Unlike its traditional counterparts, 3D CGI animation sounds more eco-friendly because the animating process does not involve the use of physical materials. Although 3D CGI Animation process does not gives off a direct detrimental effect to the environment, the prolonged usage of computer is detrimental to the users, also not to mention the E-waste when the computer has burnt out or has gotten obsolete in the long run.
As compared to the mostly organic materials used to make traditional animation, E-waste poses greater threat for human beings as the waste management for E-waste involves burning in order to extract scrap metallic materials. Burning the substance will releases hazardous smoke into the air and pollutes the water which can cause serious health risks to people.
Although working for long hours in front of a computer is inevitable for someone who wants to pursue 3D CGI Animation, recycling assets such as rigs and props can help to cut down the amount of work that animators have to do and at the same time make the use of electricity more efficient. Disney and Pixar have been recycling lots of assets that they have accumulated throughout years of making feature length animated films. For instance, the rig of Squirt from Finding Nemo is reused in Disney's most recent feature film Moana. Pixar has even made this as a gimmick to the audience by calling these reused assets 'easter eggs' and cheekily placed them on the set of their recent films.
We are left with Dan's final spirit paintings and polishing up for this week. I am done with my productions, so I spent the week cleaning up the VFX. I attempted thorough clean ups for the swooshing fire by applying some variations of feathering on different edges of the fire, adding a localised glow and also masking the bottom bit so that it looks transparent.
Result of thorough clean up
However, I got tired of doing such arduous task, and since the result does not even fits in nicely to the animation, I decided to give up on spending too much time on cleaning up and did the simple featherings to all the remaining VFX. Instead of adding transparency, I made glow effects for each of the climax smokes to make it look more interesting and use the Colour Burn blending mode on Premiere Pro. Magnus taught me how to use the Action window on Photoshop which records actions that we did to a frame, and we can just click a button and it will apply the same changes for the other frames. It is definitely a handy tool which cuts down the time spent working on a repetitive tasks with multiple steps.
Glow for the smokes
I made some quick calligraphy writings for the title of our animation. I did not have time to experiment with some other fonts, so I decided to make use of the first typeface that comes out of my mind when I think of the word 'spirits'. I think that the typeface should definitely have a high degree of fluidity and lightness because the spirits in our animation is depicted as magic that floats around midair and rises upwards.
Virtual Reality is a revolutionary technology being developed that offers a more interactive way of telling stories. It is a platform of creating new and convincing content for the gaming industry or perhaps a new form of immersive entertainment.
My first experience of VR is when I was in Manchester Animation Festival, and I think that I can see the potential of VR since it is an interesting medium that focus on immersion into the entertainment different to what the current gaming technology offers. I played BlueZoo's HooDoo simulation which is made based on the idea pitched by Catherine Salkeld, who is a junior storyboard artist at that time. The audience follows the story of a tourist cat named Claw embarking on a journey to Tibet. It is set in a Hoodoo as the title suggests.
As I played Hoo Doo, I noticed that there are more things to be considered in making VR entertainment as compared to making film animation, especially when it concerns with the choreography for a 360 degrees environment. During the BlueZoo masterclass, they identified the challenges for the animators to think about sectioning the stage in a theatrical manner and also the cinematography from the audience's point of view simultaneously. Having things animated all over the place will confuse the viewers, so the world should be designed to attract the audience's senses so that they won't miss the important segment of the narrative.
Throughout the years of producing CGI animation, I am always intrigued by Disney's stylised CGI animation and its effect on preserving the element of surrealism in a realistic-looking setting. Joseph Gilland, in his book 'Elemental Magic', described that Disney is fully aware of the importance of observation of subject matters in order to imbue soul and integrity in works of art, and they encourage their artists to go for outings to observe and understand the physics of their surroundings.
In the recent Disney feature film, Moana, the studio take visual effects on to another level. The ocean were made as a playful character, instead of just giving it a passive role in the mis-en-scene.
Having a handful of scientists who are passionate about animation to work in the company to develop new software for water simulation, they have successfully made a rig for the water, and run particle simulation as a secondary motion, to enable it to interact with Moana. They make use of the versatility of the CGI medium to engage the audience with such element of surprise has reminded me, as an animator, that if I keep a forward-looking attitude in creative problem solving while having a solid idea of the briefs that I am working on, the possibility of new discoveries is endless. This is what I like about the discipline of Animation because it is not just about making pretty things, but also involves coming up with new innovations that will enhance the storytelling.
When I was in Manchester for the Manchester Animation Festival, I drew a lot of birds for no particular reason. I particularly liked this pigeon character I circled in red so I decided to make a rig based on him.
I stylised of the pigeon that I made earlier by playing around with shapes because I want to make a good, presentable vector asset for the study task. I choose to do a 3/4 front view of the character because some of the good DUIK-rigged characters are made in such manner. It is more functional as all the features of the character are visible.
I made an asset on Adobe Illustrator, and separate them into different layers so that I can rig it on After Effects.
Stoned Pigeon Asset
For the background, I found a stock image online, copied it two times and polish it on Photoshop as the animation is going to be looped.
Background
Rigging steps and considerations:
Do manual rig rather than auto rig because you'll have more freedom with the naming conventions.
Start pinning the puppet pin from the bottom of the parts you're rigging because DUIK rigs works in inverse kinematic, also the control must be placed at the tip of the limbs instead of the joints.
Create all the bones on the puppet pins accordingly, from the top list to the bottom, and parent each of them to the bone layer above them.
When all the bones are set and parented, create controls on the lowest puppet pin at each of the limbs. i.e: the hands for the arms, and the feet for the legs. As for the body, put controls on the pelvis, and put another individual control for the head if you want it to bobs as the character moves.
To create the IK: organise the layers so that each of the controls is on top the bone layers of each of the segments. Afterwards, select the control and the bone layers in descending order and press the IK, choose 2-layers for most of the case, and DUIK will simply create it for you.
If the limbs twisted when the IK is applied, there might be something wrong with the bones. In case that happens, save a backup project before touching DUIK, and a new separate one once you've placed the bones, so that when you failed you don't have to start afresh.
I don't really bother to polish up the animation as the transitions between loops are still jagged because this is just a simple study task, and I am quite satisfied with the outcome as I managed to put up an audio for the animation.
I am an expressive person, but I am definitely not a natural born actor, so please bear with this bad acting. I don't feel bad the slightest because of this hideous acting since animators at big studios are not so much of a good actors too, but they still does reference video of themselves acting for a scene.
Exaggeration will help to accentuate the movements of animated characters. There should be squashing and stretching in the poses.
In this exercise, I considered the placement of directional lighting to convey the mood of the character even better. The more high energy poses, the directional light source is placed above the character, while for low energy poses, it is placed below. As for fear, it shines directly to the character's face to add tension.
For high energy poses, the core of the body should bend outwards and all the limbs are stretched outwards as far as possible.
For low energy poses, such as tiredness, the core of the body squashes, and so does all the limbs such that the character looks like he is curling up.
The last two poses, I've chosen more difficult poses like envy and fear. I Think they are more difficult to act because the distribution of energy is not as clear cut as high and low energy poses. Envy and fear is a reaction to something external, so first and most accentuated reaction comes from the facial expression, and the body will follow subtly.
I found 'Symphony of Two Minds' while browsing on CartoonBrew. It is an expository of the identity crisis of CGI Animation caused by its versatility as a medium. It was initially clear that CGI Animation is made to improve entertainment experience by engaging the audience with photorealistic computer generated imagery until there is a distinctive dip in the relationship between human-likeness and emotional response, that coined the term 'the uncanny valley'.
A new way of using CGI to creating organic-looking animation has gained attention among artists ever since Disney's Paperman was made. Paperman pioneers the development of the non-photorealistic rendering technology in CGI Animation. The availability of such technology allows artists to accentuate hand-made quality, and hence adds personality to their animation as mentioned by Patrick Osborne in his interview on 'Feast' by CartoonBrew. Patrick also said that 'there is no reason for CG to look in a particular way' suggesting the potential of this versatile medium, and softwares are just tools to achieve a goal, which I think is the key to coherence between the look and the narrative.
The Boy who Learned to Fly is a good example of stylised CGI animation which accentuate the narrative with textures, bold colours and lighting. The design-centric look makes the animation stands out from the basic smooth, synthetic look of CGI animation.
The potential of CGI as a tool to make hybrid animation is endless as more cutting edge technology are being invented. A gadget called StyLit is being developed to preserve the expressiveness of hand-crafted artwork. It works based on the light propagation in the scene which gives the artist freedom to use their own choice of colour to emphasise the lighting effect, instead of using the true colours in the scene.
Although the 2D-3D hybrid approach gives a new and more acceptable aesthetic than the uncanny valley to the CGI animation through incorporating some of the organic elements inherent in traditional hand-drawn animation, I still think technology could not replicate the unique craft of traditional hand-drawn animation. If the sole purpose of the hybrid approach is to replicate traditional animation, this is just another hypocrisy, just like when a vegan eats Quorn's meatless meatballs.
Before doing the actual painting, there are some pre-production planning to ensure that the production process will run smoothly.
This is one of the planning excerpts that I sticked in my sketchbook. There are some others that I did on paper in the studio, but have not managed to organised them yet. Planning is necessary to get the right timing for the special effects, although it could be changed while doing the line test. A simple approach I have used to plan out the timing is to draw a horizontal line and make a timeline breakdown on the line. First of all, I wrote down the frame number in which the scene starts and ends. Then, I add more frame numbers in between in which form or directional changes occur.
This method is just a simplified version the animation Dope sheet, which is used by professional practitioners and big animation studios to plan out their animation.
Dope Sheet
Line tests
Some of the line tests I did on Photoshop for the spirit animation, which are printed grayscale afterwards. we used them as the outline for the painting.
Normally, the spirits move in a circular manner while floating upwards.
Spirit 1.1
When Totsi sneezes, she broke the Shaman's concetration, and the spirits dissipate to the surroundings.
Spirit 1.2 (Can't loop this)
Totsi's first try in summoning the spirits.
Spirit 2
Paint on Glass
How we set the set up:
Taping the top of the glass to make a hinge so we can change over the paper we put underneath easily.
Sticking masking tapes at the back of the acrylic glass to mark a corner where the paper should be placed.
Setup
On the first day, we scraped off the paint with this plastic knife, but the next day, we couldn't find the knife, so we rubbed the paint off with tissue. The plastic knife is pretty handy because it can scrape off all the paints without leaving any marks from the glass, while wiping the paint off with tissue leaves some oil marks. Fortunately, the oil marks does not pose a major problem because it does not affect the images captured by the camera.
Scraping off the paint with plastic knife
Turning on the light box accentuate the mark making, hence makes the animation more interesting.
Climax: Smoke and Spirit
Switching off the light box out of curiosity. The textures can be barely seen.
Lightbox switched off
One thing that went wrong in the painting process is that we did not agree on which light to be switched on. As a result, the lighting of the scenes painted by Dan is different to those that I have painted. I just noticed it when I saw the colour difference between our spirit painting. I have brought it up to him, and we settled with not repainting these scenes as we don't have the luxury of time to do so.
Swooshing Fire
Clean Up
The most painstaking process of mixing paint on glass animation with CGI animation is cleaning up. I used Photoshop to clean up the images one-by-one with the Magic Wand and Lasso tool, and once each is done, save the image as PNG.
One thing that I might have to try next week is to feather the edges of the fire and spirits to make a glow effect which helps them blend into the scene.
Normal Fire
Spirit 2.1
Spirit 2.2
References
EFX showreel pencil test from the production of Disney's Hercules
Effective line test: Simple outline and straightforward, not focusing so much on details but the essential information are well-communicated. This is possibly how our line test will look like. The details are added during the paint-on-glass process.
Elemental Magic by Joseph Gilland
A really useful cheat sheet, perhaps the bible of special effects animation, just like the survival kit.
Gilland describes fire as 'luminous gases being released and rising quickly from their former state into the air', but without any external forces being added to a fire, it will be burning as a perfect featureless sphere.
Since we are animating the fire on a fireplace, it is important to consider the violent and reactionary movement of the fire to outside forces, and hence the fire that we draw should have lots of opposing curves, billows and flickers. We could also add more details, such as crackles from the burning woods, for a more believable illusion of life.
Gilland's breakdowns of the direction of fire's movement
Olivier Malric's special effect demo reel which are some pencil test for 2D animated elemental special effects
A more complex pencil test done by professional practitioner who specialises on special effect for 2D animation. The amount details fits in perfectly with the 2D aesthetic. It takes years of work to get to this level, so it is definitely not feasible for us to do our elemental special effects in this fashion. Although Malric's demo reel helps me understand the way smoke particles are drawn so that they seem to move in 3 dimensional space despite being drawn in 2 dimensional medium.
On Tuesday, we had a final crit on for this project. We are grouped together according to the medium of our animation and we discussed each of the projects and their progress so far.
I like the arrangement of the crit because we had a more focused discussion because most of us are familiar with the technical aspect of Maya. Although I did not contribute much to the discussion especially when some rigging terms, such as 'Blend Shapes' and 'Controllers', are mentioned, I am happy with just by having the experience modelling and not screwing up the rigs when animating on Maya for now.
Overall, people in the group mentioned that they like the unique mis-en-scene of our animation, some of them pointed out that it looks like a stop-motion animation. We have overlooked this 'happy accident', but perhaps the rough textures balances out the unnaturally smooth movements in CGI Animation. Therefore, having traditionally painted textures on CGI Animation kind of eliminates the uncanny.
However, there are some reservations about blending the paint-on-glass elemental effects to the CGI mis-en-scene. We are still uncertain of how to improve on this bit because we want to keep the original painted texture as much as possible.
Some suggested changes to the elemental effects:
Add smoke coming out from the fire
make the bottom of the fire darker, perhaps make a feathered mask
The night before this crit, I asked one of my housemate to help me blend in the fire into the scene, and he suggested to reduce the opacity of the fire and add a bit of glow behind the fire with a smooth brush on Photoshop. I tried it out, and it blends in quite well.
Without Glow
With Glow
After the crit, Steve came over to watch our animation and he gave us some feedback regarding the opening scene. Some of them are:
At the start, camera should focus on Totsi peeking into the tent because she is the protagonist in the story.
Shortening the timing of the overlapping action of the Shaman's hair after Totsi dashed into the tent.
There are too much thing going on at the same time when Totsi sneezes, makes the audience confused of where to focus on. He suggested to cut to a close up shot to Totsi's face when she sneezes, and jumps back to the previous shot after.
Actions taken after the feedback
After the crit, we reviewed the feedback together, and discussed the changes to be implemented. We decided to make some changes based on all the feedbacks. However, we had chosen not to have a major shot change for the sneezing scene because we will end up with too much jump cuts at a short period of time, and this will not give allowance for the audience to process what is happening in the scene. Instead, we were thinking of tilting the camera angle, so that it gives more space between the 3 instances happening together.
An unprofessional looking illustration of what I'm trying to say
We distributed the workload, such that Dan did most of the changes from the outcome of our discussion of the feedbacks, and I continued animating the paint-on-glass.
This whole week have been a stepping stone for me as an animation student. I have learnt more than I have ever had in the past months doing this project, hence the long post.
Animating
Before having to read all the rants on the next few paragraphs about my love and hate journey with Maya, I would clarify that, in the end, I enjoyed using Maya for animating and I would like to learn more about its features for other projects, and I am aware of the potential of this software, especially in the developing industry of Virtual Reality entertainment. I don't think I will ended up liking Maya if I were to have any partner other than Dan, who is really passionate about the software, who has done a good job rigging the character, and who has encouraged me to see the good side of Maya.
I started with a dread towards the idea of animating in Maya. I had a horrible experience on my first day of animating as I kept on losing my scene and Maya kept on crashing when I tried to copy some frames. I have lost everything that I produced on that day because of the messy file organisation. I talked to Dan on the next day, expressing my frustration, so we decided to create a naming convention for the scenes and be more careful when replacing and copying files or folders.
Making CGI animation is also a whole different process to what I am comfortable with, hand drawn animation. The process of CGI animation allows the animators to have emotional connections with each of the animated characters they are animating. Unlike in hand drawn animation where you have to redraw the character all over again, CGI animation only uses one modelled character throughout the animating process.
Being emotionally receptive and being able to translate the emotions into sequential movements comes in handy when creating believable illusions of life in character animation. Hence, in my opinion, the challenge I faced is when I have to animate the Shaman character designed by Dan. I find animating him harder because it is not made by me, and more often than not, when I make characters, they would reflect some of my prominent traits, just like making a carricature of myself. This is when shooting reference videos comes in handy.
Good example of the use of video reference in the animation process:
Even the most amateurish is acceptable in reference videos as long as it gives information on the series of characteristic movements . This prompts animators to apply basic understanding of the 12 principles to the acting for the animation.
Me and Dan still have a long way to be a better animation actor. We have to exaggerate more and keep in mind of the follow through movements while acting for as the characters. Our videos is not that useful in that term so we have to rely on my imagination to make exaggerate movements and fluid follow through actions.
Dan acting out the Shaman performing a ritual:
My failed attempt to frown like the disgruntled Totsi:
Getting facial expression animation right is the most challenging part of the animating process. Despite having not much controls to move, the process is more complex because there are a lot of considerations in order to make the animation effective.
Some things to keep in mind:
Squash and stretch adds more sparks of energy to the character's expression, so they don't move like monotonous automatons.
Dynamic sequence of arcs make the movement of the eyebrows more fluid.
Outcome:
After watching the outcome, Matt suggested to zoom in and out the camera to amplify the emotion in the character. It is definitely something that I could not think of since I barely move the camera when making 2D animation.
After adding camera movements:
Post Production Bits and Bobs
Sound recording with Max, who is my voice-acting coach at the same time. We have not finished rendering out the scenes yet, so we recorded the same sound in different lengths. Max voiced out the Shaman, while I am voicing Totsi. Max taught me that it is important to perform necessary gestures in order to get the right energy of the voice.
Max posing like the Shaman and Dan recording the sound
Paint on glass set up made based on research that I did last week. Before animating, the corners of the acrylics needs to be pinned down to the light box so that it won't move during the painting process. Also, lights other than the light box have to be switched off to eliminate shadows.
Camera and acrylic sheet close up
The whole set up
A cheeky picture taken by Dan of screens rendering.
I didn't do much character and narrative this week because we went to Manchester Animation Festival, but I read some chapters of this book by Keith Osborn.
The book emphasise a lot in exaggeration, mainly to avoid the character moving too realistically, but at the same time make them entertaining. Other than that, a part that intrigued me the most is polishing the animation because last time I only did the stepped tangent animation for Moom's walk cycle. I learn that I need to devote more time to refine the motion in the animation. Polishing takes the same amount of time as each of the process of planning, pose-test and creating breakdowns in order to achieve the quality of animation in feature films. Hence, it roughly gives me an idea of how much time I should spend on polishing up.
A conventional way of refining movement is to make stepped tangent to curve on the graph editor. Osborn identified a major problem of this method: curve overshoot past the value between the keys being set.
Of course, there are many ways to approach same problem as all that matters is what it looks like on camera view. Personally I prefer another method suggested by Ken Duncan to get the animation right from the start by starting off sculpting each frame in the curves from the core of the character, the hips, and move parented controls accordingly to exaggerate movements. This sounds less mathematical, and more straightforward process which also allows room for creativity and tendency to create happy accidents which makes the animation more natural.
Some useful tips to fix recurring problems:
To prevent lagging, delete all the keyframe data on a channel if there is no animation on it: Edit > Delete All by Type > Static Channels
Euler Filter under Animation Graph Editor: fix crazy rotations (gimbal lock problem) between keys if not working do the pose again from start
Other than that, I also did a little research for the research blogpost and to get the paint on glass animation for elemental effects in the animation. I worked out a simple light box set up for the paint on glass animation and wrote down a bit of planning.
I am still unsure whether this method will work or not, but hopefully, editing the animated outcome on Photoshop can help the elemental effect blend in to the 3D mis-en-scene. I talked to Dan about the issue, and we decided to have two point light source on the fireplace, so that there will be glow and shadow of the characters cast on the scene and probably help the painted animation to blend in.
A direct-under-the-camera method of animating such that the animator draw on top on a glass or a perspex sheet, usually in A4 size, with backlighting. There are lots of method out there to setup the back lighting which gives different effects. The simplest set up is to paint on top of a light box, and this is the most feasible technique for my Character and Narrative project because it eliminates shadows, which makes it less time consuming to set the setup.
Simple Paint on Glass Animation Setup with Light Box
The more complex, industry-standard the Multiplane Downshooter technique is good for a single-medium paint on glass animation because it would cast shadows of the top plane to the planes below. However, when 3D and 2D animation are combined together the shadows from the lighting could be problematic as it will not be consistent with the lighting on the 3D animation. A hypothetical solution will be placing a point light source on the fireplace, so that there will be glow and shadow cast on the scene.
The Multiplane Downshooter technique
The paint on glass technique is suitable for straight ahead animation, and good for animating continuously shape-transforming objects. Although time crunch does not allow to make the animation comparable to the quality of oil paintings, the mark makings produced while manipulating the mis-en-scene high contrast characters and objects is one of the happy accidents that contributes to the unique aesthetic of the technique. In addition, having an image that lingers for a few frames give rise to the peculiar features of rhythm in animations done in this technique allows the audience to study the feature of each images. Tones of the paintings could be varied by spreading the paint thickly and thinly by using brushes or sticks. The challege of this medium is to plan everything out and do all the necessary experimentation before doing the real thing because there is no artwork to go back to if something does not work. Nevertheless, this challenge makes paint on glass a versatile medium that allows tactile manipulation and happy accidents which introduce new possibilities to the outcome. It is a process heavily weighted on creative thinking, problem solving and exploration of new possibilities using technical knowledge.
The video below is an example of paint on glass animations done by Em Cooper:
This whole week, our goal is to finish up painting and applying the textures on the props and characters. We have finished everything by the end of the week, and I am pleased with the textured outcome. The details of the textures are retained pretty well even after being scanned, which eliminates one of our major concerns regarding the project. Dan rendered this scene with the Arnold renderer so that the final scene will have good quality of lighting.
Textured Scene
We print out the UVs on A3 papers and used acrylic paint to paint all textures on A3 watercolour papers. We used light box to make the outlines of the UVs visible when painting on the watercolour paper, and also to avoid having the polygonal edges in the final texture.
I did most of the paintings with the exception of the tent walls, rattle 3, mask, Shaman's head and Shaman's robe which are painted by Dan. I also modelled the wind chimes, and learn how to unwrap the UVs. Unwrapping seems like a complicated process when Dan showed me how to do it. After showing me several times on how to unwrap some parts of the wind chimes, he lets me try to unwrap the UVs and watch me do it. This is the perks of having Dan in the team, he is really patient and hence a good teacher. Also, if I messed up he will come to the rescue. Despite all that, I don't want to take this for granted because I want to fix the problems by myself as much as I can so that I'll take some learning points out of it.
Steps to unfold UVs: (No screenshots because I forgot to do so)
Open the UV Editor
Do the UV mapping of the object
Use UV mode, select a vertex of part of the object to be unfolded, press command + click, choose ' To Shell' to select the whole part of the object and move it onto an empty space
Select the edges which have to be cut so that the object will unfold nicely. Think of it like deconstructing 3D shape into 2D.
Next, go to Polygon > Cut UV Edges
After that, go back to UV mode, select one of the vertices, command + click on the UV Editor, choose 'To Shell' to select the whole object, and go to Polygon > Unfold (or shortcut shift+click on the UV Editor, then choose Unfold)
