Blender Add-ons
Blender Add-ons
Videos
Interviews
Events
Motion Capture
Every few years the Blender Foundation launches a new creative project to help drive Blender development forward in the environment of a real production system and to showcase the software's capabilities. Due to the success of the first such project, Elephants Dream, which premiered in 2006, the Blender Institute was founded. Fast forward nine years and we find a film that shows how much Blender has progressed and matured as a 3D tool. We've seen the open source 3D software show up in the headlines much more often in recent years due to the quality of the work produced with it, as well as to the add-ons developed by a growing and enthusiastic user base.
In addition to its technical merit, Cosmos Laundromat also surprises due to its script and sense of humor. Resources have been invested in developing an elaborate, non-decaffeinated script, which implies taking risks but also results in an original film. Thanks to this it has received wide acclaim from the public and praise from artists from Pixar, Disney, Dreamworks and ILM, among others. The production team was invited to screen the movie for artists working at Dreamworks and Pixar, and the short has recently won the Jury's Prize at the Animago Festival.
We had the pleasure to interview Mathieu Auvray, director of the film (known also for his "Babioles" TV series), who shared a detailed account of the production of Cosmos Laundromat.
Cosmos Laundromat is an ambitious project. Please tell us what were the basic guidelines set by Ton Roosendaal for the film and how you got to join as director.
I met Ton in 2011 for the first time, when the production company I work with in France (Autour de Minuit) was searching for people to make the “Babioles” series using Blender. Since that time we kept exchanging emails. Two years later Ton invited me to the Blender Conference. I was very pleased to finally go there !
Ton had already mentioned publicly that he wanted to make something special for the next open movie project, and at the conference he explained more in detail what he had in mind : “The Gooseberry Project” : an open source feature film made by multiple studios from around the world. The idea would be that each studio would make 8-10 minutes of the whole movie, each with their own style and art direction.
And on top of that, he asked me to direct it. Bam ! There was no way I could let pass such an honor and opportunity ! He told me this on the last day of the conference and I couldn't sleep that night. All kinds of questions crossed my mind.
I knew we had to make the constraints part of the story, that I had to come up with an idea that would make every scene different from an artistic direction point of view. But I also knew that to make such a thing pleasant to watch, we should have a strong story and something that provided continuity along the narrative (such as an actor's voice, the quality of the animation, performance from the actors, etc.).
So I started thinking about the story, the core meaning of the film, the main emotions and characters. We talked about it with Ton and we agreed on a story that will not make a statement, to carry a political message. What he wanted was an animated movie that was as entertaining as possible. So we had to open wide the doors of creativity !
A lot of work has been dedicated to the script. How was the process of working on it with Esther Wouda?
First, I worked with a French scriptwriter I know well : Regis Jaulin. Regis is a very talented scriptwriter but he was not very comfortable with the “love story” idea... I loved Sintel's script and when I met Esther Wouda (she wrote it), we instantly felt connected. So we talked a lot, and started developing ideas and stories almost from scratch, again...
Since we had to make each scene very different, we thought of a story in which the main character would appear every 5 minutes as something or someone totally different like “Quantum Leap” but with the drama and humor of “Groundhog Day”. The main character could become a robot, a chair, a tree, a fish, a samurai wolf... Anything !
Having characters that change appearances in every scene could be a wonderful support for a love story ! You get rid of the “appearances” of someone through repetitions and focus on who they really are and their personality will show up even more in the long run. You'll end up seeing only who he really is ! As we say in French : “the outfit doesn't make the priest”.
So we wrote a basic feature film complete beat sheet. What happened to the main three characters, how they are connected and how they will interact with each other and most importantly : why. Since the crowdfunding didn't reach its target, we focused on the first 17 minutes only.
Unfortunately, as the production started, we realized that 17 minutes was too much if we wanted to have good quality animation / render / simulation / lighting... So we had to cut it down to 10-12 minutes. And by doing so, I had to remove what mattered the most for the big story : the love story. Instead, we focused on having a strong beginning that would lead to a “good” frustration feeling when the end credits would appear : I want more of this ! And that's what we were aiming for mostly, for people to get interested in the project and to support us to continue it.
The great thing was that Esther and I shared a common feeling of what we wanted to put on screen. “Cosmos Laundromat” is about love, and how complex it can be. We wanted to write a story that would reflect the feeling of 'letting go' (without any annoying song !). When someone you love doesn't give this back to you, you have to let it go, even if you don't want to. That's what motivated us at first, that's the key thing we were after. So I don't want to spoil it (even if most of it can be read online I think) but we wrote this movie focused on this feeling, this emotion, and to make the audience care about this sense of abandon, of sacrifice, that can tear you apart…
This is something I wanted to talk about from my personal love life at the time. But also because it raised a very contemporary question : doing what's right. Making a good decision and acting like you should is something that cannot be more important nowadays. Doing what's right for someone you love, making the right decision not for you but for others. Putting your ego, yourself out of the equation and stopping to be selfish or self-centered in order to make good things for others, is really a concern I think. It's not a black and white thing, it leads to a lot of grey areas, to subtle things... It's also a story about stopping a curse, changing things that you think you can't...
Tell us a bit about your visual strategy and what you wanted to transmit visually.
For the first scene, the Island, I wanted to anchor this into reality the most I could. I wanted to plant the idea that “this is the real world”. The first world, Franck's world. So I decided to shoot it as if there was a camera guy holding the camera, almost like a documentary. To be with Franck at this very specific moment, maybe his last. And with the mood of the scene, its pace, we would feel with him how desperate and empty he feels.
But as soon as the tornado picks up Franck and he's sent into the other worlds, the camera and cinematography would have to be treated more normally with proper dolly moves in a more standard way, to be more discrete and focus a bit more on telling the story.
The pilot was supposed to be longer (17 minutes) with a longer beginning (Franck talking with other sheep, taking more time at the border of the cliff...) and a whole scene for the jungle at the end to introduce the female character. Since we agreed on focusing on quality instead of quantity, I always tried to keep the movie under 100 shots so that every one of them would matter and would be polished. Because I think that in CG having great, believable shots is a character on its own, story-wise. If you can make those images look real, you care more about what happens in the film. At least, that's what happens to me !
Comparing this short to the first open movie, the great Elephants Dream (Bassam Kurdali, Andy Goralczyk - 2006), shows how much Blender has advanced technically in the last 9 years. The aesthetic in Cosmos Laundromat is impressive. How did the team approach environment design and FX work?
Ton kept pushing us in the early writing phase to include physics in the story. And since my first idea for the main character was this sheep on an isolated island, we had to make great hair that moves. Also, originally the island was supposed to have some very long grass. It would have needed a strong refactor of the physics in Blender to make multiple hair systems act all together. And finally the tornado was an idea to keep pushing the boundaries for the development department. And on top of that we all agreed on aiming for a render quality and style close to big studio movies like “Rango”…
So all of these challenges were included in the story to push Blender as much as we could for this pilot. It was very ambitious and it also slowed down the project a lot. Its final duration, and the way it was designed suffered a bit from it because we had to balance quality and quantity. But if you think it looks impressive, we did a good job !
The work on lighting also stands out. Please tell us a bit about it.
I did a lot of tests and research. I had a lot of references from movies and pictures to define the mood and tone of the island, the laundromat and the jungle. The island changed a lot, at first I was aiming for a very dark, foggy and gray early morning with dew and low light. But we couldn't make this look nice and since we had a “depressing” opening we decided not to over charge the sadness and bring some hope to what Franck was doing. So I went for a sunset, with Franck going to the sun, to leave his darkness behind.
I'm still a bit not totally pleased by this decision since it would have made it even more obvious how bad Franck's reality/world is. Also we wouldn't have seen things because of the fog, and it would have been more clear that he lacked perspective and future because he couldn't literally see anything. We might have felt more the reasons he had for killing himself, to end this boring and pathetic life...
But it would have been a very, very depressing opening ! I tried to make a compromise that still works and make sense, and doesn't scare people away...
I worked on the lighting a lot, too. And we tried to keep it simple and consistent. So it's mostly and environment map emitting diffuse lighting and a sun. That's it ! Sometimes we added an area lamp to highlight a rim or to bounce a bit more light in a shadow. But we tried to find a simple thing that would be balanced easily through the shots and not over-tweak each setup. It helped a lot to sell the organic and natural look of the camera work. With a bit more time we would have added some subtle 2D effects too, like flares, glows, soft edges on lenses and motion blur.
But that's the time we had, so we focused on a pipeline where final frames were rendered straight out of Cycles. It's amazing because you can see and tweak final renders in the viewport, but it also brings out the situation where final frames are difficult to tweak much without re-rendering.
There's plenty of work dedicated to characters' expressions. How were the facial rigs built? Also, what improvements were added to Blender to produce the quality hair/fur and its animation?
Juan Pablo Bouza and Daniel Salazar were working remotely from South America on the rigs. And Angela Guenette who modeled them too, from North America. They did a great job !
And to me, most of the credit for the quality of the expressions goes to the animation crew : Hjalti Hjalmarsson, Sarah Laufer and Beorn Leonard. They did a terrific job interpreting what we've shot with the actors. They shot their own references also to dig even deeper into some gestures and poses.
On a side note, I also think that the rigs and the facials could have been better. Designing the characters took a very long time. And we couldn't test them enough.
Having tests with the actor's performance as references while you can still model/rig and test animations is a key part of character design and I feel we didn't have the chance to go through this deep enough. Then you “fight” with your characters to do things that are not obvious for the animators, or they fight with it to do what they think he should be able to do... But the animation team got along with that, and they compensated this a lot to make the characters as amazing as possible !
As for the hairs, and their simulation, well... we entered a world of pain ! Time was flying and we had to find a solution to make the film (sheep on a grassy island, there was no way not to show it, or we would have to redesign everything...) So Lukas Tönne solved this with some awesome tools that made us able to simulate hairs and use Alembic caches for the render farms. It was not ideal for him since it was some quickly made “dirty code”, but it worked great for us ! It did the job ! He really saved us.
How long did production of Cosmos Laundromat First Cycle last and which things did you enjoy the most working on?
I worked almost two years on this. Close to a year on pre-production, so half time writing, thinking, designing, organizing and talking about it ! And one year at the institute to make it.
I really enjoyed the writing part with Esther Wouda, the recording with the actors and the animation dailies. All of this gives you realtime feedback and a good sense of what the movie will be. It's easy to make changes, to talk and see how the characters are or should be. And the technology to work with on those steps is very simple.
For the rest, 3D is so complex that is can kill the joy of making a movie. But having a great team to work with, with very nice people helps a lot to keep having fun and doing great things ! Andy Goralczyk is some kind of magician that can make incredible things even when they're very complex. He took charge of the tornado like a boss ! From the 3D design to the final renders, I'm still amazed by what he did.
Manu Jarvinen was also amazing to work with. His skills and style are very unique. Unfortunately, because the project kept changing during the early phases, his work and talents don't shine in the final movie as much as they could (he did all of the laundromat for example). We wanted him to design a whole world for the movie, but we didn't have the time for that in the end...
Since I'm originally a film editor, I ended up editing the movie, making the animatic and layouts myself. It was surprisingly pleasant with Blender, as long as you don't have video footage and too many audio tracks. It's a very intuitive and fast tool to edit with. I really liked it !
What characterizes the open source production system of a film?
Everything is different !
Ton Roosendaal heads one of the weekly meetings broadcasted online. From left to right: Elysia Brenner, Ton Roosendaal, Lukas Tönne, Antonis Ryakiotakis, Francesco Siddi and Andy Goralczyk
First you have to use only open source software. Luckily now there are a bunch of them that are really good for production but I suspect a few years ago it might not have been that easy. And the big problem is not the tools, it's finding professional artists knowing how to use them. Many of them are used to other commercial packages, and artists are often not happy about learning AGAIN how to work...
Also you have to use only creative commons content or free to use elements because everything has to be shared later on. This is also something most artists aren't really used to either. But this is actually a very good and clear process. It forces you to create and share a lot of assets and tools. So everyone benefits from this. It also helps you as an artist to work more cleanly knowing someone else will be using them later on.
And finally you have to share and show everything you do, as you do it, online. This has its pros and cons. Having realtime feedback on unfinished work from anyone has the potential of being disturbing and ruin the confidence you have on what you do. But it can also be a very good source of information, when you succeed at not taking criticism and insults (or over supportive remarks too) personally. It's a realtime test audience and it's of great value if you can handle it and manage or filter this raw feedback.
There is one thing that bothered me a bit, it's that the open source production pipeline places the tools at the center of everything. Art and filmmaking become secondary. This was the deal on this project, it's defined by those constraints, so there is no problem about it, but it's something I felt fighting against in the long run.
Rigging setup for the cassette player and headphones, by Manu Järvinen and Hjalti Hjálmarsson - one of the many shared assets from the film
Let's take an example: you write a movie that takes place in the North Pole and your D.O.P. wants to use an Alexa (digital camera) to shoot it because it can deliver the look you want. But shooting digital wouldn't be recommended because the temperatures are too low to keep the batteries and camera running, so it might be best to go for film instead. Your art direction might change. You have to discuss this with the D.O.P. : the look of the movie might change. And creatively you have to adapt to this constraint... It open doors for creativity. You have to rethink things and adapt to it.
But if you have a production that works only on digital tools, you just can't shoot the film or do it with very bad conditions. Or you would need to change the project's DNA and re-write the film to take place somewhere else, because there is no artistic compromise anymore. Shooting in the North Pole isn't possible. Art becomes secondary in the decision-making because the constraints offer no real alternative.
When making a movie, technical constraints and limitations are always a problem you have to handle, but what is important I think is to decide on those based on what the movie requires, not the other way around.
So here it was the other way around. And it turns into a very clever way of trying to think outside the box and re-invent things. Also to push the tools further into development. If you take the same example as before, the BI would have to develop a camera that can shoot under the snow way below 0 degrees ! But it would need testing, a lot of extra effort and courage, and a lot of intelligence in the process. So in the end, it's not what a filmmaker is used to, but it's understandable and something one has to value, as it has helped push Blender development significantly and that is one of the main goals of the open movie projects.
How did you feel about the feedback the film has been receiving from the public? What comments did you get when you screened it at Pixar and Dreamworks?
I wasn't at SIGGRAPH and LA when the crew showed the movie at Pixar and Dreamworks so I don't exactly know what happened there. There were some great and kind words from CG professionals from what they told me... Ton told me that big companies and studios congratulated him personally, too. And we heard a lot of compliments on the technical side, especially from people who know how hard it is to pull a movie like that.
As for the online audience and non-professionals, the feedback was very positive too ! For sure our movie isn't the new “Kung Fury”, but it wasn't aiming to be. Of course we would have loved to have such an online response, but the themes, the tone and poetry we tried to have was meant to be in a feature film rhythm and perspective. So it feels short and slow paced for a pilot, mainly because it's not a short film on its own...
But what I like is that most of the people that have watched it find it original, keeping their interest until the end (and for a 10 minutes online short with such a pace it's pretty impressive, I think :) ). And they always want to see more ! Everyone keeps asking me when the rest, or next episode, will be made.
Esther and I are very proud of the tone, the poetry and the feeling that shines through the movie. That is something we were aiming for more than a year ago on paper, and when we saw the final movie together at the NFF, we were very happy to feel this and to see the audience react to it !
The film has succeeded greatly in demonstrating Blender's capabilities as a 3D software. Adoption of Blender has really taken off in the last couple of years, what would you attribute this to?
First to Blender itself. It's a great tool that keeps improving over the years because of the Blender Institute work and Ton's vision for it. He also has a very strong and talented team of developers around him. They fight and work hard to make it as it is.
Secondly, I think the other software are doing a great job at not improving, too ! CG tools are very complex, unstable, slow to work with and to learn, ever since they were created and none of them is really majorly improving or innovating on that. They're just going toward the same direction for years... Big refactors like Blender did for 2.5 are quite rare.
Also the lack of standard formats between CG software is annoying for the professionals that have to work with pipeline and file exchanges every day. Pixar is going open source on that, and some others too, so at least we can be free to use whatever software we want. And of course if it's open source, Blender can use it and improve it ! So that is definitely something that motivates people to use it.
And finally, Blender is free. And that is also something small, medium and even big size studios can't ignore these days...
The Blender Cloud has been created to finance open movie projects and Blender development. Please tell us how it works and what it offers.
Recording characters' voices with actors Pierre Bokma (Franck) and Reinout Scholten van Aschat (Victor)
You subscribe to the Blender Cloud for 10$ per month and you have access to everything the Blender Institute does, their repository for current and past projects, tutorials related to these projects, all the assets of those projects. You also support Blender Development related to those projects and help developing their vision for the future of Blender and the artists using it.
Basically you support the Blender Institute developing tools, assets and tutorials. And it's also a new way of supporting software development, a very transparent and clean way. You really get to see what your money is used for. You can subscribe and unsubscribe anytime you want, no extra fee or whatever commercial bulls****...
It's the Blender Institute way of doing that with Blender, some others are also doing other commercial models to support Blender, help its development and improve its free learning database and assets. But the Blender Intitute is the only one doing open movies or games so far. So they're also investing in art and it's a risk both for them and for the people that support it. But in a world where no major studios are taking any risks anymore, or betting on developing talents, I think it's a very healthy “risk”.
In the long run, I'm pretty sure it's the way to make things that matter and last in the spotlight more than a week or two.
What's in store for the future of Mathieu Auvray?
I hope we can find a way to make the full version of Cosmos Laundromat because I really believe that what Esther wrote is some very good material. We really have a unique and great feature film project for independent CG animation made with Blender. But finding the money for such a project is very difficult because, again, it's a bet and we're in a period of crisis where money has to make money, even before you pay one dollar...
I'm currently developing two TV series for kids based on comic books with Autour de Minuit in France. But I'm also writing and developing some personal projects. One of them I've been working on for years, now. I hope to finance a pilot next year. So I'll keep you posted !
Lightwave 3D
Breakdowns
Plugin News
Software - 3DS Max
Terms and Conditions | Privacy Policy | Cookie Policy
Copyright ©2000-2024 CGPress. All rights reserved.