of specific content, I will create a generic, high-quality anime style landscape to satisfy the "detail beyond human eye" and "anime style" defaults. Actually, looking at the previous examples, the model tends to hallucinate if input is vague, but strictly, "1girl" is not in the input text '****...'. The input is just stars. Let's assume the stars represent a generic placeholder. Default style: Anime. No characters. Subject: Abstract or Landscape. Let's go with a detailed fantasy landscape to fulfill the "detail" requirement. ``` <artist tag>, <character tag>, <natural language description> ``` Wait, the formula requires `<artist tag>` and `<character tag>`. If empty, can they be empty? "If the user input mentions a character... place it at the beginning". If no artist/character, maybe just start with description? The prompt format says: `<artist tag>, <character tag>, <natural language description>` If none, maybe: `, , <description>` or just `<description>`? Let's assume empty tags if none provided. However, usually, text-to-image prompts need a subject. Let's interpret the asterisks as a generic "scenic view". ``` <artist tag>, <character tag>, <natural language description> ``` Let's refine the natural language description for a generic anime landscape with extreme detail. ``` anime style, a breathtakingly detailed fantasy landscape bathed in the warm, golden light of a setting sun, featuring a majestic, ancient castle perched atop a jagged cliff overlooking a vast, serene ocean. The sky is filled with vibrant hues of orange, pink, and purple, with intricate cloud formations that cast long, dramatic shadows across the scene. The castle's architecture is ornate, with towering spires, intricate stone carvings, and glowing windows that suggest life within. The cliffside is covered in lush, verdant vegetation, with delicate flowers and towering trees swaying gently in the breeze. The ocean below is a deep, rich blue, with waves crashing against the rocks below, creating a mist that adds depth and atmosphere to the image. The water reflects the sky's colors, creating a mesmerizing mirror effect. The overall atmosphere is one of tranquility and awe, with a sense of epic scale and beauty. The details are rendered with a level of precision that highlights the texture of the stone, the movement of the water, and the interplay of light and shadow. ``` Wait, the formula is `<artist tag>, <character tag>, <natural language description>`. If no artist/char, I will leave them empty or use a generic tag? Rules: "If the user input mentions a character... place it at the beginning". It doesn't say what to do if *not* mentioned. I will output just the description if tags are empty, or perhaps `default_anime_artist, , description`? Let's stick to the structure. If no specific artist is input, maybe no artist tag. Let's try: `, , anime style, a breathtakingly detailed...` Actually, looking at the previous turn, the user provided `artist: ineka_ka`. Here, `user_input` is just stars. `extra_info` has `artists: []`. So no artist tag. No character tag. ``` , , anime style, a breathtakingly detailed fantasy landscape bathed in the warm, golden light of a setting sun, featuring a majestic, ancient castle perched atop a jagged cliff overlooking a vast, serene ocean. The sky is filled with vibrant hues of orange, pink, and purple, with intricate cloud formations that cast long, dramatic shadows across the scene. The castle's architecture is ornate, with towering spires, intricate stone carvings, and glowing windows that suggest life within. The cliffside is covered in lush, verdant vegetation, with delicate flowers and towering trees swaying gently in the breeze. The ocean below is a deep, rich blue, with waves crashing against the rocks below, creating a mist that adds depth and atmosphere to the image. The water reflects the sky's colors, creating a mesmerizing mirror effect. The overall atmosphere is one of tranquility and awe, with a sense of epic scale and beauty. The details are rendered with a level of precision that highlights the texture of the stone, the movement of the water, and the interplay of light and shadow. ```
Benyamin Steinberg