Matyas Rathonyi

Interstellar is my favourite film, and Gargantua is one of cinema's most scientifically grounded black holes. Its image was generated with a renderer based on relativistic light propagation, although some effects were simplified so the audience could read it. I wanted to understand the visual logic behind it in Blender.

I rebuilt the mathematical shader from a tutorial. The difficult part was not placing the nodes, but understanding why the coordinate transforms, curves and math operations produced the distortion. The final accretion-disk material contains 102 nodes and 117 connections. Every model and the final composition were made by me.

The project took more than a week because I kept returning to the mathematics until it made sense. It became my favourite still project and gave me a much better understanding of procedural shading. Feedback from a friend helped me refine the final lighting and framing.

I have worked through every Polyfjord tutorial because I like the way he combines clear technical systems with playful forms. The mechanical Kraken was my favourite, so it became the only direct tutorial recreation I selected for the portfolio.

The model and construction follow Polyfjord's tutorial closely. I recreated every part myself in Blender, but the original design and method are his. My own work begins with the colouring, materials, lighting and framing. I used those decisions to turn the same 212-part model into a cold blue and a glowing red version.

The two renders taught me how strongly colour and light can change the personality of an unchanged object. The project is also a useful marker of how I learn: first I rebuild a system carefully, then I use the finished structure to test my own visual decisions.

I wanted to see how much character I could create from an extremely simple body and a readable silhouette. The basic rig came from a Polyfjord exercise, while the dark setting, grey body, red edges and wide composition were my own choices.

The creature uses eight armatures, IK targets and curve-based limbs. Rebuilding the rig made it clear how a few controls can coordinate many legs without making the system impossible to pose. I kept the modelling intentionally simple and spent more time on the final visual balance.

It is one of my simplest projects, but also one of the most visually direct. I built an animation for the creature as well, and I still think the movement is one of its strongest parts, even though I did not have time to finish it for this portfolio version.

I gave myself a simple constraint: choose two unrelated assets at random, then make them feel as if they belonged in the same scene. The scale difference between an industrial KUKA arm and a small rubber duck created a strange mix of tension and humour.

The two models are external assets, while the idea, animation, materials, lighting, camera movement and sound design are mine. I first made a still to establish the spotlight and composition, then turned it into a short vertical sequence in which the arm enters the darkness and removes the duck.

The project showed me that a clear constraint can produce a stronger idea than starting with an unlimited scene. It also became an exercise in visual storytelling: the viewer understands the relationship between the two objects without any explanation or dialogue.

I found an image online of a vehicle covered in vegetation and used it as a starting point for a Blender recreation. I liked that the object was immediately recognisable as a car even after its normal hard surface had been replaced by something soft and organic.

I modelled the simple car silhouette by hand, then used particle systems to build the grass surface. The main challenge was balancing density, length and lighting so the result looked like vegetation instead of fur. Keeping the entire scene within a narrow green range made the silhouette more important.

This was a compact experiment, but it taught me much more about particle-system control than a larger scene would have. It also showed me how a single material change can become the entire idea of an image.

A friend asked me to make a desktop background, which gave the project a clear format and purpose. I chose a loose cluster of connected bubbles because it could remain calm and minimal while still changing across the wide image.

The modelling stayed simple so the shading could carry the image. The difficult part was creating clear edges, transparency and slight rainbow dispersion without turning the forms into heavy glass marbles. I adjusted the material and lighting together until the bubbles remained visible on a pale background.

The final image works as a background because the detail stays concentrated around the forms and leaves enough quiet space around them. The project mainly improved my understanding of transparent materials and how much their appearance depends on the environment.

The project began with a Ducky3D tutorial, but I did not want the final frame to feel like a direct copy. I treated the radial form as a piece of light art installed on a rough architectural surface and pushed the contrast between the precise glow and the imperfect wall.

I rebuilt the core technique, then changed the form, materials, surface treatment and framing. The scene combines a repeated radial object, internal lights and layered concrete textures. Most of the work went into keeping the pink light readable without flattening the wall around it.

The finished image feels closer to an installation photograph than a motion-graphics exercise. It taught me how to take a tutorial technique and give it a different atmosphere instead of only changing colours.

For a visual-culture assignment, I had to create an artwork or artefact connected to Rococo. Instead of reproducing a historical object, I designed a piece of furniture that exaggerates decoration, softness and luxury through contemporary 3D materials.

I modelled the entire object by hand. The segmented lid creates a jewellery-box or piano-like rhythm, while the twisted stacked legs make the table feel ornamental and slightly unstable. I used translucent pink surfaces against pale stone so the object could feel both precious and artificial.

The result is not a historical Rococo reconstruction, but my own interpretation of its excess. The assignment showed me that 3D can be used for object and art direction, not only for realistic reconstruction or animation.

I wanted to build an animation around one clear physical interaction rather than a large environment. A hard cylindrical object enters a cloth surface, and the entire image is defined by the tension spreading out from that contact point.

The difficult part was matching the scale of the objects with the cloth settings. Small changes to size, stiffness or collision distance made the surface look either solid or weightless. I ran repeated simulations until the material stretched dramatically without breaking the basic shape.

The monochrome lighting keeps attention on the deformation instead of the materials. The project gave me a better instinct for simulation scale and for designing an animation around the behaviour of one system.

Both pieces began from motion rather than a finished still. I was interested in simple loops that could keep changing without needing a story, character or camera cut. Their shared subject is how a procedural system can create rhythm from repeated forms.

Orbit was based on a Polyfjord tutorial. I first watched the finished animation, then tried to rebuild it independently and only returned to the tutorial when I became stuck. This made the Geometry Nodes simulation and the movement around the columns much easier to understand than copying every step in sequence.

Noise is a simpler Geometry Nodes wallpaper, loosely inspired by Ducky3D. Instead of a readable path, it uses shifting procedural variation to keep a small surface alive. Together, the two loops taught me the difference between directed motion and ambient motion.

This was a direct test of how well I could design a large environment without relying on another program or a prepared scene. There was no longer story behind it. The goal was to make one simple symbol feel monumental through architecture, repetition and scale.

I modelled the corridor as a repeated sequence of dark vertical elements, then placed an inverted neon triangle at its centre. The reflective floor extends the geometry and gives the gate more weight, while the red light is the only strong colour in the frame.

The final still proved that I could create a convincing sense of scale from a small number of elements. It also became a useful lesson in restraint: most of the atmosphere comes from spacing, darkness and reflection rather than modelling detail.

I wanted to recreate the uncomfortable familiarity of the Backrooms without treating it as a static hallway render. The scene needed to feel recorded by a person who was already moving, confused and possibly being followed.

I modelled the environment and creature, built a 352-frame camera path and rendered the sequence in Eevee. The most important part was the camera shake. I adjusted its direction, rhythm and strength until it felt connected to running footsteps instead of looking like a generic handheld effect.

Backrooms is one of the projects that best represents what I can do because its effect comes from many layers working together: modelling, lighting, camera movement, pacing and atmosphere. None of those parts is enough on its own, but together they make the short sequence believable.

ZORD is a real drink brand connected to Balázs Kicks. I created the two pieces independently as concept advertising. The first uses a bottle breaking out of a clear ice cube to communicate cold refreshment. The second turns the word launch into a literal rocket launch because the product had just been released.

I made every model and both scenes myself. The ice setup contains around 800 fractured cells, while the rocket scene combines the bottle, launch structure, animated cameras, lights, fire and fluid-based smoke. The rocket took several weeks because each layer had to work both technically and as a short vertical advertisement.

I sent the animations to the brand, but did not receive a reply. The direct feedback came from friends instead. Even without becoming a real campaign, the two pieces are among the clearest examples of my ability to build a complete product world from concept to final animation.