The project is inspired by the now-extinct Arena Coliseo de Monterrey, which first opened its doors in 1955 and was demolished in 2023. Known as the cathedral of wrestling in Monterrey, Mexico, this iconic venue held deep cultural significance. While studying Visual Development for Environments and Characters, I wanted to pay tribute to this historic place by creating an animated scene entirely from scratch.

Rule of Thirds

The most commonly used guideline, making a scene feel more natural and visually pleasing.

50/50 Composition

Creates a sense of balance through symmetry, harmony and a feeling of formality within the scene.

90/10 Composition

Emphasizes a specific area by dedicating 90% of the space to it, drawing the viewer’s attention.

80/20 Composition

Showcases more of the background while keeping the most important elements within the 80% area.

Using procedural workflows speeds up the creation of objects that involve heavy repetition while still providing unique variations for each instance.

First, I start by modeling the Mesh Input (1). Then, I add a Mesh Line, setting its direction along the X-axis (2) with a fixed number of points, 11 in this case. I use this Mesh Line as the points for the Instance on Points node. The Mesh Input is assigned as the instance, and I apply a Random Value node to rotate it along the Z-axis for variation. (3)

The base starts with a Plane (1) as a Mesh Base. Then, Separate each face and Scale them individually (2). After that, Extrude (3) the shapes and add some Distortion (4) to the top section for variation. Finally, set the Material (5) to complete the look.

I start with a Cube primitive, modifying its shape to resemble a brick and applying a Material (1). Then, I instance the brick multiple times (2) and use a Random value to give each one a slightly different depth (3). To break uniformity, I introduce horizontal offsets (4) and randomly remove some bricks (5). Finally, I apply a Boolean operation (6).

I start with a Grid as the base (1) and a Cube primitive to serve as a brick (2). Then, I create two staggered rows and merge them to form an interlocking Brick Wall (3). Next, I align the excess bricks on the left side (4), followed by the right (5). Finally, I adjust their Position and Scale (6) before applying a Material (7).

I start with a base plane (1), then create Instances of cubes using the Cube node and randomly remove some (2). Next, I apply distortion (3) to break uniformity and finish with a Material (4).

By leveraging procedural techniques, we can create animations that would be time-consuming or impractical with traditional keyframing. This approach allows for dynamic behaviors, automated motion systems, and parameter-driven effects that adapt in real time.

We start with a Base Mesh (1). Then, duplicate and position it to form the fan structure (2). Finally, use Scene Time to animate the rotation along the Z-axis (3).

Start with a Path Curve as the base (1). Then, select specific vertices using Index (2) and animate them with a Noise Texture (3). Finally, add Thickness to the curve (4) and apply a Material (5).

Start with a Sphere as the Base Mesh (1). Then, create a simple moth shape entirely within Geometry Nodes (2). Using the Base Mesh as a Volume, I distribute the moth instances inside it (3). Finally, animate them with Scene Time (4) and apply a Material (5).

Start with a Path Curve as the base for the cable (1). Then, create the shape of a single flag (2) and animate the cable using Scene Time and a Noise Texture (3). Next, use the curve and the flag to generate an array of flags along the cable (4). Animate the flags (5) and randomly delete some of them to create variation (6). Then select specific flags to apply the Red Material (7), Blue Material (8), and White Material (9). After that, give Thickness to the spline and smooth it with a Subdivision (10). Add a Switch to toggle the visibility of the flags (11) and, finally, combine everything using Join Geometry (12).

Creating Masks was essential for achieving varied effects, as I wanted to avoid flat, solid colors. By Combining Multiple Masks, I was able to generate more complex textures while maintaining a stylized look that complemented the modeling.

In this section, I will showcase different node combinations used to create the masks for this project.

This mask is created using a Voronoi Texture node. By utilizing the Distance output, we can generate distinct cell-based patterns. To refine the mask, a Color Ramp node with Constant interpolation is applied, allowing for sharper transitions between values.

This mask is created using a Voronoi Texture node with the Distance output. By increasing the Randomness parameter, the pattern becomes more irregular. A Color Ramp node is then applied, using Constant interpolation to create sharp transitions.

This mask is created using a Voronoi Texture node with the Color output. A Color Ramp node is then applied with Linear interpolation to achieve a smooth transition between values.

This mask starts with a Voronoi Texture node. A White Noise Texture (4D) node is then used to modify the pattern by adjusting the W float value. Finally, a Color Ramp node with Constant interpolation is applied to create sharp, defined regions.

Use the Voronoi node in 2D with Distance to Edge, followed by a Color Ramp using Constant Interpolation to refine the effect.

Use the Wave Texture node, followed by a Color Ramp with Constant interpolation to create sharp, well-defined lines.

Use the Brick Texture node, allowing adjustments to Scale, Mortar Size, Width, and Height to refine the pattern.

Use the Gradient Texture node in Linear mode, though other modes offer interesting variations.

We start with a Voronoi Texture node, combined with a Color Ramp set to Constant to create distinct point patterns.

Next, we add a Gradient Texture, controlled by a Mapping node and a Texture Coordinate node, allowing for adjustments in position and orientation.

Finally, we blend both elements using a Mix Color node with the Multiply blending mode, merging the gradient and point patterns into a cohesive mask.

We begin with a Voronoi Texture node, followed by a Color Ramp set to Constant, adjusting the ramp’s position to refine the pattern.

Separately, we use another Voronoi Texture node with a smaller scale, also processed through a Color Ramp.

Finally, both patterns are blended using a Mix Color node with the Screen blending mode, creating an overlapping square effect.

We start with a Voronoi Texture node set to output Color, followed by a Color Ramp with Constant interpolation and an additional stop to refine the effect.

Separately, another Voronoi Texture node is used, this time set to output Distance. This output is processed through a Color Ramp, inverted, and also set to Constant interpolation.

Finally, both patterns are blended using a Mix Color node with the Multiply blending mode, creating a scattered dot effect within specific areas.

Start with a Voronoi Texture node set to output Distance, with Randomness enabled. This output is processed through a Color Ramp with Constant interpolation to define distinct regions.

Separately, another Voronoi Texture node is used, also set to output Distance with Randomness enabled, creating a complementary variation.

Finally, both patterns are blended using a Mix Color node with the Multiply blending mode, resulting in an organic, interconnected worm-like structure.

Start with a Voronoi Texture node set to output Color, with a slight Randomness variation. This output is processed through a Color Ramp with Constant interpolation, where an additional stop is introduced to define clearer regions.

Separately, a Wave Texture node is used, combined with a Mapping and Texture Coordinate node to allow rotation along the Z-axis. This pattern is also passed through a Color Ramp with Constant interpolation to create sharp transitions.

Finally, both elements are merged using a Mix Color node with the Multiply blending mode, producing a structured interplay of lines distributed by defined Voronoi areas.

Begin with a Voronoi Texture node with slight Randomness, followed by a Color Ramp using Constant interpolation to define distinct areas.

Separately, a Wave Texture node is used alongside Mapping and Texture Coordinate nodes to control rotation along the Z-axis. To distort the wave pattern into an organic shape, a Magic Texture node is introduced. This distortion is further refined with two Vector Math nodes, while a Value node is used to uniformly modify three axes with a single input.

Finally, both elements are blended using a Mix Color node with the Multiply blending mode, creating a structured yet dynamic wave pattern influenced by the Voronoi regions.

The process begins with a Voronoi Texture node with slight Randomness, followed by a White Noise Texture node set to 4D mode. This noise is then refined using a Color Ramp with Constant interpolation to control contrast.

To manipulate orientation, Mapping and Texture Coordinate nodes are used, allowing rotation along the Z-axis.

Additionally, another Voronoi Texture node is introduced, combined with multiple Vector Math nodes to further deform the primary Voronoi pattern, adding complexity and variation to the final texture.

This setup begins with a Voronoi Texture combined with a White Noise Texture in 4D mode to control the dark areas, followed by a Color Ramp for contrast adjustment.

Separately, a Brick Texture node is used, enhanced with multiple Color Ramps to generate different effects from the same texture. Finally, these elements are blended together using Mix Color nodes, applying Multiply for depth and Screen for highlights, creating a more intricate layered brick effect.

To set this up, we first add a Geometry Nodes modifier to the mesh. Inside the node setup, we only need two nodes: Scene Time to retrieve the animation time and Store Named Attribute, set to Float and Point mode, with a custom name like “time”.

In the Shader Editor, we then add an Attribute node set to Geometry, using the same attribute name. With this, we can animate textures by moving them along an axis and control the speed of the movement using a Math Node using Multiply.

For this project, I relied on various references that clearly outline the essential elements of character design. While there isn’t a single correct method, understanding core concepts such as shape language, the line of action, and silhouette can provide a structured approach to designing characters effectively. In the following sections, I will break down each of these key aspects to demonstrate how they contribute to strong and visually appealing character designs.

Realism: The closer a design is to this point, the more detailed and true-to-life it becomes. This approach is often used for unique, highly individualized characters, such as hyperrealistic portraits.

Symbolism: This category includes simplified and stylized designs, making characters more universally recognizable. Examples include emojis or the iconic characters from Peanuts (Charlie Brown).

Abstraction: At this end of the spectrum, the focus shifts towards pure visual expression rather than literal representation. Think of cubist art or abstract compositions that use shapes and colors to convey emotions and concepts.

Breaking shapes into large, medium, and small sections helps maintain balance and visual interest. A strong hierarchy prevents clutter and ensures the design remains readable at a glance.

Shapes influence how a character is perceived. Circles create a friendly and inviting look, squares suggest stability and reliability, while triangles add energy and sharpness, making a design feel more dynamic.

Weight distribution affects a character’s presence. A lower center of mass makes them feel grounded and strong, while a top-heavy design can suggest agility, instability, or dominance. In this case, is a top-heavy type.

The Line of Action is an imaginary curved guideline that defines the character’s posture and movement. It creates a natural sense of rhythm and direction, enhancing the clarity and intent behind each pose.
A strong Line of Action helps convey emotions and storytelling. Whether a character is leaping, slouching, or striking a dramatic stance, this foundational shape reinforces their attitude and purpose within a scene.

A clear silhouette ensures that a character’s pose and actions are readable even without colors or linework. Strong silhouettes make characters more memorable and their gestures more expressive. Beyond recognition, silhouettes also help define weight distribution, creating a balanced and believable stance. The way shapes are arranged within a silhouette can also guide the viewer’s eye, leading attention toward key areas like the face or hands.

In this vision, Blue Demon stands as a guardian, keeping the doors of the Coliseo open. As long as he remains at the entrance, the arena lives on, not in reality, but in our collective memory. This concept influenced the character’s pose, emphasizing his role as a protector. Another defining aspect of Blue Demon was his signature formal attire. Unlike traditional luchadores who are often seen in wrestling gear, he was known for his elegant suits. To reflect this distinguished presence, I chose to depict him dressed in a more serious and dignified manner, reinforcing his status as both A legend and a guardian of the past.

To achieve this, we use the File Output node, which allows us to automatically direct multiple render passes to a designated folder. This node provides flexibility by letting you specify the file name, format, and additional properties for each output. By setting up a structured file output system, you can streamline your workflow, keep your files organized, and avoid the risk of missing important passes during post-production.

To achieve this, we use multiple File Output nodes, ensuring that each pass is neatly stored in its own folder. This setup makes it easier to import and manage footage in DaVinci Resolve without sorting files manually. While using an EXR Multilayer would be more convenient, it doesn’t retain the applied Color Management, so I opt for 16-bit PNGs instead to maintain color accuracy.

I will break down each step done in Photoshop so it can later be replicated in DaVinci Resolve. Understanding each adjustment makes it easier to translate the look into video. While these techniques and workflow are not mandatory for every project, they help shape the final aesthetic. In this case, my goal was to create a nostalgic, aged, and slightly faded atmosphere, evoking the feeling of a memory.

First, in Camera Raw, increase Texture and Clarity to enhance sharpness. Then, I use Levels to brighten the darker areas. I apply the Ambient Occlusion Pass with Multiply at 50%, adjusting it with Levels to fine-tune its effect. I also add the Color Denoise Pass with Screen at 5% to smooth the image.

Layer all Light Passes with Screen at 100%, blending them with the Mist Pass and applying a warm tint to enhance depth.

Use multiple Levels to selectively lighten and darken areas, creating a Vignetting effect. Finally, I apply several LUTs and reduce Saturation, shifting the image towards muted, grayscale tones.

To ensure compatibility, LUTs need to be properly exported from Photoshop, following specific rules. First, any elements that shouldn’t be included in the LUT must be merged into the Background layer. Before exporting, select only the layers that contain the color adjustments. Then, navigate to: File > Export > Color Lookup Tables, choose the desired formats, and save the LUT for use in other programs.

DaVinci Resolve has become my go-to tool for video creation due to its speed and efficiency. The free version is incredibly powerful, and the full version is available at a very accessible price. It offers specialized workspaces for color grading, editing, audio, and compositing, making it a versatile choice for post-production.

For this project, I worked mainly in Fusion, DaVinci’s Node-Based Compositing System. My biggest challenge was replicating the adjustments made in Photoshop within Fusion. After experimenting, I found that many techniques translate well between the two, with masks being one of the key differences. This section will focus on those unique aspects.

For the Polygon Mask, I created a Polygon Node and connected it to the Effect Mask Input. This is the simplest method and works well for precise manual selections.

For the Mist Pass Mask, a direct connection doesn’t work, so I used additional nodes. First, I adjusted the Brightness and Contrast to control the mask’s intensity. Then, I applied a Luma Keyer to isolate the bright areas. Finally, I used a Channel Booleans Node to combine the mask with the target image, integrating it seamlessly.

Initially, I planned to render the animation directly at 12 FPS to save rendering time. However, I realized that DaVinci Resolve does not provide a straightforward option to change the playback frame rate in this way. Instead, I used the Stop Motion Node, which allows selecting how many frames should be held before advancing to the next. By setting it to hold every second frame, I effectively reduced the animation to 12 FPS. If I wanted an even choppier effect, I could set it to hold every third frame, resulting in 8 FPS.