During the production process, I encountered a lot of difficulties. For example, I tried to use two different process bindings to compare the results, repeatedly tested fabric simulation and rigid body simulation, how to use cel shading material to make the character’s look more natural, designing stylised scenes, and how to light the character in the night scene. I learned more new skills and solved these difficulties through many experiments and explorations.
I also enjoyed designing and building fantasy scenes. I used to model more sci-fi scenes and vehicles, and I encountered a lot of challenges in the process, for example, there were a lot of masks in the temple that needed to be modelled, and it was a challenge to complete several different models quickly. The modelling course at school helped me a lot. It made me realise that there are many ways to achieve results in modelling.
I encountered a lot of difficulties in subplot and action design, George helped me to analyse the subplot, design the character’s action, and give me advice and feedback. Overall, what I learnt in this year is not only the technology, but also the overall knowledge of 3D animation, including the whole process of animation production. I treasure the opportunity to learn this and the application of these skills in many cross-disciplinary collaborations.
Finished two other scenes this week: the temple and the city (the first two in the video)
Temple Reference Pictures Inspired by Hayao Miyazaki to make a mysterious oriental temple hidden in a stone cave. This is where the mysterious monster mask craftsmen live.
In Chinese temples you will see Sculpture of God, each representing a different symbol, the fierce idol at the entrance is the door god, in addition to the common Goddess of Mercy, representing happiness and well-being, schooling, health and peace and so on.
I referenced some Tibetan Buddhist temple designs, including the symmetrical building structure, the columns of the main hall, the place for the god is a workbench full of masks of the god’s envoys, lanterns, red drapery, some saws on the table, carpenters’ tools and so on.
Some carpenter’s tools have taken the place of tributes and idols, and here is the monster’s workbench.
The scene when the little girl walks into the cave and finds the temple: Striking the light on the temple, the red colour draws attention, while the cave and the little girl in the near distance are in cooler tones, and the stones of the cave form a good picture frame for extending the sense of space.
Production process:
Various masks hung from the screen and all four walls:
What the model looks like without materials:
Try a variety of lighting effects: spot lighting to illuminate the character’s face, parallel lighting to illuminate the scene.
I created blender’s sunlight here, which is not reasonably available at night, but it’s appropriate for the temple as a top light, and even though the main character is very gentle, I wanted to reflect the majestic atmosphere of the temple
There is no light or shadow on the character’s face if I only use parallel lighting. So the spotlight needs to follow the character’s movement and be animated together.
Cities and Kongming lanterns:
Here’s another scene， this scene is the end of story. I’m going to adding the town in the distance. and also use warmer tones like the reference image below. I will trying to make it to look more cosy.
At the end of the monster and children together to fly the lanterns, and finally chose to do the modern city, I personally feel that this is closer to reality more immersive. Whether it’s London or Tianjin, it’s all about old historical buildings standing in the middle of the city, and there’s something beautiful about this combination that transcends time and history.
Its lifting principle is to use the principle that the air density inside the lantern is less than the air density outside the lantern. It consists of a paper film, a bamboo and wood skeleton and fuel.
The Kongming lantern is a traditional Chinese lantern that rises into the air using the principle that the density of the air inside the lantern is less than the density of the air outside the lantern. It consists of a paper film, a bamboo and wood skeleton and fuel. When the fuel is ignited, hot air is formed inside the lantern. Since hot air is lighter than cold air, it rises upwards and drives the lantern upwards, eventually floating into the sky.
The rise of the lantern requires suitable airflow, no wind and no rain to ensure the stability of the lantern. In addition, the temperature and density of the hot air inside the lantern need to be properly controlled to ensure the safe and stable lifting of the lantern. If the temperature of the hot air is too high or too low, or the structure of the lantern itself is not solid, it may lead to unstable lifting of the lantern.
This image brought me inspiration, I refer to some classical illustrations, in the plot the main character is alone in the forest at night to put lanterns. This posed a challenge: how to do a night scene? You can find a lot of blender tutorials on how to make a summer sunny day, but very few people make a night scene, on the one hand, it’s hard to show the volume of the model at night because of the insufficient light, on the other hand.
The forest
Scenes Design:
Here the characters first time release the lantern . To illuminate the character’s expression in the dimly lit scene at night, I fill in the light on the front of the character.
Final rending:
it’s difficult to match the colours of the night scene, it doesn’t have the rich props such as the blue sky, green grass, clouds and flowers during the daytime, therefore, in order to make a better night scene, I refer to a lot of illustrations from fairy tales, which often have the protagonist walking alone at night in the forest. picture, the moon and the stars are essential, hazy fog light effects and distant mountains, fireflies become the composition of the night scene of several important elements.
Process:
The cave
the tunnel to the monster’s home, with a huge stone statue of the monster as the entrance.
The entrance to the nuo mask maker’s house is a giant stone statue, and the character runs after the lantern into the mouth of the giant stone statue
Final rending:
The gate of the temple often fits the theme of the temple, Lingyin Temple Qingjing in the valley, its gate looks rustic, there is a pine tree inside the symbolism of solitude, independence and fortitude and serenity. In the anime as a monster craftsman of the temple gate, is a stone mask greatly drop open mouth, inspired by the Chinese gate on the door ring, similar to a kind of door god, the duty is to deter thieves, so it all looks ferocious eyes.
temple is a sentiment that is more than just a place to meditate and pray for blessings. temples contain many exotic treasures and pass down all sorts of mysterious legends. A part of them will tell that the temple is hidden in the forest and can only be discovered by those who are destined to arrive.
First use auto rig pro to bind character body. Physics: creat rigid body simulation, Here’s what it looks like when it’s done. I also tried blender’s fabric effect, comparing the two, the fabric simulation can show more details. So in the end I chose the fabric simulation
Final fabric simulation effe：
The effect of rigid body simulation:
For the first time, I tried rigid body simulation in blender. Rigid body simulation tends to be cartoonish, i.e. by controlling the clothes through the skeleton, you can make some special effects (you can give the clothes posing action k keyframes), and I refer to some Japanese cel-shading animation games which tend to simulate the fabrics with rigid body, and their characters’ movements are more exaggerated and cartoonish. Then I started to try over and over again.
The principle of the rigid body simulation is to control the model (the dress) through these squares, but the density thickness of the squares and their distance from each other have a great influence on the effect of the dress, and at the beginning there will be a lot of effects of the dress taking off or the fabric blowing up.
I first changed the weights where I used a brush to make each weight correspond to a bone, erasing the excess weights made the rigid body effect much better.
On the other hand I modified the number of bone segments and density a couple of times, denser bones and and thinner thicknesses, and again the simulation works much better (through-moulding because the legs haven’t been added to the volumetric collision).
Fabric Simulation:
The rigid body simulation was done, but the results were not as expected. As the character moves the rigid body costume moves very stiffly. I started trying fabric simulations.
The principle of fabric simulation is similar to installing the vertices of a piece of fabric on the character with a small nail. Mastering this principle can be used in various scenes, for example, if you make a cloak, you can fix two vertices on the back of the character, if you make a belt, you can fix a point on the top of the belt, and so on. A skirt should be a circle of points around the waist. I tried this theory on this character and was able to realise it by brushing a weight on the top point of the fabric, so that this point completely controls the top point of the fabric (otherwise the whole fabric would fall straight down once the animation starts).
After making the fabric simulation and finding that the fabric still molds through the character’s body, I also needed to make physical volume collisions for the character’s body.
Volumetric collision just needs to be set in blender, but since the character model is quite large and the stance needs to be very strong, setting it to 1 has no effect at all, I’ve tried it a few times and found that around 3 is relatively stable.
Solving for the fabric:
Fabric solving needs to adjust the movement first, then solve the animation, after that you can see the effect of the fabric, according to the effect of the fabric to adjust, to prevent the emergence of through-moulding, and then solving then rendering.
I animated the character in order to see the effect of the fabric, this used the knowledge from the modelling course and the animation course, I posed for the release of the peacock in order to see the fabric simulation and the final effect.
the decorative elements on the clothes are clouds, butterflies, flowers, cranes, most of which I drew in blender. use ps to draw complex patterns and then transparencies to the clothes.
You can see in the model that there are a lot of line textures, such as an edge line on the dragon’s eye sockets and eyebrows, which you can’t achieve with the cel shading material alone. At first I tried to add edge lines to the colour nodes, so that the whole object would have edges, but what I wanted was the effect of the colour blocks being stitched together, and without the edges the model is blurred and a lot of detail is lost, which is a problem with cel shading. This is a problem for cel shading, because cel shading relies a lot on light effects, especially in my scene, which is almost always at night, and without a strong light source, the cel shading material will only show one colour of the dark side (solid colour).
So I modeled the lines at the turn of the character’s facial structure. For example, the dragon’s eye sockets, the corners of the mouth, the brow bones, the inside of the ears, the marks in the middle of the tongue, the decorative lines on the cheeks, and the indentations of the nose.
The legend of Dragon Nuo mask:
Nuo culture in the legend, the evil dragon is the representative of the evil, ghost legend, and open the mountain Mang will be (Dragon King) and the mountain king and other Nuo God, is specialised in driving away the evil dragon such as the evil evil ghosts and positive God. When they appear, break the horns of the evil dragon, to defeat the demons to ensure the safety of the people’s lives, the essence of Nuo can be prevalent in folklore passed down to the present day. Now more dragon masks are used to pray for blessings and blessings of the open mountain mang will, on behalf of the five grain harvest luck.
In the short film the protagonist is designed as a dragon on the one hand is very much like the dragon, the dragon in the culture is often an auspicious symbol, but also suitable for the emergence of the protagonist as a representative role, the powerful and mysterious qualities and the short film it is kind and full of affinity for the behaviour of the formation of the contrast.
In the Song Dynasty, the technology of weaving Luo can be said to have reached its peak. The Song dynasty Luo’s many names and technological advances made it a famous fabric of the time, known as “Song Luo”.
Nobles and officials of the Song Dynasty usually wore costumes in bright colours, such as red, green and purple, and embroidered designs, such as dragons, phoenixes and cranes, in gold and silver thread. Merchants, craftsmen and farmers of the Song Dynasty also had different costumes. Merchants’ costumes were plainer, usually in black or dark blue; craftsmen’s costumes were more practical and comfortable, usually in linen, cotton and other natural materials; and peasants mainly wore blue lab coats and black trousers, with simpler styles. The interior of Song Dynasty clothing was also very elaborate, with details such as collars, cuffs and belts being handled with great care.
All manual labourers in the Song Dynasty period, tambourers, hawkers, peasants, boatmen, and so on, had a common feature, that is, their tops were short, mostly short jackets or short shirts, not exceeding the knees, or just above the knees.
Mapping：
Nuo opera costumes pattern: Nuo Nuo opera clothes mainly red, on the one hand, the fire, the worship of the sun, part of the reason is that the red symbolises good luck, people think it can dispel bad luck.
Embroidery is an essential part of Nuo opera costumes, common dragons, cranes are more representative, but generally only high-powered roles will appear on these clothes, such as the dragon belongs to the emperor, cranes belong to the emperor’s officials. So here I chose clouds, swallows, butterflies and peony flowers, these are very common people elements, the common people like butterflies and flowers in the fields, clouds and flying swallows are also very popular. Here I draw a good pattern in Photoshop first, and then use the transparent paste printing to the character.
Mapping process, surrounded by reference images.
During production I often compared the two characters together to keep the style consistent.
Cartoon Character Binding: this brings some challenges, the most difficult is the facial binding, cartoon character expression changes need to be made with blendshape, I use faciel finished character facial binding. Then used auto rig pro to do the body rigging and finally merged the two controllers together.
Face rigging
Body rigging
Body Rig：
auto rig pro to do the body rigging：Body binding in blender is similar to the process in maya. First build the skeleton and put the joints in the right places, head, chin, neck, wrists, ankles, and the end of the spine.
Building the skeleton is followed by binding and brushing the weights，With IK on the legs, the character’s feet don’t elongate with movement, preventing the character from turning into something like Disney Rubberman during the animation process.
Weights: auto-generated weights can easily connect fingers together, so some adjustments have to be made to the hands, in addition to this auto weights often connect hair to clothes, this time it is necessary to go from bone level to bone level to check and erase the excess weights.
Face rigging:
How to use faciel：
1.The first step is to place the dots on the character’s face in the positions that determine the controller’s eye sockets, nose, chin, brow bone, orbicularis oris muscle, etc., which is suitable for all kinds of exaggerated cartoon images.
2.Refinement of weights – Correction of expressions (sculpting/skeletal method) – Baking，This generates a semi-finished product of the controller，
Generate an animation of the character’s expression, and next tweak it in blendshape to create smiles, sadness, anger, surprise, and so on. And make a pout, wide open mouth, crooked mouth, and pursed lips. Raise your eyebrows, frown, etc.
3.Irregular binding
how to bind irregular eyelashes? How to change their nature during expression changes? How to bind them to the character’s eyelids?
I learnt that binding is based on a hierarchical relationship. Forming a parent-child set of lashes and faces allows the lashes to follow the face.Lower eye lashes were found to be immobile, at first I thought it was a problem with the binding or because of the asymmetry (which didn’t work even after removing half the mirrors) Finally I tried to increase the weights and found that they weren’t automatically assigned to the upper weights.
Eyeballs: inwardly recessed round pieces were used here. I tried that the eyeballs of the spheres would be incongruous on the character, although that would be easy on the bindings (as with traditional character bindings), though less visually pleasing. So ended up making flaky eyes, with the depressions appearing to be deep eyes. Here brings a new problem: how to bind irregular eyeballs？
In order to make the eyeballs (2D character eyeballs are slices not spheres, based on the principle that the faciel binding is rotating around the object, a new sphere is created as the centre to control the face of the character’s eyes)
4.Since the two controllers are not compatible, I tried binding the expression controller to the body skeleton before making the blendshape, which worked but the face controller maintained a half-finished state, I initially bound the expression controller to the bone controller and found that I couldn’t follow the movement. Then bound the expression controller to the neck skeleton, which produced a lot of reported errors. Then I explored further and found a way that I could keep the face controller and the two controllers work properly as colleagues, binding the face to the head object through the parent-child hierarchy.
This week I finished modelling the character as well as the textures. Here is a three view of the completed character’s head. cel shading materials and realistic materials in parallel, there is also a part of the node production by hand-drawn mapping to join.
Material:
Skin material:
I tried a few different ways to make the skin material, the first is the crayon material, use the crayon tool in blender to draw on the surface of the model, you can achieve a very stylised effect, it can mimic the crayon drawings, watercolour paintings, more fairytale texture, but this material will not change according to the influence of the light, the similar works are the game Hades (more inclined to the hand-painted mapping). The second one is the colour node, I got the inspiration from the League of Legends. It has more realistic materials in it, and one of its features is that it draws shadows in a painterly manner
Material node: change the colour of an object through the relationship between light and dark. Node to divide the material into two parts, light side and dark side, first in the light side connected to the painted face mapping, adjust the gradient of the black and white strips can be changed in the light and dark intersection of the turn of the line, the scope of the shadow.
Three views of the head after finishing the material：
Map detail: The eye material, blush, and skin details were drawn in ps. The eye highlights are separate models using a material that always glows regardless of light, this little round piece is attached to the top of the eyeball model to make the character’s eyes look more intense
Hair design and modelling：
As it was my first time using blender, before modelling the hair I researched how to make hair using blender, the principle is to make a cross section of the hair with a circle and then use curves to turn the cross section into a hair piece. A strip of hair was used to attach it to the model head.
Consider the position of each set of hair first, before pasting the model from the inside out. Straight hair will be much easier than curly hair, I wanted to try a higher challenge, I designed several versions of the hair and finally chose the curly version.
Start with the bangs and do it in clusters symmetrical left and right. Then lay it all over the back of the head. It takes quite some time and patience to adjust to the ideal position in order to prevent the hair from moulding through each other.
Here’s the hair converted from curves to geometry, you can see the mesh, indicating that it can be edited. Some fine tuning can be done here.
Body modelling:
Modelling the body according to the draft, building a rough outline with geometry first, then gradually adjusting the points and lines, adding details with sculpting
Clothing modelling: Relatively simple, just copy and extrude the faces of the body and adjust them.
I tried a cartoonish character design. This charater have interesting facial expressions which providing more space for character performances.
this brings some challenges, the most difficult is the facial binding, cartoon character expression changes need to be made with blendshape, I use faciel finished character facial binding. Then used auto rig pro to do the body binding and finally merged the two controllers together
Design sketch
But first the character design, I drew some very cute drafts in the early stages, and tried out a lot of different ideas here, such as designing the protagonist as an ancient person in Chinese dress, here are a couple of sets of ancient Chinese dresses for her, or a more stylised look with pea-eyes, and also thinking about the head-to-body ratio, and hoping that the protagonist would have both the innocent inquisitive traits of a child, but also be full of courage, and exploring a semi-realistic character I’m exploring a semi-realistic style for the character.
In terms of expressions, I referred to the exaggerated expressions of recent cel shading animations, which bring challenging for the later modelling and expression binding.This raises a number of questions： The character’s shadow outline and colour need to be achieved in which way? How to make the character expression move flexibly? How to make the material of three rendering? How to bind the skeleton? How to make the 3RT character move more naturally?
Modelling：
Start by modelling the head, extruding from a flat surface to the general shape of the skull. However, stylised character modelling requires a few details to be considered to make the result better, here are the tips:
Nose：To make such a small nose you need to use the sculpting mode and pull it out with the smallest brush possible
Mouth：The lips of a character with such a small mouth are easy to mould through, especially when viewed from the side (the character has a large angle of dip from the middle of the person to the chin. Still use edit mode to make the mouth, and then use embossing to paste the lips for better effect.
Eyeballs: inwardly recessed round pieces were used here. I tried that the eyeballs of the spheres would be incongruous on the character, although that would be easy on the bindings (as with traditional character bindings), though less visually pleasing. So ended up making flaky eyes, with the depressions appearing to be deep eyes. Here brings a new problem: how to bind irregular eyeballs？
Eyelashes: The eyelashes of a character cartoon are made up of upper and lower eyelashes, and each part is made up of three three-dimensional pieces. Adjusting the eyelash model so that it perfectly covers the character’s eye socket from all angles of observation has encountered some problems here: how to bind irregular eyelashes? How to change their nature during expression changes? How to bind them to the character’s eyelids?
Teeth Material：The character’s mouth is small and his teeth are in a position where they are difficult to be affected by light. I started out with a triple-touch material, which almost always showed the colour of the dark side (it looked like a slice with no light or dark), so I ended up using blender’s material to adjust a realistic material for the teeth
Nuo opera is a kind of Chinese festival procession where people wear masks and play messengers of the gods，in order to pray that everything goes well in the coming year.
In the Nuo festival, Nuo opera mask plays a very important role. Nuo wind prevalent in the Shang and Zhou periods, in order to obtain a strong ancestral effect in the stalls, presided over the Nuo festival Fang Xiang’s wearing a “gold four eyes” mask. “Fang Xiang’s palm masked bear skin, gold four eyes, the end of the Xuanjie clothes, holding the Go Yang Shield, handsome hundreds of subordinates and when the stalls to Suo room to drive away the epidemic.” Fang Xiang’s into the “ghosts and epidemics, disaster relief Na Ji’s deification of the image, the appearance of the mysterious and fearful. In the stall performance, Nuo opera masks have become an important means of Nuo opera modelling art, but also Nuo opera is the most important, the most typical props. Actors wearing Nuo opera masks is Nuo opera is different from other important features of the theatre.
To further design clothes：
According to the role of wearing a painted mask, commonly known as “face”, divided into a Wei, two net, three students, four Dan, five ugly, six outside, seven sticker Dan, eight small students; its performance is commonly known as “jumping exorcism”, accompanied by gongs and drums on the scene. Representatives of the repertoire are “catching yellow ghosts”, “Zhang Wenxian”, “Xue Rengui East”, “Pang’s female”, “Dragon King’s daughter”, “Liangshan land” and so on.
From 2006 to 2014, Nuo opera (Wuan Nuo opera, Chizhou Nuo opera, Dong Nuo opera, Yuanling Chenzhou Nuo opera, Dejiang Nuo Nuo play Wanzai opening Nuo, Hefeng Nuo opera, Enshi Nuo opera, Renzhuang fan drums Nuo opera, De’an Pangong opera, Meishan Nuo opera, Libo Buyi Nuo Nuo opera Linwu Nuo opera, Qingtan) have been included in the national representative list of intangible cultural heritage projects.
Nuo mask carving techniques require strict rules, square and round fall road, deep religious tradition. From the selection of materials (balsam camphor wood) to the completion of a complete Nuo face, please God, open blanks, facelift, polishing, primer, painting gold, raw beard, light, worship, An abdominal offal, recruiting, and other processes are missing one not less. Finally, with gold, red, yellow, black and white five-colour paint according to the mask form on the paint and paint gold, before finally completed a Nuo mask.
The story is about the friendship of a monster with a scary mask and a lonely child. they meet each other by chance in the forest and become friends. The theme of the story is affirmative action. The child sees the goodness of the monster through his scary mask, their friendship is not limited by race, age, or other conditions.
This week has been a further tweak from last week’s preview video. This week I have reset some of the camera positions, changed the movement of the character movement, changed the speed of the camera movement etc. It made the shots more fluid.
The process
I changed the character movements in Maya and set keyframes at different points in the timeline.
When I adjust the action before, I have to turn on the controller that shows the character, and then turn it off after I have adjusted it so that the controller is not shown when the video is exported.
The camera position is then changed according to the set character movements.
Finally export the video via preview，Right click on the timeline to export the preview video.The inside of the viewfinder is the final export, I was more concerned with the composition inside the viewfinder
There are a few things that could be improved in this week’s video and I have ranked them below.
1. This series of shots is about a young girl who finds a broken lantern and starts to chase it.
The first shot shows the girl with a worried expression, the third shot shows the girl chasing the lantern as fast as she can. The third shot shows the girl chasing the lantern. The middle shot of the girl chasing the lantern has been removed to make the plot more concise and clear.This way the action of the first shot can be linked more coherently to the third action, where the girl is surprised and then immediately runs.
2.The girl follows the lantern into the hole, this shot lingers for an extra two seconds.
3.The girl walks straight into the cave and instead the girl first peeks out from the right side of the camera, looks around carefully and then walks towards the camera and stands in place looking around.
4.Zoom in on this shot and change it to a close up.
5.Change from flat to elevated view
6.This set of shots shows the monster picking up the lantern and turning around ready to place it on the workbench for repair.This shot has been removed here and replaced with a close up of the monster above the waist and then the monster turning around
7.Change the position of the two main characters to be closer to the centre of the picture overall. And turn the front of the characters slightly towards the viewer. Move the girl slightly to the left to pull the two protagonists apart.
