Making of “The Rose of Turaida”

Apr 15

Making of “The Rose of Turaida”

Mon 13th Apr 2015, by Mike Hepburn | Production

[embedded content]

The Rose of Turaida from Ryan Grobins

Ryan Grobins is a creative force of nature. He is best known for his work on Happy Feet (2006), The Hunger Games (2012) and X-Men: Days of Future Past (2014). On top of his exquisite work on these blockbusters, Ryan is also one part technical whiz, one part handyman, one part visionary artist and most of all – he is all auteur. CGSociety talked to him about his stunning and award winning short film “The Rose of Turaida”.

You are working at Rising Sun Pictures in Australia; can you tell us what you do there? What movies have you and will you work on?

I am a Lead Lighter, contributing to the great work being done in this Adelaide based studio. I spend my days leading a great team of lighters, problem solving renders, tracking down technical issues, and if time permits, maybe actually getting to do some creative lighting. At the moment, RSP has three films it is working on, Pan, Tarzan and Gods of Egypt, the latter of which is the show I am leading the lighting of currently. While at RSP, I have worked on a number of great films, including The Hunger Games and The Wolverine. But I guess the work I am most proud of was leading the lighting of the Quicksilver Pentagon Kitchen sequence in X-Men: Days of Future Past, which has received praise from all over the industry (Oscar nomination, 2 VES awards, and more). It has apparently also become one of the most memorable VFX sequences of 2014.

[embedded content]

VFX Showreel – 08.12.2014 from Ryan Grobins

Can you tell us about the Latvian story “The Rose of Turaida”? Why did you choose this rather grim but ultimately beautiful tale?

I have always wanted to find out more about the country that my grandparents came from, Latvia. So, as the production on “Sneeze Me Away “, my first animated short film, was nearing an end, I decided that my next film would have something to do with Latvia. To find the story I wanted to tell, I delved into the many myths, fairy tales and legends of Latvia, eventually paring down the list of options to two final candidates. One candidate told of the well-known history behind the Latvian ring that many Latvian people wear today (including me). The other candidate was about the tragic tale based on a true story about a young and beautiful woman who sacrifices herself to maintain her honour. In the end, I chose the tragedy, because I wanted to tell a type story that I have never told before. I researched more about the story of The Rose of Turaida, and realised that like much long past history, there were several different versions of the story, and I then had to decide on the version that I felt would be the best one to explore.

What festivals has it appeared in? What prizes has it won?

Rose screened in 76 festivals, with some of the stand out ones being: Animation Block Party, Australian Effects and Animation Festival, Catalina Film Festival and the Oscar affiliated Foyle Film Festival. It has also won 13 awards:

Kingston NY Film Festival – Best Animation

Uncanny Valley Animation Festival – Best Animation

The Snake Alley Festival of Film – Best Animation

Barossa Film Festival – Best Animation

FirstGlance Fall 2013 Short Online Contest – Winner

Savannah International Animation Festival – Best Experimental Animation

Festival Internacional de Animación – Mejor Cortometraje de Animación Experimental

Amsterdam Film Festival – Special Jury Prize, World Cinema Animation

Canada International Film Festival – Rising Star Award

Campfire Film Festival – Best History, Geography, Economics & Business

Australian Effects and Animation Festival – Silver Medal

Los Angeles Movie Awards – Honourable Mention

Santa Monica Independent Film Festival – Honourable Mention

What tools and techniques did you use to create that wonderful sand effect throughout the movie?

In traditional sand animation, the animator would draw some imagery, scatter more sand on top to “reset” the frame, and then draw some more imagery. At first I was thinking of ways to emulate something similar, coming up with examples of a scattering system over existing simulations in much the same way. Until it dawned on me that I should take advantage of the fact that I can do whatever I want with the camera in a way that traditional sand animation never could. The idea of a dynamic camera diving from layer to layer of sand not only would create a fantastic sense of movement. It would also help develop the uniqueness of the style, and solve the problem of finding an easy and simple way of “resetting the imagery” of the particles by simply moving though the layer of sand to another layer beneath, and ignoring the need to “reset” the layer of particles altogether! I then came up with the idea of the film being one long shot, and while this added extra complications, it gave the film a feeling of the story continuously flowing in a narrative stream. 3DS Max was used in conjunction with Krakatoa, which is able to render millions of points very fast.

The dynamic camera had the potential to provide a great “depth of field” look as the camera passes through various layers of particles that come in and out of focus. Quick tests to render all of the defocusing in the 3D camera, proved time consuming and complicated to maintain, in that multiple particle simulations need to exist in a single 3D file, slowing everything down. The solution turned out to be to render out each of the particle simulations separately from a static camera in a separate file, map them onto geometry cards in Nuke. Then I generated a zdepth image sequence from each of the particle simulation’s file that used the position of the camera in relation to where the glass pane was in space. This zdepth image sequence was also used to grade the particle renders, giving the effect that the glass panes appear from a depth fog. Feeding the resulting image sequence into a ZFDefocus node in Nuke gave a beautiful look to the dynamic camera. Generating motion blur from mapped cards in Nuke with a 3D camera was also a lot faster than if I had to do it all in Max. The cards and camera were exported from Max to be used in Nuke. Managing the 3D data then became very easy: I had a “master” 3DS Max file that will contain the animated camera and all of the glass panes, and each glass panes particle simulation was able to exist in its own separate file.

The final camera move with all of glass panes in the film in position.

One of the glass pane files showing off the different groups of particles that will be simulated.

Another glass pane file.

As I was still making this film in the “straight ahead” method, the process to create this animated short film went roughly as follows:

1. I would imagine a key image or two that would show the next particular story point, then in Photoshop, paint up some rough imagery to be used as a placeholder.

2. In a “master” scene in Max, I would play around with the position of cards in 3D space while at the same time experiment with the camera movement for the new part of the film I was focusing on. Quite a lot of tweaking of the camera animation went on here until I was completely happy with the position of the new cards, and how the camera shows them off.

3. I would then export a new version of the camera, open up a default particle template scene in Max that I had previously set up, then experiment with how the imagery should resolve itself in Nuke.

4. Then I paint up an image in Photoshop that looked like a scattering of sand on a lightbox. This was used as a birthing mask for the particle simulation. The final imagery would also be cleaned up in Photoshop at this time too.

5. Next, there were a few rounds of experimentation with the particle simulations, often redoing simulations many times to get the movement I was after.

6. After rendering the particle and zdepth sequences from the static camera files, they would be loaded in Nuke, where I would put together the next group of Nuke nodes that represent the new sequence of particles. I needed to precomp every glass pane, otherwise the Nuke script render time would be completely unmanageable. The precomp graded the particle sequence and applied the depth of field look.

7. Then there may be some effects in Nuke that need to be produced for the new part of the animation, including flares, glints and godrays.

Over the course of production, I became increasingly displeased with the sand scatter images I was painting in Photoshop that were being used as birthing masks of the particle systems. They looked unnatural, amateur and poorly made, not to mention the fact that they were also time consuming to make. I wanted natural sand patterns to photograph, which meant that I would need a lightbox. A quick search for lightboxes to purchase online showed me that they were either too small to be of any use for photographing scatter patterns (which I estimated I would need in excess of at least a metre squared), or that the lightboxes that were big enough were way too expensive. So I decided to make one. What I wanted didn’t have to be too complicated, a simple wooden frame, a custom cut piece of translucent white perspex, a panel of lighting and some masking tape was all I needed really. I nailed together the frame, then used nails on the inside of the frame to rest the perspex on, and sealed it all up with the masking tape. I purchased a really handy type of fluorescent light that can be daisy chained so that I only needed a single power source. I went down to one of my local beaches that I remember had the highest quality sand, and grabbed a bag full. As soon as I poured some sand on the lightbox and started to push it around with my fingers, it became extremely apparent that I obviously hadn’t taken into account the static electricity that quickly accumulates by rubbing fingers on a perspex surface full of sand. Within minutes, the sand was jumping all over the place every time I pushed it around and became almost unmanageable. I figured that a piece of glass would do the trick, so I ordered a custom square of glass that had a low iron content (iron tends to turn the glass a slight green colour). The next weekend I picked up the glass from the shop, ripped off the masking tape on the lightbox, laid the glass over the perspex and resealed it all up again. It worked perfectly. I had so much fun playing with this, and I felt a sense of satisfaction that I made something with my hands!

Raw sand plate

Sand can sometimes sparkle and refract light in interesting ways, giving off a range of beautiful tiny flashes. I gave myself the task of having a go at getting these flashes in my film, realising that when defocused, they would create beautiful looking bokeh artifacts. The bokeh effect needed to be dynamic, not only does the size and shape of the bokeh change in relation to how the camera moves, but a grain of sand may refract light for a short part of the camera move only, sometimes fading up and down in brightness (but not size) without even coming into focus; also the colours sometimes change very slightly. I first severely reduced the particle count of the particle simulation to roughly 0.01%, then I coloured the points using a few layers of procedural noises in different hues, then multiplied the result with another procedural noise of pure black which simulated the particles going in and out of brightness randomly. The procedural noises were linked in object space to the position of the camera, so that any colour and brightness change would be driven purely by the speed and position of the camera. This layer was rendered, and then defocused using the same zdepth image sequence the particle system was using, and merged back on top of the “sand” particles.

Adding the bokeh element made me think about what other visual elements I can add to make the look of the particles more beautiful. By pretending if I were magically able to photograph these many layered particle simulations, it came to me to add a shadow effect, so if it was in the real world and one layer was close to another, the background layer would receive a thin shadow from the foreground layer if there was a light attached to the camera. I ended up using the precomps of the particle simulations from Nuke, mapped onto their respective cards in Max, and use a raytraced light linked to the camera to produce moving shadows from the moving light relative to the stationary cards. While this is a relatively simple effect to set up, it gave the animation a fantastic sense of depth as it now felt like the layers were interacting with each other, and not just existing independently in space.

The final look of the bokeh effect

The final major direct effect that I would give the particles was an embossed illumination look. If a light source existed in the frame, certain particle layers would exhibit a brightening on the edges “facing” the light source. I setup controls in Nuke to be able to control the “depth”, brightness and falloff of the illumination effect. This effect I would only use sparingly, just to give the animation an extra dimension here and there.

I developed a number of tools in Max to minimise the amount of manual input I needed to do. For example, import and export tools between Max and Nuke, auto caching set ups, and a very easy to use render layer tool specifically designed for the film.

The film is all told in a single shot, and so the entire film fits inside a single Nuke script. Near the end of production, the Nuke script was taking longer and longer to open and load up. It ended up taking a couple of minutes at the very end. There were even times when I was editing the script directly in a text editor because it was faster to change values by searching for them.

The techniques you created have received praise from peers. Can you tell us about that? Have you used the techniques since in other movies?

The ultimate praise I received in relation to the technique I developed was its inclusion in the handful of animated short films in Siggraph’s Electronic Theatre in 2013, the same year that also included Pixar’s technically brilliant The Blue Umbrella. Consequently it was then replayed in a large number of ACM Siggraph screenings around the world, thus extending the reach to even more of my peers.

Because I am never satisfied with staying still artistically, I have not used the technique I developed in Rose again, which I admit, is possibly a bit of a waste, especially since I spent over a year alone in look development. I have thought about telling another story in the same style though, maybe the other Latvian story I originally had in mind.

What other personal movies have you worked on? What lessons did you learn from those?

My first animated short film, Sneeze Me Away, was a baptism of fire. What I thought was going to be a fairly simple enough 6 minute toon style short turned into a thirteen minute mini episode that took 50% longer than I had originally planned. When one works on an animated short film that takes a year or two to complete, one of the greatest obstacles to overcome is a loss of motivation. I’ve observed that motivation can take on many forms, and in my own personal experiences, I can distil them down to four main types. The reason why I think it is important to differentiate one from another is because motivation, in all its forms, usually does not last for the duration of the production of the film. In fact, it rarely lasts for any real good stretch of time. Therefore it is wise to be able to try and jump from the back of one form onto another:

Type 1: Internal Inspiration – A usual starting point for the genesis of an idea. Internal inspiration must be strong enough to carry oneself all the way from the initial creation of the idea into pre-production, and hopefully, at least into the beginning of production, or else the idea probably wasn’t good or strong enough to begin with. During the course of production, internal inspiration can also often spring unexpectedly from reviewing milestones in one’s own work, like watching a monthly edit, or a great piece of polished character animation. A sense of self admiration of one’s own good work is always welcome.

Type 2: External Inspiration – The next form of motivation comes from without. Watching a movie, looking at a piece of art, noticing how the light falls on the ground while walking on the way to work. This is usually a great instant hit when one’s own inspiration begins to slow down. But like any drug, while that hit can be powerful, external inspiration rarely lasts for long. Watching a great film on a Friday night can be the difference between a productive weekend, and a weekend spent trawling Facebook and too many games of Bejewelled.

Type 3: Motivation by Guilt – This most commonly works when you first go out try to tell as many people as possible that you are making a film, a great film, a revolutionary film. A film so good, that when it’s released, critics around the world will be singing your praises! And I mean tell everyone, your family, your friends, your hairdresser, your plumber, the weird old man that you see every day at the bus stand. After a little while, you begin to hear all of these people ask you how your film is going. You might tell the first couple of people that it is going well, even though you are fully aware that you have been playing Bejeweled for 8 nights in a row, and have not worked on your film the entire time. You may even be satisfied thinking that it is ok to be telling this white lie to just a few people, but you know very well that saying so causes a little bit of guilt. But then when more and more people continue to ask how your film is going, this guilt builds up more and more, and starts to turn into something darker. As the 17th day since you have worked on your film rolls around, you realise that it is getting harder and harder to be content with lying to others and yourself that your film is still going well, Hopefully you may start to feel like you are not just letting yourself down, but by some strange connection, you will be letting others down too. Insomuch as if you don’t finish the film and become famous, you will be denying them the eventual opportunity to tell other people that they know someone famous! So now there is a lot more riding on this film, and you realise that you now have to deliver, if only so that your grandmother can boast to her friends at bingo that her grandchild is screening a film in some arty film festival in a far away country where actual foreigners will be watching it. At that point, you pause the game of Bejeweled you are in the middle of playing, and shout out “For Grandmother!” You then start opening up your project files once again. As you then continue working on your film well into the night, only now do you realise that this was the only way to make the guilt that has been building inside you, go away.

Type 4: Motivation by Competition – Then there are times when competitiveness can lead to motivation. I’ll admit that this may be a little controversial, but nonetheless, at least to me, it has been a tool of usefulness. There are certain people in the industry that I measure myself against, a good number of these people I have worked with in the industry, and the rest whom I admire but have never met in person. It is a secret competition which none of them know about. When I hear of one of these people succeeding, for example gaining some sort of promotion, or completing a great short that has won awards, or has worked on a successful feature film, I try to use this news to spur me on to also do great things. Competition is a fascinating force in humans; it can turn the meekest of us to do great things in our lives. Why not try and harness its power to produce great art and film?

[embedded content]

Sneeze Me Away from Ryan Grobins on Vimeo.

What’s the most indispensable tool in your creative arsenal?

Quite simply, I’d class the internet as the best tool I have. I am able to research tips on how to write maxscripts to speed up my workflow, trawl for inspiration in periods of low motivation, look up tutorials to help develop my skills and find all sorts of reference.

Is there a personal project you are currently working on, or are excited about starting, that you can tell us about?

Yes! I have finished pre-production on my next animated short film, Cyan Eyed (http://www.cyan-eyed.com), an awesome steampunk adventure film that I am very excited about. The film is now in the asset build stage, with a lot of cool concepts painted by talented concept artists to work from (some of the amazing concept images can be found on the website). Stylistically, it is completely different to my first two films, this time going for a realistic look that is aiming to have the same quality as one of the cool looking Eldar Scrolls game cinematics. Once the film is finished, I am then going to release it online, bypassing the festival circuit this time, unlike my first two films. I can’t wait to complete it!

At the same time, I am also working on developing my first animated short film, Sneeze Me Away, into a feature film. I have written the treatment, and commissioned a lot of amazing concept art, as well as working on a lot of the details of the world the film is set in. I am hoping to begin pitching this project around soon.

Finally, what would be your #1 advice to other artists?

Give yourself deadlines, print them, and then stick the prints all over in places you look at often in your home. Nothing is more important than finishing what you started, and deadlines, even self-imposed ones, are a great way of making this happen. I also print up every piece of art made during the course of the film and pin them up on several cork boards hung on the walls directly near my computer, so that I am constantly looking at something to do with my film.

Hasta La Motion!