It has taken the Franco-Swiss artist Yannick Jacquet three years to create Élytre, a forty-metre-long generative work on display at the foot of the the Alexander III bridge in Paris. The piece was commissioned as a permanent design feature for Le Flow, a floating building moored along the new pedestrian area on the banks of the Seine.

Yannick Jacquet drew inspiration from the dark mass of the barge between the sky and the river to fine-tune his response to the immediate surroundings, calling on the instability and permanence of the flowing water, the infinitely nuanced shifts of light, and the interplay of transparencies between its large plate-glass windows and the glass dome of the Grand Palais just across the water. Drawing on the barge’s organic, cocoon-like architecture, he came up with a highly sensitive, reactive work in the form of an installation that reverses the overall structural inertia of the barge’s four hundred tons of steel, as if echoing Reyner Banham’s principle of regenerative architecture.

The installation is linked up to a battery of sensors so that it varies according to the time of year, season, atmospheric pressure, wind speed, temperature, and so on. It is in a constant state of flux, permanently subject to imperceptible shifts. As a generative work, it has its own dedicated, custom made software programme and required the artist to work closely with a team of specialists in engineering, craft manufacture, electronics, programming, and architecture.

As night falls, the work softly rises up into the surrounding cityscape. The impression on the viewer’s retinas is deep and lasting. The colourful stimuli of elements emerging and fading seem to mirror the shimmering river and foliage and mimic the circadian rhythm of breathing. The artist also devoted considerable research to the issue of colour. A metal mesh with its own unique structure is fastened over the cladding from the roof to the hull, creating a pointillist effect by means of an infinite palette of pixels. The material resists the quantity of light and contrast: the artist has sought to create nuances and shadings of colour by pushing LEDs beyond their usual capacities.

Yannick Jacquet explains that the installation is part of a broader project exploring cycles and our relationship with time. The work is designed less as an invitation to a journey as an order to slow down. To take the time for contemplation.

Elytre in 6 stone-cold facts:
957 custom-made Leds sections
789 micro perforated sunshield panels
340 different sizes
372 universes of lighting information
4 weather instruments
1 cinder application

> Making of  & Artist interview:

Concept, design, animation: Yannick Jacquet
Producer: Nicolas Boritch Label: Antivj
Software: Eric Renaud-Houde, Simon Geilfus
Hardware engineering: LedPXL

Elytre – video teaser: https://vimeo.com/channels/antivj/225525522

Making of credits: Camera: Corentin Kopp, James Medcraft – Editing: Corentin Kopp – Music: Thomas Vaquié
Pictures by James Medcraft

Remote Memories is a new project by Yannick Jacquet, in collaboration with composer Laurent Delforge (Before Tigers). Remote Memories is a polyptych in panoramic format, a large canvas of video and sound. This highly pictorial work resists immediate apprehension; rather it needs to be observed a moment, contemplated in order to grasp its minor details. Textures are superimposed and interlaced, creating atmospheres that vibrate with neither line nor contour – a sort of “sfumato video.” The image that seems fixed at first is criss-crossed by almost imperceptible waves, like a brownian movement that shakes a gas’s particles. Glimmers, colours, shapes unknown or anxious seem to emerge and disappear as if glimpsed through thick fog.

Remote Memories from ANTIVJ is a visual label on Vimeo.

The soundtrack broadcasted through a series of 11 vibration speakers makes use of the canvas’ wooden panels as a sound box. Just like its visual counterpart, it is composed of vibrant textures superimposed on one another and creating a “drone” that alternates between brooding moods and more luminous sounds.

There is no question of audio-reactivity here since the images and sounds evolve at their own rhythm, mixing, losing ground, drifting, letting chance and coincidence create new interactions endlessly.

The installation Remote Memories invites the gaze to pause and apprehend the impossibility of immobility in an age where data is overabundant.

Remote Memories & the Six Cycles Orchestra – performance view

Remote Memories & the Six Cycles Orchestra

Remote Memories & the Six Circles Orchestra is a performance work adapted from the installation Remote Memories.

Before Tigers (Laurent Delforge) uses 6 turntables to play a collection of custom-made handcut vinyls with locked grooves, conjuring up an abstract composition by adding in or stripping out harmonies and textures to paint a soundscape in conversation with the work’s visual element. Images and sounds develop at their own unique rhythm as the sound and visual loops form intricate layers, intertwine, shift, and fall out of sync, triggering in nite new interactions at random.

The musical composition is spatialised by means of a unique setup combining standard speakers and transducers, or vibrating speakers, creating an echo chamber that encompasses the screen polyptych and the surrounding space. The installation and the live performance explore the tensions engendered by a certain conception of slowness, inviting the viewer to engage in contemplation.

Over the last few years, Yannick Jacquet’s twofold research into colour and the notions of time and natural cycles has led him to flesh out a new paradigm: slowness. Slowness as one possible path to the urgently needed restoration of sensibility.

Remote Memories (installation + performance)
Scopitone festival, Nantes (Fr)
20-24 September

More dates to be announced soon.

Inspiration.

After a few brainstorming sessions with the rest of the team, it appeared pretty quickly that we were going to work with themes inspired by the breathtaking differences in scale found in nature. When looking at the curves of the architecture, the idea of flow came immediately to mind and we started looking into the idea of the sea and the crazy organisms that populate it. I did some research on deep sea organisms and found a couple of articles about the paleodictyon nodosum, its incredible habitat and supposedly faculties for bacteria farming. Without getting too much into details here, the similarities that we found between these organisms and Shigeru Ban’s architecture seemed to be too much of a coincidence not to be looked at.

Our storyboard gave me some pretty clear leads on how to build the the software upon. I had to come up with visual and technical solutions to make our ideas possible. Ideas like this concept of multiple individuals creating, on a higher scale, a big complex organism or this idea of “skin” being a structure constantly rearranging itself in reaction to different stimuli, and so forth… From that point it was kind of easy to see that I was going to play around with agents, particles and group behaviour. But since we’d all been experimenting a lot with these themes in the last decade I really wanted to try to push this further. I had to find a way to make it more interesting for me and not just to create an nth Craig Reynolds’s Steering behaviors implementation.

From storyboard to software.

Instead of hard-coding a particle system as I usually would, I decided that it was time to have a more modular approach to designing particle animation, and invested quite some time trying to find the right solution in terms of usability and creative possibilities. I might have been wrong but it was pretty clear to me at the time that in order to achieve this i would need a good user interface.

I quickly implemented a typical scene explorer, similar to the ones you find in most graphic softwares, and came up with some easy-to-use code to create new objects that could be added to a scene. Having the hierarchies that a scene graph can offer allowed me to design a number of objects and re-arrange at will how those objects could influence each other. I quickly decided to limit myself to a small number of object types per scene. Particle groups, particle behaviour and effectors felt like a good starting points to build what I wanted. It kind of summarized quite well those ideas of internal/external world and stimuli that we had in our storyboard.

Complexification.

When it comes to programming physical processes, I’ve always been fascinated by how combining different layers of complexity can be so powerful. Each layer bringing its new set of rules and surprises. Combining can sometimes result in something greater than the sum of their parts, and that is where interesting and unexpected things can happen.

Usually the first layer that I play with gives each particle different properties, sizes, masses or shapes. “Press play” and see what happens, what kind of interesting patterns or animations emerge when exploring with those different parameters. Sometimes it does really feel like putting your finger into a Petri dish just to see what might happen.

Another layer that I wanted to add was the ability to create separated group of particles and apply behaviours or constraints to a group rather than to every single individual. This layer may seem quite simple or obvious at first, but it allowed for some really nice things to happen, and helped a lot to create complex interactions between particles. Make a small group act like a flock of fish and another one more as a fluid and you already have some nice interactions.

That is where the design that I chose (keeping particle properties, behaviours and constraints as separated concepts) came really handy. Playing with particles and giving them different properties, sizes and shapes is always interesting, but the fun really starts when you can mix different groups of behaviours together.

Timeline animation.

Why bother with coded animation when you can do it with a timeline? This might seems trivial but the level of complexity increased quite drastically when I added a time dimension to those two layers.

I knew that injecting any animated data into a physical simulation can often lead to surprising results but still I was really amazed to see how changing those behaviours over time would create such unexpected reactions. Anyone who has played with Craig Reynolds’s Steering behaviors knows how a small set of rules can create such compelling animations, even if none of the rules parameters are animated. Well if you are that kind of person, then you can probably imagine how animating those parameters can create such surprisingly organic and complex animations. This system helped me to create the different reactions that we wanted for our living organism, like skin contraction and dilatation, structure’s construction, re-organisation and deconstruction, and other organic animations. This was already a big part of our idea of an organism reacting to external stimuli. Here are a couple of examples of that “ever-changing state” structure:

Node Graph and sound design.

There was quite a lot of ping-pong between our composer, Thomas Vaquié and myself. More than ever, the music that he wrote influenced our approach to visual production. More than just a highly collaborative way of working together , we wanted to give the music a real literal role in the piece, making it one of the actual inputs in our (eco)system, like the very stimuli I was talking about previously. The music became quickly the main antagonist in our story, attracting/repelling those organisms, controlling their every move. It also helped a lot to structure our narrative around the birth, life and death of this weird organism, and even led up to an interesting new aspect, that of the balance between two other worlds, light and darkness.

This is one of the reasons why we needed that strong symbiosis between music and visuals. The last few years I’ve been experimenting a lot with audio and particles systems as part of my ongoing collaboration with Murcof and I really wanted to try something new in terms of creation and experimentation possibilities. This is where a node graph came really handy, allowing me to visually route any part of the audio to any part of the visual/physical system. When you are used to re-design the code every time you want the music to influence a part of the animation, well, a user interface like this one is definitely a huge time saver and gives much more space for experimentation.

Stimuli and working with motion designers.

From the start, it was pretty clear that Yannick was going to focus on the more graphical and geometric parts of the piece, and that I would be taking care of the procedural and organic parts. Instead of giving a 4 week old software full of bugs to Yannick, I decided to put my efforts on building bridges between my software and the ones that Yannick would be using. In order to develop fully our story it was really important for us to make those two worlds meet, fight and live together, not only in terms of collaboration and compositing techniques but also conceptually.

I decided to look into computer vision to find simple ways to work with Yannick’s footage rather than the other way around. I built an OpenCV module that was taking care of analysing Yannick’s videos and extracting interesting data that I could use for my animations. This idea gave birth to a nice list of new effects, some of them would extract the polygons out of videos to create collisions, some other would use grayscale gradients to influence the strength of another effect, etc…

It ended up being a really powerful tool, allowing me to use those graphical animation to physically collide with particles, scaring them off or attracting them. And after a few tweaks to the computer vision engine the result was quite convincing. I could just click one button, import new videos, assign them to different effects and, voila, please meet interactive physical compositing!

This module was the last piece I added to the software. Because time is often the main constraint for that kind of project, especially when you are the lead and only developer, there’s always a moment when you have to stop building new toys and start playing with the one you already have!

Artistic direction by Simon Geilfus, Yannick Jacquet, Thomas Vaquié
Producer Nicolas Boritch
Visual content by Simon Geilfus, Yannick Jacquet, Romain Tardy
Music composed by Thomas Vaquié

Header photohraph by James Medcraft.

You can find more information and pictures about the project here.

The software that I built for this project was made possible thanks to the huge efforts and energy of Andrew Bell, the Barbarians and the amazing Cinder community. A big thanks to the whole team for creating such a powerful framework! Cinder rules!

I used Gwen GUI to build the user interface, Gwen is a small library written by Garry Newman, and it is definitely worth having a look at it! I started playing with this library several months before this project and had to hack it quite a lot to make the timeline and nodegraph widgets possible, but it is without any doubt a really nice piece of code!

Post written by Simon Geilfus