34 - Pitagoras

Shubham, aka Pitagoras / Prickle Graphics, is a 20-year-old designer from Delhi, India. He developed a passion for digital art at an early age. He originally started out as a YouTuber, creating VFX shots on his mobile phone, which gradually introduced him to the world of design and visual storytelling.

Over time, thumbnail design became the perfect way for him to combine his love for YouTube with his passion for creating art in Photoshop. With a strong interest in photo manipulation, he spent years experimenting and refining his skills.

In 2025, after buying his first laptop, he began pursuing thumbnail design professionally and turned his long-time passion into a career. Pitagoras is available for hire.

Twitter (X) → x.com/pricklegraphics
Portfolio → pricklegraphics.framer.ai

The theme for this thumbnail was a treehouse. I enjoy experimenting with angles and new approaches to challenge myself.

After reading the WC#43 brief, I knew the moment was meant to feel wholesome, so I aimed for a vibrant and colorful look. At the same time, I wanted to challenge myself by exploring more creative angles and approaches.

With this concept, I wanted to emphasize the height of the treehouse by using an elevated, top-down perspective. This helps establish its distance from the ground while keeping the subject clear and engaging. Since the moment represents a dream come true for the character, I aimed for a vibrant, dreamy, and colorful environment to enhance that magical feeling.

I started with a free 3D asset of a tree trunk that I found on Sketchfab, along with another practice thumbnail I had made for myself, which was a recreation of a Sidemen thumbnail included in the reference. It came in handy because it had the exact pose I was looking for.

I used these two assets together with my sketch to build the basic composition.

I didn’t actually use the 3D model directly. I just rotated and positioned it in the Sketchfab viewport, then took a screenshot — a simple trick you can try too, haha.

I took the first rough sketch and ran it through AI because I was curious to see where it would take the design. I’m usually very detailed with my prompts, so it gave me some interesting directions to explore.

The first result wasn’t great, so I refined my sketch by making it more specific and assigning different colors to different sections. With a better prompt, I ran it again and got a much more usable image to work with.

Then I added highlights and shadows to the treehouse completely manually in Photoshop using clipping masks and painting with curves adjustment layers. After that, I ran it through AI again to generate a new background. I usually describe the kind of background I want to ChatGPT and let it generate an accurate prompt for me.

I usually generate the same concept with 3–4 different models because each one approaches things differently and produces unique results. If I like something from one generation, I keep it, add it into the sketch, and run it through AI again until I get the perfect image I have in mind.

I also removed the small trees on the ground and the fog because they made the treehouse look more like a skyscraper, lol. Instead, I replaced them with a simpler background to keep the scene feeling more natural.

I then cleaned up the thumbnail by removing unnecessary elements, adjusting some proportions using the Rule of Thirds, adding a red slide with kids on it, and finally replacing the guy in the thumbnail with the character from the provided challenge assets using Nano Banana Pro.

At this point the thumbnail was almost done. I just made a few minor adjustments like face retouching, overall lighting, and color grading.

If you found this edition of Unlayered helpful, please consider sharing it with someone who might benefit from this workflow too! 💙