For our Animated Narratives project we were tasked to create a 30 second animation that involved the theme of “Escape”. I was in a group with Andrea, Ellen, Aaron and Kaitlyn, and although my doubts got the better of me and there were times I shied away from the work, looking back, I’m glad I got to work with this group again. We all worked together in the first semester and always found ourselves working alongside one another again, and I’m glad to say I couldn’t have found a nicer and more comfortable team to work alongside.
Author: Nicolle Watters
Animated Narratives Portfolio.
My work and contribution towards the “Get a Grip” short!
Character Creation Process.
I was stumped on what to do for a creature design so I looked up creature’s online to pull inspiration from.
I felt out of my comfort zone with animal characters but I still wanted to give it a go.
I learned that we could also design humanoid characters, which I felt a bit more inspired by. I made a list of creatures that I would potentially like to make/was interested in alongside scanning for anything of visual interest.
I enjoyed modelling Cthulu for a previous assignment and fancied creating a character with wings. I wanted to create a fae like creature with typical characteristics of a harpy, and with Henry’s lecture about creating creatures in mind I wrote down some notes on how they live/reasons for physical characteristics.
I got inspired by the bottom right design and painted over a colour palette idea. I liked the thought of a fairy who looks after the environment and in turn the environment looks after them. There is a DnD race called Eladrin whose appearance changes with the seasons, so I thought they were so in tune with nature that their physical attributes shifts with the seasons.
I had also created more colour palettes based off this concept. I liked the first design the best as they’re complimentary colours that offer the design a nice contrast.
If I were to go back I would explore a few more concepts as I settled very early on my fae design.
I began the block out stage of my character, mostly being formed out of squares and sphere’s. The ears and wings were made out of plains that I cut to the desired shape and solidified. I made the antennas bevier curves that I solidified and turned into a mesh.
I showed my progress to Henry and he gave me some notes after I progressed. I realized that the wings weren’t proportionate to the fae’s body and made any other adjustments before advancing.
I was originally going to craft leaves out of the one mesh with sculpt but to get better control I made leaves out of a plane, edgelooped and shaped a single leaf. I used proportional editing to get some variation to block out the hair shape. I joined the body and smoothed the form, adding sphere’s for the eyes.
In realized the hair I drew wasn’t applicable in a 3D space so I had to change the hairstyle on the fly, however I am happy with how it turned out.
I then followed Henry’s guide to retopology, which made the concept less daunting. It made sense to plan out loops where you were later connect them. I used colour planes to keep myself right. I gathered as much reference for the retop stage as I could as the wing/should connection confused me, alongside female forms to see where I could connect faces. I used the mirror modifier so the geometry stayed symmetrical.
This video helped with the types of topology and how to break down polygons.
I intended to keep everything bar the body sculpt hard surface, the torso topology needed to make sense as I ended up selecting the torso’s faces, mirroring the other half, scaling and altered it to form the overdress. I created the under-dress by duplicating the garment, deleting top faces and reshaping the bottom.
The wings were difficult to do as I had to keep adding loops to adjust to the chest area.
I copied and pasted the leaves from the hair to the collar, resizing them to drape around the shoulders. I copied and pasted the long cape and slid it under the leaves as overhang, the ties are leaves that I altered.
I moved onto UVing my model, which was an happy enough process as I took time to ensure the topology was ok. Despite this there was a few strange pathways when selecting faces, however nothing damaging.
I decided to use UDIM’s as there were a lot of parts to my model and I didn’t want to lose out of resolution. I wanted to attempt hand painting my model as I liked the stylised effect so I distributed the UV’s across 4 tiles at a high res.
Once UV’ed I joined everything and started on the sculpt. I made a leaf alpha in Procreate and imported it to blender as a brush. I subdivided the mesh to 5 levels so I got a good amount of detail and I placed the leafs around the hairline to blend the leaf meshes into the fae’s face.
Giving face and leaves definition, leg detail:
I baked my normals and uploaded them to substance to paint. I had to stop the project due to personal matters but I finished my textures but freaked out by how many there were for UDIM’s.
——-
Getting back into this project was very jarring as I’m a bit more practiced with blender and substance at this stage. I had lost my original backup file from where I left off so I attempted to export my mesh from my old texture pass for the model but it didn’t carry over my UV’s, which meant I would’ve needed to redo this and a lot of progress I originally made. I found a better up to date file on my onedrive thankfully and was able to continue on from there.
As the model was old I wanted to check over and make a few adjustments, the main ones being utilization the UV islands for better resolution (when I painted over under the leaves it would bleed onto the upper side, despite wanting to save island space for surfaces being shown this bugged me enough to want to fix it) and I wanted to double check for any bad geometry. I selected by trait/faces and filtered through the amount of faces on my model.
I fixed these up by merging vertices and creating edge loops that I linked together. I’m glad I resolved this as I didn’t want any issues when importing into Unreal. After quading my model I deleted some inner faces of the leaves that couldn’t be seen.
I had altered the OG UV’s, mostly for the leaves as the low res would cause bleeding, and island packed them. After I finalized the UV’s I baked my multires as a normal and my ambient occlusion, however when I plugged the baked normals back in it had a strange pixelation about the leaves edges, alongside the AO having a weird result.
I baked these a few times and couldn’t think of a reason why this kept happening so I ended up bringing bring the PNG’s into procreate to paint over problematic areas/smooth out surfaces. I used the mesh as a reference to where the problems persisted, located them on the image file and colour picking to blend.
Before:
After:
The same wrinkled issue happened with my AO so I smoothed these over and painted with B&W. Despite it being a timely process I’m glad I was able to manually fix my image textures before importing them into substance so the problems wouldn’t persist, this just meant I had to keep sending the images back and forth until no blemishes could be seen on the mesh when I plugged in the nodes.
I had also went back to revisit the textures. I looked over the class content as a refresher, and knowing how to import to unreal made this task not as scary.
I baked my normals and AO and ticked low poly to high poly mesh. I then painted on my textures. I was using one of my older files as a base but mostly used black masks and different opacity’s for hues. Having the materials separate was very handy when painting, less so when exporting. There’s probably a more efficient way to export these textures, the easiest just painting on the one mat, but I had a lot of control over painting the different areas of my mesh. I experimented a bit with height maps and noise to try and get her skin to look like moss, but it took away a lot of normal detail so I thought to keep her sorely hand painted.
I exported these textures, one packed for Unreal and the other to bring into blender to see if the mats worked. Which they did! So I stated-
Weight Painting:
Beginning the weight painting process, I created an armature for my mesh based on how I wanted my object to be posed. I wanted to test out the automatic weights feature after parenting the armature to the mesh, the wings and legs worked well so I kept these as they were. As for the rest of the mesh I went in and altered it quite a bit so it would logically pose. A problem I came across was the legs as it would deform the white part of the dress and clip through the upper fabric, so I had to go in and trial-and- error this quite a bit due to wanting her leg to be raised.
As I was struggling with the flow of the torso I left the thigh bones alone and focused on the other areas, hopefully to solve the problem of the hidden faces that clipped through when bending the mesh over. I ended up deleting a lot of faces inside that model that weren’t seen.
I had to paint the inside of the dress and erase inbetween when it wasn’t successful. I felt a bit ambitious doing clothing on my creature as I wasn’t sure on how to make the fabric bend over the models form. I began to feel really frustrated at this stage as it was really killing my motivation and stunted my workflow quite a bit. I also altered between the normalize function, blocking it in then switching it off to blend out the edges. I had the armature and rig highlighted so when I clicked on a bone it would show me the weights painted.
Because of the amount of moving parts my character had I looked online and saw this:
I would accidentally paint other parts of my mesh and would get stretching, so knowing I could select faces to focus on the painting helped a ton.
After an hour or two of editing the weights and watching a lot of youtube I figured I would use the smudge tool to pull the weight amount then smooth. I positioned the bones in my desired pose rather than going back and forth to know if the bones would deform, this improved my workflow a lot and I eventually was able to fix the torso/leg by influencing larger sections of the mesh. I had strange stretching around the sleeves with the wings so I connected the sleeve to the wing bone and did the same method. If I had any weird stretching I would select the faces then remove the weights, then reassigned them the the bone.
Finished rig!
I hadn’t gotten the chance to export a armature and mesh to unreal, however I was dealt a bit of trouble trying to understand how when there was no pose library. Blender updated their program so poses move to an asset drawer. I saved my pose through the dopesheet, exported the skeletal mesh and then was able to pose my model in unreal.
Unreal Engine:
My narrative was that the fae resided in the forest to look after it, so they abandon burnt down areas as a form of guilt. I wanted my character to stumble across nature reclaiming a burnt wasteland (foliage).
I had originally attempted to model an environment in Unreal last year but was unhappy with the results as they looked rushed, so I revisited this too.
I first began by looking at some references that I wanted my scene to replicate.
I first imported a height map from motionforgepictures. I chose a spot in the map with a distant hill scape and created a small rise for my flower to settle. I deleted any of the surrounding map that wasn’t going to be seen in the render.
I wanted to replicate a burnt area of my forest, for that I came across this tutorial with assets from QuixelBridge. I downloaded the Unreal Plugin which let me have access to these assets as I was struggling to find a burnt forest floor MAT/items to fill the scene.
I used a mixture of Unreal and Quixel textures to paint the landscape. I downloaded some foliage from Quixel too, this was a rough has as to what I had in mind with the landscape. With the nodes set up for painting I began to block out the scene. The surface looked too shiny so I adjusted the roughness value a bit.
I had downloaded a collection of trees from the Unreal marketplace and began placing them in the scene. I wanted variation in the types of trees surrounding the fae, and to match the colours of her design, but placing them singularly cause Unreal to lag. To fix this I imported the trees I wanted to use to the foliage tab which let me place them faster. I discovered a fast way to add hue variation through this video, which showed being able to change the appearance through the global foliage actor.
I dragged these tree’s to the foliage tab and placed them down singular for better control.
With globalfoliage editor I changed the tress hue, added the scenes center-point and added dead trees and grass. The place was devoid for life for a while but looking back I felt like my scene have a lot more character as it’s quite sparse.
I wanted to change the sky as it washed out the scene so I used this tool to enhance my enviroment.
I altered the sun value and changed it’s hue to orange and changed the time of day to dawn. I played around with the light intensity to get a dark atmosphere, using a light point so you can still see my characters face and their point of interest.
Final renders:
Experimental:
Reflection:
Overall I’m grateful I got the chance to see this project through. I like the end result, although looking at it now there is a lot about it I would probably change just because the final renders look the same tone/dull.
I’ve grown a lot more comfortable in my problem solving, every challenge I was able to find some way to rectify it. Despite this I’m quite stubborn when it comes to asking for help as I know if I did I would have been able to cut down on a lot of time and headaches. I will need to come to realize asking for help isn’t a bad thing, and I think I would’ve improve my project if I had asked for more feedback and is something I hope to become better with in the future.
Despite keeping to myself about the issues that arose I really enjoyed the character creation pipeline, as scattered as it was. I learned a lot from the Vertical slice group project and was a lot more comfortable with importing my materials into unreal, and UDIM’s feel less daunting now that I understand how to export between programs! I also have a newfound love for hand painted textures, and if anything I would’ve liked adding stylised objects to my scene opposed to realistic ones for my character could fit the surroundings a bit better.
Things I need to improve on would be my topology and.. weight painting. I hope to get more practise at this over the summer so I can grow my confidence in my 3D abilities and improve the quality of my work. I also need to improve on my engagement and communication growing forward as the more I begin to panic the more silent I become. I’m glad to have been able to work on this project again, going into it with a clearer mind, and hope to expand my skills on blender and unreal with similar pipelines.
Animated Narratives 3: 3D Previs.
PREVIS 1:
In order for us to get proper timing down for our animated sections we were told to create a “3D Previs”, which simply just means visualising a scene before creating it. As I was off this day I had trouble trying to understand how to animate using the cameras and basic models, however through the online videos I was able to create a really rough visualisation based off of the refined storyboards. I used very basic shapes to nail down the framing etc as final designs/models for our characters weren’t finished at this stage and I wanted to animate from a neutral standpoint.
This is the video I used to base my previs around, using very general shapes to get the basic idea across of what the scene will play out like:
Animated Narratives 2: Pipeline Planning/Design.
The following post details my designing and modelling processes involving my characters and props!
INITIAL MODELS & CONCEPTS:
We started to discuss character designs and how to allocate, mostly from discussions and meet ups we had prior to this. We were focusing mostly on using “prey” animals with a soft design.
Continue reading “Animated Narratives 2: Pipeline Planning/Design.”
Animated Narratives 1: Introduction and Initial Planning Stages.
In the very beginning of this course we were tasked with rolling a picture dice in group, and whatever these dice resulted in we had to create a story on. We had gotten pirates, and when it came to developing a short story we used a whiteboard to briefly come up with a narrative based on the pictures we received. This was a nice small ice breaker exercise and introduced us to creative thinking through narrative and working in teams to develop ideas and tell our input. We also gave a brief presentation to the class to discuss our over arching idea, this exercise giving us a glimpse on what the main project would expect of us.
In terms of telling a story we were also shown the format of the “Hero’s Journey”, which could be found in a large selection of fairytales and well known stories, such as Lord of the Rings and Star Wars. The structure always involves a hero who goes on an adventure, learns a lesson, wins a victory with that newfound knowledge, and then returns home transformed. This narrative structure helped us form our narrative for our 30 second animated short. I always enjoyed writing my own original works and ironically never heard of the Hero’s Journey before, so I came away from the lesson with the newfound knowledge and skillset to improve my own narrative telling’s.
Continue reading “Animated Narratives 1: Introduction and Initial Planning Stages.”
Small Group Research Poster: Korean Animation.
For the second half of our World of Animation course we were tasked to get together in a small group to research the history or development of one nations/area’s animation from a specific angle. My group formed up of me, Bailey and Robert based from our interest in the Korean animation topic. As I never really looked into Korean animation I had thought it would be a good opportunity to learn about the works they produced considering I was well informed about the aspects of the Japanese animation industry, it was a good chance to go out of my comfort zone.
Continue reading “Small Group Research Poster: Korean Animation.”
Savannah Swift: Process, Reflection and Analysis.
Researching and Groupwork:
In our World of Animation class we were assigned to groups of 4 to research and formulate our own presentation slides based off the short film “Savannah Swift” by Savannah Swift team. This small team consisted of Pauline Gregoire, Lucie Bonzom, Theo Pierrel Et and Benoit Parias. It was frustrating to not delve further into the small teams background and personal thought processes behind making the film, the most I could find about the background of the short would be it was produced under Taiko Studios in 2018 and was a French award winning short film.
Continue reading “Savannah Swift: Process, Reflection and Analysis.”
3D Digital Literacy, Assignment 2 Refection.
I won’t lie, assignment 2 was met with a whole lot of trial and error. Reflecting back on the assignment I had little to no idea what the steps entailed and I chose an design that needed a lot of delicate care when it came to UV mapping and topology. If I were to go back and do everything over I would ensure I followed along with the steps every week to ensure what I was applying to my models wouldn’t cause issues down the line.
Continue reading “3D Digital Literacy, Assignment 2 Refection.”
UV Mapping Hotel Transylvania Character.
When it came to UV wrapping, I got very nervous as my model had so many moving parts that didn’t fit the mold of the goblin tutorials on blackboard. I typically look for other guides as reference, however staring at my map alongside the knowledge where this is the area, I needed to improve upon from my previous assignments I got nervous. However, I didn’t want to give up from the get-go and I intended to neaten my UV map before pushing forward. I also feel that my UV mapping has always been sporadic in where I mark my seams, so I need to take this into consideration and research UV mapping further.