Terracotta Courtyard Dhunuchi Naach – Bengali Durga Puja Portrait

♡ 15

Published October 24, 2026

Create a traditional Bengali Dhunuchi dance portrait with your real face, an ochre silk saree, terracotta architecture, glowing oil lamps and realistic incense smoke.

Using my uploaded photo as the reference, create a photorealistic 4:5 portrait of me performing Dhunuchi Naach in a traditional Bengali terracotta courtyard during Durga Puja. Preserve my exact recognizable facial identity, facial structure, facial proportions, eyes, eyebrows, nose, lips, jawline, cheeks, skin tone, hairline, hairstyle, distinctive facial features, natural age and overall likeness. Do not replace my face. Dress me in an ochre-yellow Bengali silk saree with an indigo woven border and subtle antique-gold details. Capture a graceful three-quarter dance turn while I securely hold exactly one terracotta dhunuchi with both hands. Include natural incense smoke, glowing embers, carved terracotta columns, blurred dhak drummers, a subtle Durga Puja shrine and a semicircle of small oil lamps. Use warm amber lamp light balanced by cool twilight ambience, realistic skin and silk texture, anatomically correct hands, a sharp face and shallow depth of field.

Works with

✦ChatGPT image generation✦Google Gemini image generation✦Midjourney✦Stable Diffusion XL

Why this prompt works

The prompt separates identity requirements from scene styling, specifies a single burner and controlled two-handed grip, and gives the model a clear pose, background and lighting plan. Terracotta textures and ochre silk create a coherent warm palette, while cool twilight ambience preserves depth. The semicircle of lamps provides visual structure without competing with the face.

What does this prompt combine?

Traditional Bengali Dhunuchi Naach, reference-based identity preservation, ochre silk saree styling, indigo textile accents, terracotta courtyard architecture, semicircular oil-lamp composition, realistic incense smoke and professional cultural portrait photography.

How to use this prompt

1. Upload a clear photo showing the subject’s face. 2. Paste the prompt into an image editor that supports reference photos. 3. Keep the uploaded photo active as the identity reference. 4. Generate a 4:5 portrait. 5. Verify that the face remains recognizable, the saree border is coherent, exactly one dhunuchi is present and both hands grip it naturally. 6. Regenerate or request a targeted correction if hands, smoke or clothing are inaccurate.

Recommended Settings

Aspect ratio: 4:5 portrait. Upload a clear reference photo with the face unobstructed. Enable reference-image or character-reference features where available. Prioritize identity preservation over stylization. Use high-quality photorealistic settings and moderate stylization. For SDXL image-to-image workflows, use a realistic photography checkpoint and moderate denoising, with a face-reference method if available. Inspect hand anatomy, burner count, smoke placement, saree border and facial likeness before accepting the result.

Compatible AI Image Generators

Ready to create? Try this prompt with the image tool you enjoy using. Each one works a little differently, so feel free to experiment.

◉ ChatGPT Image Generation ↗

Great for creating images or making changes to an existing one.

G Google Gemini ↗

Handy for detailed prompts and image-editing ideas.

✦ Midjourney ↗

A good choice for rich visual styles, compositions, and creative concepts.

☷ Stable Diffusion ↗

Useful when you want more control over the image-generation process.

How to Use This AI Prompt

Copy the prompt above and try it with your preferred image generator. Start with the original, then make a few small changes to get closer to the look you have in mind.

✓   Tips for Better Results

▣

Start with the original prompt: Give it a try as written before changing anything.

▧

Use your own photo: If your tool supports image input, add a photo to guide the result.

☷

Change one detail at a time: Adjust the lighting, background, colours, or style one step at a time.

↻

Try a few versions: Make a couple of variations and keep the one you like best.

Note: Results can vary depending on the AI tool, model, settings, and any photo you use.

Share Your Prompt & Results

Made something you love with an AI prompt? Share the prompt and the result with the SGPrompts community. We'd love to see what you create.

What to include
  • Creator name — the name you'd like us to credit.
  • Category — choose the topic that best fits your prompt.
  • Before image — the original or reference photo, if there is one.
  • After image — the final image you created.
  • Exact prompt — the full text you used to make it.
➤ Share Your Prompt & Results

Follow SGPrompts on Instagram

Come see more AI creations, prompt ideas, and experiments from the SGPrompts community.

What you'll find

  • Fresh creations: A closer look at new images and visual ideas.
  • Prompt inspiration: Ideas you can try and make your own.
  • Behind the results: See how prompts turn into finished images.
  • New posts: Keep up with the latest from SGPrompts.
◎   Follow on Instagram

Join SGPrompts on Facebook

Like the page to catch new prompt ideas, share your creations, and see what others in the community are making.

Let's create together

  • New prompt ideas: Find inspiration for your next image.
  • Community creations: Discover what other creators are making.
  • Tips and experiments: Pick up ideas to try in your own workflow.
  • Page updates: See what's new from SGPrompts.
f   Visit our Facebook Page

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *

Related prompts