Since the renders take incredibly long it is only a playblast for now, but this is the finished animation. The renders have the nicely added effect of reflective glass, making the tiger visible at all times:
The first render using maya’s default sun and sky settings:
After some tweeting to get a more subtle result. I like the resulting contrast of blue and orange, so I’ll use it for the finished result:
I might, however, increase the brightness still.
The technicalities of this tiger rig are incredibly difficult to work out. Next to several plug-ins I have never used before there is an abundance of controls that I have to find and understand before I’m in a position to pose it out properly. Luckily I have a great reference animation by Krzysztof Boyoko, which reassures me that the rig is far more versatile than I am seeing at the moment. With that in mind it is easier to explore the possibilities and I am already getting more familiar with the controls by imitating some of his poses for training purposes and comparing the results I’m getting with his.
I searched a long time for the appropriate reference for this shot. I used video reference of myself, but I didn’t quite get the “giddiness” as much as I felt necessary. Finally I went back to the Disney princesses. Are archetype GIRLS they provided great reference for my own main character. I mainly relied on Anna from Frozen:
It gave me the idea to work in more skips and the “bouncing” movement of her upper body. Together with the realisation that I needed more rhythm it allowed me to add a lot more dynamic to her performance, which was originally a little stiff.
After seeing The Great Budapest Hotel I did some more research into Wes Anderson’s films and found this lovely analysis of his tendency of centered shots:
It made me realise that while in theory my shot plays on a balanced, centre perspective the middle line doesn’t actually focus on anything significant. Due to both subjects moving there are moments where the idea of central placement gets muddled – I’m not entirely sure yet how important these moments are for the overall idea and success of the shot, but maybe I will have to keep the camera focussed on the centre of the tiger at all or most times.
Another interesting thing to consider is the design of the background in these shots. By looking at his design of the areas outside of the focus point I could maybe understand what is needed for my own set up. However, in the listed examples there is every possible combination of objects: Only the focus point with a lot of empty space on both sides. Two equal objects in perfect symmetry. A slight offset with either two different objects on each side, or only one object on one side and empty space on the other, or even a seemingly random collection of a lot of different objects.
In the end I realised that simplicity and balance would be the best way to go to sell my idea and not distract from the shot. A simple balanced background would enhance the end moment of the tiger jumping across the crouching girl, by not taking any attention away from their placement.
With the Tiger blocked in and a clear To Do-List for the girl’s performance in place I have some time to think about the look of the finished project. I want to achieve that high-end finished feel, just like my previous work. But whereas in SS1 I wanted to achieve the best result with the smallest effort and experiments it was more or less a default to go for the usual “Pixar look” of smooth surfaces in bright colours. This look has a lot of advantages, is easy to do and might well be the solution for my current project, but since I am getting tired of it myself (and heard a lot of criticism of it everywhere in the industry) I want to at least look at different possibilities.
(Low budget productions like THIS for example:
might have actually been a charming and worthy contribution to the animation market if they had managed to stay away from a cheap reproduction of the overused and uninspired smooth look that the big companies developed!)
If you look at the Animation Mentor showreels you find a lot of high quality, finished pieces, that utilise exactly this style. But what is nice about them is actually just the good composition and strong lighting choices:
This blends with my experience of concept art, especially landscape drawings. In this area I have been told several times that the thing that makes or breaks a piece is not the level of detail or the brushes used, but simply a solid composition and good lighting.
Based on this it should be possible to get the set for my animation done in many different ways as long as composition and lighting work to support the shot.
In several other examples of work I have seen the purposeful hard-edged design that, in my opinion, works nicely to break up the usual “slick” look of 3D animation and creates a more interesting style unique to 3D:
Due to the nature of my models I doubt I could go for exactly this style, but I like the idea of rough minimalism and would love to try myself on a “less is more” base.
Instead of changing the geometry, which is difficult to do for my existing models, I want to look at the use of textures. Jon Klassen’s work has been a big inspiration to me regarding textures. Besides being amazing children’s books, his work features rich textures combined with very strong edges and borders, being arranged minimalistic on the page:
Especially after seeing his animation work I am more and more curious about textures and how they could enhance digital work and break up the usual, boring look. Though the shapes here are smooth meshes, the textures make them appear rough and edgy:
In all this I need to keep in mind a finished look that is held together. Though the idea of a drawn background (or at least appearing drawn in its use of textures) appeals to me there is no point if it can’t be tied in with my rigs.
Paperman is an interesting film to study for this style. I was very disappointed when first seeing it, and felt that the film was (next to a terrible story and a cheesy soundtrack) overpraised for “looking like hand drawn” when it was so clearly digitally animated and didn’t feel “flat” enough for my liking. However there are certain elements that are quite interesting. The white highlights that wash out all information where they fall and, in a way, flatten the image down, or the clear outlines of her arms.
The backgrounds also have a very hand drawn feeling to them, mainly due to the textures used:
Another thing to keep in mind is the design of the actual background set, regardless of textures used. I really enjoy Jon Klassen’s minimalistic illustrations, so it is possibly not necessary to render a complete scenic background as detailed as can be.
In a song interlude in Frozen, for example, the backgrounds get very simplified and childish to contrast the “real” environment of the film:
I researched some more animation with that minimalistic approach, but I found that, unlike I thought before, it is not the lack of detail that attracts me to this style. The example “Myosis” (by Emmanuel Asquier-Brassart, Ricky Cometa, Guillaume Dousse, Adrien Gromelle and Thibaud Petitpas) shows that, while simplicity may play a part in its design, it is not about absence of details, as some shots are incredibly complex and intricate. Instead it is about a strong and clear composition that strips away anything unnecessary to avoid clutter or mess.
Colour obviously also plays an important part in the visual language of this film. For my own project I want the foreground of the cage to be brightly coloured and full of contrast, whereas the bakground – behind the glass – is set in a slightly less saturated, possibly almost grey tone. I want to enhance the feeling of the camera / audience being inside the cage, the girl being behind the glass, shut off from experiencing the tiger in its true wild and natural form. Her world is tame and orderly, the tiger’s world is wild and intense.
After analysing the visuals in Myosis I am more determined to play on my chosen camera angle and design a background that channels the focus towards the action, rather than simply filling it like I did before. In this early study of what the final piece could look like, for example, I placed a more or less random array of items in the background, anything that came to mind when thinking of a zoo. I migth ahve to completely revise the scenery or design and place those objects in a way that enhances the composition.
My previous camera angle was chosen for balance, but it was a messy composition and the balance didn’t add to the idea of the shot.
I went through several potential new camera angles before settling for a frontal shot. It’s slightly risky as things tend to look flat, but since most people advice to avoid this angle it might be a good way to make a shot that stands out, assuming I can pull it off well enough.
The front on angle, with the tiger walking once close to the glass and once close to the camera, poses some difficulties in framing all of the action without cutting.
First of all I like the composition of the girl dead centre with the tiger moving from left to right. However I really want to keep the turns as visible as possible since they add dynamic, character, and are good moments to show some creature animation and weight. I will have to add a slight camera pan to keep everything in shot.
Secondly I need to make sure the tiger frames the shot nicely when closer to the camera, while still being fully in shot when closer to the glass. To have a nice framing effect I need to put him on higher ground as he passes the camera, so while the camera is eye level to her it needs to be at foot level for the tiger. The problem here is that the higher ground blocks out the view of the tiger when closer to the glass.
Since the camera is already in motion, however, I can shift the angle slightly to go from a more frontal view to a slightly tilted point, blocking out the floor while keeping everything else in frame.
The remaining problem with all of these shots was that they were distancing the girl character by not only having her furthest away from the camera, but also clearly focussing on the tiger movement wise. (I seem to unconsciously manipulate my audience away from the characters, whether it’s lack of speech or through visual language..). I found that there were two ways of solving this problem and making her more approachable to the audience:
A longer focal length gives a narrower field of view, but a greater magnifying effect. There was a good balance between making sure everything is in shot, while at the same time avoiding the girl becoming too small and distant.
More importantly though, I also opened the scene with a close up of the girl character, to ensure that the attention and empathy is on her, before the camera shifts focus to follow her gaze and give way to what she is focussing on. The audience now shares her experience and has more context to the scene.
Here’s a quick version of what this camera move will look like in the final: https://www.dropbox.com/s/ovwwzrgnrp6vkj3/camera_front_tiger_up_04.avi
I’m still in the progress of finding the exact story for my final project. It will be 15 – 30 seconds long, involve creature animation and human acting and dialogue, or rather, spoken monologue.
Originally I expected it to be nothing but cheesy and silly, with a cheap joke that’s only a vehicle for animation, but I will try to push it to a slightly more solemnly acted ending to give everything a layer of sense behind it.
The currently favoured camera angle is very balanced and mostly chosen to have both characters in clear sight while favouring the tiger and bringing it closer to the audience. It was important to have the camera within the cage and the woman on the other side of the glass. I am still considering a more contrasting or dramatic camera angle and trying to find ways to incorporate light to sell the idea of contrast between the obnoxious zoo visitor and the wild animal.
I have already done some intense research into the movements of tigers and recorded the sound acting. Since the voice will be the centre of attention in this I wanted to get a voice talent in to ensure good acting, but unfortunately those are hard to come by. Even with semi-professionals like I found for my previous project the result wasn’t convincing enough. I tried to record some real reactions in Dudley zoo, but people rarely get loud around animals but watch them quietly and fascinated. In the end I did my own take, and there is a possibility that I will just stick with it.
While recording my own video reference I realised that the recorded words were far too rushed. When actually encountering an animal and watching something there are far longer breaks inbetween the sentences. A person watches the animal, and only a new observation or thought kicks off another sentence.
I first blocked in a performance that was very close to my video reference. I made the mistake of ignoring the tiger and just blocking her out independently. But since her actions are motivated by the tiger they need to be closely related to the timing of the tiger pacing the cage. I redid the blocking and based it on an estimate of the tiger’s speed.
As soon as a camera angle is decided I need to improve the clarity of the poses towards the camera and pay attention to contrast and good silhouettes. At the moment she never really comes out of her default pose, making for a fairly boring shot: