Rookie Awards 2024 - Open for Entries!
Mattia Binda - 3D generalist portfolio
Share  

Mattia Binda - 3D generalist portfolio

Mattia Binda
by mbinda on 29 May 2022 for Rookie Awards 2022

This is my 2022 portfolio, created during the MA 3D VFX Production course at Escape Studios. From a stylized short movie, to photorealism, I strived to learn a wide variety of software and techniques to soon become a professional 3d generalist.

4 1431 1
Round of applause for our sponsors

This is my final showreel, which includes projects I've worked on in the past year as a student at the 3D VFX Production MA course at Escape studios. Below you can find more information and breakdowns of my various projects.

Introduction to the Light

Starting from my latest work, and the one I'm the proudest of, Introduction to the Light is an experimental short movie I created as part of my final module.

In this project, I decided to explore VFX as a storytelling medium, to also test my filmmaking and project management skills.

As opposed to some of my other work, in this case I went for a less photorealistic and more stylized approach.

Everything CG was created by me from scratch, backplates were shot with the help of my friend and Director of Photography Samuele Bogni.

Inspiration and references

The short was partly inspired by Miracle Musical's Introduction to the Snow, which also serves as soundtrack, and by UVIVF photography.

Another source of inspiration was the Zeniarai Benzaiten shrine, both for its look and for its legend, which was an inspiration for my story. It is in fact said that the temple's sacred water springs hold the power to multiply money.

UVIVF photography is quite interesting: when shining a UV light on certain subjects, it gets reflected back as visible light, and colours can change completely, often becoming more saturated and intense. Below you can find an example with Hellebores lit under normal and UV light. Pictures were taken by me. Aside from the camera's native colour management and colour temperature corrections (to account for the UV light's natural blue spill) the colours were not changed in post-production.

Other than more pictures I took myself, other references for the flowers were some of Craig Burrows' UVIVF shots and some footage from the paper Plants with genetically encoded autoluminescence (Mitiouchkina, T. et al., 2020, Nat Biotechnol 38, 27 April, pp. 944-946.)

Modelling

All models were created in Maya and Houdini. 

The structure of the main temple was based off of the aforementioned Zeniarai Benzaiten shrine, among a few other references, such as the Higashi-Tenno shrine, which I used as inspiration when modelling the roof.

The Temizuya (the water pavilion) wasn't inspired by a particular temple or structure; I decided to create something that would fit with the style of the existing temple by re-using parts of it (such as the wooden beams and roof).

The rope and flower are both procedural. Some variations can be found below.

Texturing

All textures (except for the Shimenawa, the sacred rope on the temple, which uses a procedural shader) were created with Substance Painter.  The circular shape, which I used as a recurring detail in some textures and models is a Mitsudomoe, a symbol often found in Buddhist and Shinto temples.

Simulation

All simulations were created in Houdini; FLIP for the droplet and coin splash, RBD for the Ema (small wooden planks) moving in the wind and particles for shots 4 and 5. I also used Maya's Ncloth to create some static props used in the background, inside the temple.

Rendering and compositing

To keep as much freedom as possible in comp, I decided to render three separate layers, one for each type of light: environment, lantern and particles. This allowed me to balance the three light sources in comp, which helped with integration, especially as the backplates were shot in daylight and had to be colour-corrected to fit the night setting. Everything was rendered with Arnold. Render times averaged about 10 minutes per frame for the least complex layers in simple shots (eg. for the particle light layer in shot 1) and up to 45 minutes for more intensive shots and layers (eg. for the environment light layer in shot 4).

Below you can find the breakdown for shot 3.

An Old Dusty Boombox

This was my first module at Escape, which revolved around integrating a CG object in a photographic backplate, making it as photoreal as possible. Everything, including the backplate (except of course the logo designs), was done by me.

For this project, I decided to recreate my father's old radio. Not only this allowed me to use the real object as a reference, but the radio itself felt like it had a story: this radio had been used for who knows how many years and had been who knows where, and both the wear and stickers on it show this perfectly.

The reference images I shot allowed to create an accurate model without issues.

The model was created in Maya and textured in Substance painter.  For lighting, I used an HDRI I shot with the help of a reflective sphere along with a simple light setup to better recreate the correct shadows on the model.

Rendered with Arnold, composited in Nuke and colour-corrected in After Effects.

Lake Pier

This project was once again focused on photorealism. Tracking was done with 3DEqualizer.

Working on a set extension was quite interesting: at first, I was unsure of the photographic composition when shooting the backplate, however I soon realised that the CG itself would play a very important role.

Creating a quick mockup in Photoshop over a picture taken at the location helped define roughly the shape of my pier, thus allowing me to keep in mind how the final product would look while shooting.


Once again, all modelling was done in Maya, textures in Substance Painter and Zbrush. The model was lit thanks to an HDRI shot on a Theta SC2 and a simple light setup consisting of a sun lamp and some area lights.

Rendered in Arnold and composited in Nuke.

The water reflections were created by using a reflective plane, and using a water node in Hypershade as a bump map, which was animated to match the real waves.

Dancing with the Ivy

This project was a bit different: rather than focusing entirely on photorealism, the goal here was to create either something procedural or simulation-related with Houdini, so I decided to focus on both, by creating a procedural ivy system that could grow by following a guide curve, while chipping out bits of the wall, which also create some dust.

Everything CG, except the leaves' textures (which are from Quixel Megascans) was created by me. Backplates were shot by Yuri Tuena and his crew and tutors as part of a second-year graduation project in CISA Locarno (Conservatorio Internazionale Scienze Audiovisive). The footage was kindly provided by CISA.

The ivy generator is completely curve-based: some curves are generated on a target point, then rayed on the geometry surface and distorted via a path deform SOP. Points are then spread on the resulting nurbs structure, which are used to generate a random leaf or berries. Leaves were animated using CHOPS and bend SOPs.

The guide curve is also used to transfer "active" and "velocity" attributes, which allow to control the progression of the animation and simulation at the same time.

The simulation is based on an RBD material fracture node; only the pieces generated using the chipping option get simulated. 

Being my first project in Houdini, there are a couple of things I'd do differently, if I had to recreate this project from scratch. Rather than using a nurbs-based system (which introduces some limitations when controlling the overall shape of the ivy), creating a particle-based system similar to Disney's V-GRO tool (Konersman, 2010) would allow for way better creative control, along with providing a framework to use the same tool to create other types of plants.

Everything was rendered in Mantra and composited in Nuke.

Closing remarks

I'd like to thank my friends, classmates, and especially my tutors at Escape, for their precious feedback and help.

I hope you liked my entry!


Comments (2)