Stable Diffusion is an AI-based image generation tool that allows you to create high-quality images simply by using descriptions (prompts). For beginners, understanding the structure of a prompt and how to create it is essential to achieve the desired results. Here’s the complete guide:
1. Understand the Structure of a Prompt
A good prompt consists of four main parts:
Opening: A description of the desired image quality.
Subject: Specifies who or what will be the main focus of the image.
Details: Mentions specific attributes or visual elements.
Closing: Adds aesthetic details such as lighting, expressions, or additional elements.
2. Separating Prompts with Commas (,)
Commas are used to separate each element in the prompt, making it easier for the AI to interpret the description.
Example:
masterpiece, high quality, photorealistic, solo woman, white t-shirt, black jacket
This prompt instructs the AI to generate an image of a woman with high quality, wearing a white t-shirt and a black jacket.
3. Emphasizing Elements with Double Parentheses (( ))
If you want to highlight certain elements in the prompt, use double parentheses (( )). This gives higher priority to the emphasized elements, ensuring the AI focuses more on them.
Example:
((masterpiece)), ((high quality)), ((highly detailed)), solo woman, white t-shirt, black jacket
This prompt ensures that the image quality is extremely high (masterpiece, high quality, highly detailed), while still considering the other elements.
4. Steps to Create a Prompt
a. Define the Image Quality (Opening)
Start by describing the quality of the image you want. Use keywords such as:
masterpiece (the best artwork)
highres (high resolution)
highly detailed (highly detailed)
photorealistic (photorealistic)
Example:
((masterpiece)), ((photorealistic))
b. Choose the Main Subject
Choose who or what will be the focus of the image, such as:
solo woman (a single woman)
1girl, 2girls (one or two women)
1boy, duo (a single man or two people)
Example:
solo woman
c. Add Specific Details
Include more specific visual attributes, such as clothing, poses, or background:
white t-shirt, black jacket (white t-shirt, black jacket)
street background (street background)
sitting (sitting)
Example:
white t-shirt, black jacket, sitting, street background
d. Add Aesthetic Details (Closing)
To enhance the result, add aesthetic descriptions such as:
realistic light (realistic lighting)
beautiful face (beautiful face)
soft lighting, blurred background (soft lighting, blurred background)
Example:
beautiful face, soft lighting, blurred background
5. Combine All Elements
Combine all the parts into one complete prompt.
Example of a Complete Prompt:
((masterpiece)), ((high quality)), ((highly detailed)), solo woman, white t-shirt, black jacket, sitting, street background, beautiful face, soft lighting, blurred background
6. Example Prompts I Created
Portrait of a Woman
((masterpiece)), ((high quality)), ((highly detailed)), ((photorealistic)), potret, up close, solo woman, using climbing clothes and mountain bags, following the curves of her body, the red and white Indonesian flag flies on the top of the mountain, brown hair, realistic face, realistic body, beautiful mountain background, realistic light
Fantasy Character
((masterpiece)), ((high quality)), ((highly detailed)), elf woman, long white hair, green eyes, wearing armor, standing in a forest, glowing light, magical atmosphere
Realistic Landscape
((photorealistic)), ((highres)), mountain landscape, clear blue sky, green forest, river flowing through rocks, sunlight shining, soft mist
Dynamic Action Scene
((high quality)), ((highly detailed)), superhero, flying through the city, red cape flowing, futuristic city background, dynamic pose, glowing lights, intense expression
Futuristic Sci-Fi Concept
((masterpiece)), futuristic city, neon lights, flying cars, cyberpunk style, solo man, black trench coat, standing under rain, glowing holographic screens
These prompts demonstrate how to combine elements for various creative results, from realistic portraits to fantastical concepts. Modify th
em to suit your vision!
Bahasa Indonesia
Panduan Lengkap Membuat Prompt di Stable Diffusion untuk Pemula
Stable Diffusion adalah alat generasi gambar berbasis AI yang memungkinkan Anda menciptakan gambar berkualitas tinggi hanya dengan menggunakan deskripsi (prompt). Untuk pemula, memahami struktur prompt dan cara membuatnya sangat penting untuk mendapatkan hasil yang sesuai harapan. Berikut panduan lengkapnya:
1. Memahami Struktur Prompt
Prompt yang baik terdiri dari empat bagian utama:
Pembuka: Berisi deskripsi kualitas gambar yang diinginkan.
Subjek: Menentukan siapa atau apa yang akan menjadi fokus utama gambar.
Detail: Menyebutkan atribut atau elemen visual yang spesifik.
Penutup: Menambahkan rincian estetika seperti pencahayaan, ekspresi, atau elemen tambahan lainnya.
2. Pemisahan Prompt dengan Koma (,)
Tanda koma digunakan untuk memisahkan setiap elemen dalam prompt agar AI lebih mudah memahami deskripsi.
Contoh:
masterpiece, high quality, photorealistic, solo woman, white t-shirt, black jacket
Prompt ini menginstruksikan AI untuk menghasilkan gambar seorang perempuan dengan kualitas tinggi, mengenakan kaos putih dan jaket hitam.
3. Penekanan Elemen dengan Tanda Kurung Ganda (( ))
Jika Anda ingin menonjolkan elemen tertentu dalam prompt, gunakan tanda kurung ganda (( )). Ini memberikan prioritas lebih tinggi sehingga AI lebih fokus pada elemen tersebut.
Contoh:
((masterpiece)), ((high quality)), ((highly detailed)), solo woman, white t-shirt, black jacket
Prompt ini memastikan kualitas gambar sangat tinggi (masterpiece, high quality, highly detailed), sementara elemen lainnya tetap diperhatikan.
4. Langkah-Langkah Membuat Prompt
a. Tentukan Kualitas Gambar (Pembuka)
Mulailah dengan mendeskripsikan kualitas gambar yang Anda inginkan. Gunakan kata kunci seperti:
masterpiece (karya seni terbaik)
highres (resolusi tinggi)
highly detailed (detail tinggi)
photorealistic (fotorealistik)
Contoh:
((masterpiece)), ((photorealistic))
b. Pilih Subjek Utama
Pilih siapa atau apa yang akan menjadi fokus gambar, seperti:
solo woman (perempuan tunggal)
1girl, 2girls (satu atau dua perempuan)
1boy, duo (satu laki-laki, atau dua orang)
Contoh:
solo woman
c. Tambahkan Detail
Tambahkan atribut visual yang lebih spesifik, seperti pakaian, pose, atau latar belakang:
white t-shirt, black jacket (kaos putih, jaket hitam)
street background (latar belakang jalan)
sitting (duduk)
Contoh:
white t-shirt, black jacket, sitting, street background
d. Tambahkan Rincian Estetika (Penutup)
Untuk memperindah hasil, tambahkan deskripsi estetika seperti:
realistic light (cahaya realistis)
beautiful face (wajah cantik)
soft lighting, blurred background (pencahayaan lembut, latar belakang buram)
Contoh:
beautiful face, soft lighting, blurred background
5. Gabungkan Semua Elemen
Gabungkan semua bagian menjadi satu prompt lengkap.
Contoh Prompt Lengkap:
((masterpiece)), ((high quality)), ((highly detailed)), solo woman, white t-shirt, black jacket, sitting, street background, beautiful face, soft lighting, blurred background
---
6. Contoh Prompt yang Saya Buat
Potret Seorang Wanita
((masterpiece)), ((high quality)), ((highly detailed)), ((photorealistic)), potret, up close, solo woman, using climbing clothes and mountain bags, following the curves of her body, the red and white Indonesian flag flies on the top of the mountain, brown hair, realistic face, realistic body, beautiful mountain background, realistic light
Karakter Fantasi
((masterpiece)), ((high quality)), ((highly detailed)), elf woman, long white hair, green eyes, wearing armor, standing in a forest, glowing light, magical atmosphere
Pemandangan Realistis
((photorealistic)), ((highres)), mountain landscape, clear blue sky, green forest, river flowing through rocks, sunlight shining, soft mist
Adegan Aksi Dinamis
((high quality)), ((highly detailed)), superhero, flying through the city, red cape flowing, futuristic city background, dynamic pose, glowing lights, intense expression
Konsep Sci-Fi Futuristik
((masterpiece)), futuristic city, neon lights, flying cars, cyberpunk style, solo man, black trench coat, standing under rain, glowing holographic screens
