The topology of the animal must facilitate deformation.
Produce a 3D asset of an animal wearing a detachable saddle or saddlebags, placed and lit in a UE5 scene. A level of abstraction or stylisation is permitted and may even be required, particularly when it comes to animals with fur/feathers, (The given budget will not lend itself to creating a project heavily in need of hair cards)
If stylised, the animal should still be close to its natural proportions and should have convincing anatomy
I have chosen to sculpt a racoon! Like the brief suggest, This might mean I will have to find a good balance between stylised and realism so that I can depict the fluffiness of this animal without the need for hair cards.
My personal goal for this project is to successfully texture my assets!!! I will have succeeded once I create believable and visually appealing textures! I also hope to pose my raccoon on a plinth and use Niagara systems and alphas to effectively dress the scene.
To make sure I succeed on the 'Natural Proportions' and 'convincing anatomy' part of the brief, I have collected a range of raccoon references, including different angles, close-ups, expressions with its mouth open and closed, standing and walking, detailed views of the front and back paws, skeletal structures, and a small section exploring potential saddle inspirations.
BLOCKOUT
Week 1 - Thursday
Notes from todays lecture:
Brush > Curve > AccuCurve allows you to push out sharp points. Useful for pointy things like hair strands and scales.
To multiselect using the gizmo tool; Ctrl + Shift + EITHER Drag on negative space to select / deselect everything OR Click on parts of your sculpt to select / deselect them individually
You can easily and quickly inflate the mesh if you hold Ctrl whilst dragging yellow gizmo square
Duplicate mesh using Ctrl + Alt
Whilst on move Gizmo, Click Y for Transpose tool
The Brush 'Move Infinite' is very useful for making drastic changes.
This is my third time using ZSpheres, so I’m familiar with how they work, but it’s the first time I’ve used them to block out anatomical parts. My previous attempts were much simpler - a braid (following a step-by-step tutorial) and an eel (basically just a stick). It was very funny working with the ZSpheres, as you can see on the top left - a real Waterbear! (Tardigrades will just have to give up that nickname).
I didn’t use a ZSphere to block in the head because, as my lecturer mentioned, it’s often better to create the head separately - it tends not to work well when included in the main ZSphere structure because it can be cut off in strange ways. A good example of this is when I was creating the eel in the last block; the ends of the ZSpheres would turn square, which wasn't ideal for a round head.
One thing I learned while using ZSpheres this time is to enable symmetry (pressing X) before building off the main body. Unlike other tools, ZSpheres don’t have a Mirror and Weld option, so symmetry needs to be set up from the start. Or you can just wait until you are happy with one side to make an adaptive skin... but its better to see how the anatomy looks from both sides whilst moving things around.
They are really quite chunky...
It was a bit challenging to make it look like a raccoon - it kept resembling a pig, a bear, or even an opossum. I think this is because the body is a little longer than it should be, and the face needs to be a little flatter too. It will probably start to take shape once I add the fuzz and work on the proportions a bit more.
I’m still undecided about how realistic or stylized I want the anatomy to be. Raccoons are naturally adorable but should I push the design toward baby-like proportions or stick with the look of a more mature, adult raccoon?
I came across a 3D model of a raccoon that really impressed me (Seen above). The artist used hair cards and even animated the raccoon performing various actions, which was incredibly useful to study. However, I noticed something small felt off about this ‘realistic’ raccoon - it looked more like a dog than it should have in some areas, and it lacked the characteristic cuteness that raccoons naturally have. Curious to learn from this, I decided to investigate why this might be the case so I could avoid similar mistakes. One challenge I discovered was finding good reference images of raccoon feet, especially the back paws. This is likely why the artist ended up using dog paws as a reference for the sculpt (having four toes on their model rather than the five they are meant to have). Another issue - apart from the eyes which clearly don’t resemble a raccoon’s - is the chin. A raccoon’s face is much slimmer and more pointed than the one in the sculpt. While working on my own model I realized this might be because the artist sculpted the mouth open, just like I currently am. This likely made it harder for them to notice the differences in the chin’s shape. I believe this is the case because if you hide the bottom jaw on their model, the face already looks much more raccoon-like. Despite these minor issues, this was by far the best 3d racoon I have found which means getting it to look like a cute raccoon - without stylising it too much - will probably be more difficult than I am expecting.
I later found this low poly racoon sculpt which I personally prefer. I think it has to do with it's proportions because it retains the cute racoon qualities and proportions (but the same inconsistencies in the back paws can be seen as this was made by the same artist who created the realistic raccoon I talked about before. I should try to combine the qualities I like from both outcomes to help make my own project.
Racoons look like a cross between a lemur, panda, and fox... it's easy to make it look like a rat but it does NOT look like one.
Here is another example of excellent work but I still feel like the cuteness is missing from it and the face doesn't look exactly like a racoon's either.
I worked on refining proportions in this lesson. As you can see from the images on the left, I shortened the torso so that the ratio between head size and body size is reduced. This is a common way to easily make animals / humans appear younger and cuter and since that is my goal, this worked to achieve that. From the front view, I also brought the front legs closer together so that it is anatomically accurate to how racoons usually and naturally hold themselves whilst walking. Above you can see I was making decent progress with the face too!! Whilst I was happy with how my racoon was starting to look, I decided it would be best to start thinking about what I should do for my saddle. I knew the kind of vibe I wanted for it but I didn't know how that would fabricate into a functional saddle.. so I did a crude blockout of the rider's size and made quick indications of a harness and saddle to spark some ideas for my designs.
To save time on creating the 20 fingers / toes, I just made one cylindar, and duplicated them; plaacing them in the right place and mirror welding the other side so that I would only have to place 5 fingers. When I was duplicating the front toes to bring them to the back feet, I hadn't notice that Ctrl + Alt duplicates the mesh within the same layer. This resulted in a mess of fingers all over the floor when I was adjusting the front / back and didn't notice it had moved the others. I tried to cut them off but I was having issues with TrimRect. This is because TrimRect only cuts geometry off of meshes, it refuses to erase an entire floating mesh. So to delete entire floating meshes, use Ctrl + Shift + Alt and drag over what you want to delete (this will hide the selection). Then go to Modify Topology > Delete Hidden
ANALYSING INDUSTRY EXAMPLES & STYLE DIRECTION
As I start to wonder how I will refine my sculpt, I am still struggling to decided on the level of stylisation and realism I will choose to settle with. To combat this I went out to look at industry examples of creatures in different games to look at their approaches so that I can way out what level of realism is feasible for the animal I have chosen with this budget and skill. Firstly I tried to look for animals that share a simillar build or concept to that of a saddled racoon and I found some in Genshin Impact.
Looking at these first two creatures, Although quite low poly, The texturing and minimalistic use of topology to create the illusion of fur is very effective. Looking at their uses in game to understand why the game artists chose to buid them this way; they are small creatures that are only seen in the open world for very short periods of time because they are meant to be chased: if caught the player is rewarded with low value resources. This explains not only why they use a cheap budget, but also why their designs utilise one bold colour to make spotting them easy, simple shape language to recognise them, and cute repeating animations and sounds to attract player's attention. Although They utilize a smaller budget than my project provides me with, it's good to see how similar creatures are depicted in games and how they are efficient with their budget withought compromising aesthetic quality.
In contrast to the simpler animals I tlked about before, I decided to look at a creature that utilises a higher budget so that I can see what is possible for a fluffier animal that is more 'important' than a sidequest animal. As expected, more detail and budget is given to enemies, bosses, and playable characters (listed in ascending order). I think it is interesting how they make the topology reletivly flat across the entire body but then exagerates the fur in areas that wont deform such as before or after limb joints.
To give the players a memorable experience they also have to make their designs stand out. Since it is a fantasy game, they really emphasise the genre through the use of emissives, more complex colour pallets, and merging creature elements to make them look more magical - in this case, butterfly wings are being used to enhance this enemy.
Despite feeling more informed about how games treat their topology and silhouettes for thier creatures in games, I was still unsatisfied with the conclusions I had made on stylisation. I didn't want to use such a fantasy style like genshin, I wanted something with more detail. This is where I looked into artworks from Dishonoured 2! I absolutely love Dishonored 2's taxidermy animals and how they manage to balance their level of realism and stylisation to give the world a mature look which still manages to bring with it hints of wonder within it's subtle painterly rendering. I aways find hyper realistic games to feel uncanny so I often gravitate towards games which provide some stylisation - no matter how small. I personally think Dishonored 2's artstyle is perfect for an action-adventure game because even just by looking at these taxidermy animals alone makes me want to go out and explore the place these are being kept in, how they got here and where they come from.
In terms of proportions and anatomy, they don't over simplify anatomical parts like genshin does, they instead retain all aspects of their real life references but change their sizing instead - to bring attention certain parts of the animal. For both the bird and the monkey they decided to reduce the eye size and extend / enlarge anything that would give them a more triangular, uninviting appearance. These triangles are seen in the bird's head, beak, chest, wings, and line of action in the way it's standing. For the monkey, triangles can be seen in it's fur, frown, ears, collar, pendant, and wall stand. These eocurring angular shapes help match these assets to the darker world of dishonoured and also helps make you admire these dead creatures for their elegance because of the sharp and delicate connotations triangles have visually. This is very important; keep the anatomy the same but adjust proportional sizing subtly to comunicate the emotion / feeling / narratve you want to tell through them.
Their use of colour is something I also really want to take onboard for my creature project. It's mostly made up of naturally dull colours seen in the real world but then it surprises you with this new saturated accent colour that makes you crave for more. Since it's mostly made up of neutral / dark / dull colours, the vibrant hues get to shine and be all the more impactful. It utilizes darker, dull colours to make the saturated accents all the more impactful. Subtle textures are also what makes this style effective. the textures on the bird's beak and the monkey's skin and hair introduces believable irregularities / realistic noise and texture which balances the more styalised and simplified areas of the assets: like the tail feathers on the bird and the clothing on the mokey.
Not only their animals but the foliage is fantastic inspiration for my plinth. The level of detail they display whilst keeping it intriguing and fantastical is so inspiring. It appears they use a mixture of alphas and assets to create these (though I might be missing something). Here yet again you can see how it achieves high level of reaism but that 10 ~ 20 % of a fantastical touch through vibrant colours and whimsical patterns help to keep this art style out of uncanny valley territory - not that you would feel that towards plants but you understand the point I'm trying to make.
PLINTH DESIGN
After this research, I decided to get back to my own project, bringing with me the things that I learned to plan for my plinth design. I spoke with one of my classmates and they suggested a fun idea: having the raccoon carry a bag of coins which are spilling out, with the plinth being a coin waterfall arcade machine. The idea was that the raccoon had gotten into the machine to steal all the coins. I loved the concept, it’s so comedic and unique. but while it's an idea I’d love to explore, I feel the nature-inspired plinth designs I’ve been considering work better for my project. This is because I want my insect rider to complement the saddled creature, and since we’re required to use insect parts for the photo-bashing textures, I really want to make the most of the beautiful green hues and fresh colours often seen in bugs. With that in mind, I want to focus on a natural theme for my project.
While I’m shifting away from my classmate’s modern setting, I love the idea and am going to incorporate elements of it into my design. Instead of a coin waterfall, I’m reimagining it into a sort of wishing fountain, pond, or stream. I definitely want to include a bag of coins on the saddle, and while it won’t be an arcade machine, I’m still planning to feature coins in the plinth’s water.
BLOG TASK 1
Discussing My Creature
[RIDER AND RACCOON NARRATIVE]
The raccoon and rider are nomadic thieves who travel from place to place. I want to design my ridder and saddle inspired off of Verdigris statues. This appearance will help them to blend in with the environments they often visit and will allow them to stealthily steal the treasures that are left at shrines and wishing wells / fountains etc. an epic patina statue that has come to life. They steal the offerings of spiritual places to return to their own saviour. Think fountains with statues that use it for people to tip coins into their waters. A silent, elegant, and stealthy beggar.
FUR ALPHAS
Week 2 - Friday
In this lecture we were introduced to two new tools: Morph Target and Layers. (in Layers you HAVE to click Record otherwise it won’t record any information onto that layer). These were extreamly helpful to learn about. There have been many times in the past where I have tried to introduce textures onto my sculpt only to end up not liking it so I would then had to smoothen the textures out with Polish which made the sculpt very muddy. Layers are more useful to use later in the process when defining details like strands of hair rather than making big changes because modifying topology like Divide, polishing, ZRemeshing, Mirror & Weld, Dynamesh don't work when Layers are being used so it's better to use subgroups and different meshes to create things like this. As for Morph Target, you must StoreMT and then use the Morph Brush.
We were also taught how to make fur alphas. Once your fur texture has been drawn out on a plane, set your brush to StandardBrush and stroke to DragRect then go to BrushAlpha > GrabDoc (at the bottom of the window).
For this fur alpha on the left, I used the massking laso to select a fur strand shape. I then used the move gizmo to rotate and push out the unmasked area to give it a clean elevation. I then polished the edges.
^ Focal shift from -100 to 50 ^ Focal shift effects the opactiry of the brush's edges. this helps to remove undesired artifacts showing in the corners.
I ended up disliking the repetetive look of the use of fur alphas and i also felt that it didnt capture the fluffyness of a raccoon so I decided to try out other methods to achieve the desired effect. The Slide brush was interesting but it distorts the mesh very drastically. I am using layers to apply hand drawn fur because the alphas felt too predictable and repetitive but I don't know how to retopologise the mesh without having to bake down the layer of fur (Dynamesh and ZRemesher refuses to work when a layer is being used apparently?) I much prefer the Alt version of the ScribeStandard brush because it creates very clean lines quickly which is perfect for stylised fur.
I then tried Inserting spheres to bend into large chunks of fur. This made the fur look too long. I then tried using a large DamnStandard brush to carve and extrude flowing fur lines (middle) but this still didn't look right because it wasn't flowing naturally across the body. In the final image on the right you can see a mixture of my second attempt and my third attempt; I was using small spheres where I used the slide brush to slide fur out of it. This had a much nicer look to it but I still have the rest of the body to figure out.
The stylised fur from this Wayfinder Warbear is a really good example of the fur I would have wanted to create but the retop might be too messy. The top two are my fur attempts and the bottom two are closeups from ..
analysis on this beautiful piece where they have captured the fluffiness of the animals so well.......
Painted to see how it changes the look of proportions once eyes are darker. This looks a little too human like and silly
I then made further changes to the face by adding the brow fur and upturinging the cheek fur and making changes to the eye shape.
I dragged in spheres and used the slide brush to pull out strands. This worked to some extent but when I started to overdo it, it started to look like feathers - especially at the joints. I found that using the slide brush only worked when I utilised Accu Curve.
FUR & PROPORTION / ANATOMY TOUCH-UPS
Another attempt at fur. I like this a little better but the strands look like they're melting due to their bulbous, thick, and clunky shape. They should taper sharply to resemble clumps of hair rather than clay or wax. You can see how I adjusted the proportions to make it look more raccoon-like. In the image on the left, the S-shaped face and prominent brow ridge make it feel too much like a bear. I modified this by tapering the face slightly, giving it a more mouse-like appearance. I also pushed the body in to reduce its length and raised the spine so the highest peak of the back is more centered rather than toward the rear. Additionally, I thinned the lower parts of the legs for a more refined silhouette.
After so so many attempts, I FINALLY found a method I liked for creating believable semi stylised fur!!!! I used a combination of the brushes Clay [CL], Slide [SD], Dam Standard [DS] and Flatten [FA] . To produce the effect, I build up sections using the Clay brush by drawing the fur pattern I want. Since the brush leaves behind some harsh edges where the brush stroke begins and ends, I use the flatten brush to ease the transition from flattened fur to thick and fluffy. In some areas where it requires more details - like protruding fur strands - I use the slide brush on the highest peaks, and follow some strands using DamStandard to carve defined strands.
I was worried it woud be too noisy but I ended up liking the effect. I can always reduce the noise through the use of textures and colours in substance painter.
SADDLE DESIGN
Week 3 - Monday
For the saddle I wanted something basic because the budget isn't too high. I also didn't really give myself time to plan something fun for it so I used my references as the base of it's design. My raccoon is meant to be quite large in comparison to the rider so I made the saddle smaller than it is seen on horses. If I gave myself more time I would have loved to give the saddle a more leafy, sharp look to it just like the examples bellow from 'How to train your dragon' saddles because it would match my rider far better.
BLOG TASK 2
Movement & Animations to Consider
Screen Recording 2025-01-07 190948.mp4
Screen Recording 2025-02-04 104432.mp4
Raccoons have highly dexterous hands, so I should retop the fingers and wrists like a human’s—using gridded loops rather than crocodile loops. Their flexible limbs allow them to move sideways for climbing, so the topology needs to support that range of motion.
For animation, the key areas to optimize are the tail, nose, and ears. The ears and nose only twitch slightly, but they still need clean loops around them for proper deformation. The tail moves side to side but doesn’t curl upward much, so I’ll account for that in the topology. When posing my raccoon, I want to push both the head and tail down, as that’s its most natural posture when walking.
RETOPPING
Week 4
Started retopping hlf of the body because it seemed like it had the most straightforward approach. For some reason it showed me this horrific looking issue on the boarder so I simply just deleted those extra faces.
I was wondering if the ear was bring retopped too simply...
I left out the spaces for ears and joints so that I could fill them in after I export them out of ZBrush. I should have done this before I began to retop to save time.
From grided to loop when I added the tail... note to self.... add all HP decimated assets of the character so I don't waste time creating a neat grid to just end up removing it to create a different edgeflow.
Notes : Wrists don't need crocodile loops because it is more human like meaning it folds both ways. The leg loops should be placed a little higher, the ears don't need too much complex topology if the animations doesn't require it.
When retopping the toes, I started at the tips and worked my way up. The most challenging part was shaping the knuckles while reducing the poly count for the ankles and the rest of the body.
This is the final retopology for the raccoon.
Reflections: I should have kept symmetrical areas separate from asymmetrical ones—for example, keeping the head and body independent from the tail would have been much better if done from the start.
What went well: Naming my layers was incredibly useful, especially with seven sets of teeth. Without proper labeling, managing them would have been a nightmare. Also, creating loops before gridding them proved to be a better approach than starting with the easier areas.
UNWRAPPING & PACKING
Week 5
I started unwrapping the mouth first since it can be symmetrized and easily cut up. I scaled it up to ensure the grid pattern and size were clearly visible.
These are the nails being unwrapped while checking for consistency and ensuring the checkered pattern is evenly distributed.
I felt the torso took up too much space on the UV sheet, so I cut it in half and used the symmetry modifier to optimize space. The tail requires more room since it's asymmetrical, but I think this trade-off works well. The head and tail get unique UV space, while the symmetrical torso allows for efficient packing - especially since the saddle will add the desired asymmetrical detail.
Issue: Using symmetry for the head created extra vertices. As expected, the UV unwrap struggled with the new topology, so I had to go back and fix it.
I had 7K polys remaining from my 23K budget (16K used). I felt the silhouette could be improved, especially with the sharp fur. I tested this, as seen in the center image, but it was too excessive for minimal silhouette change. In the end, I only kept the most impactful adjustments—those that made a difference when viewed from a distance.
The final packed UVs.
1) Eyeballs 2) Body 3) Head
I chose to pack the eyeballs seperated like this because in the past I have always needed to go back and seperate them so that the painted stylised highlight wasn't mirrored strangely. If I don't end up making them unique eyeballs I should go back and merge them together to be more efficient with my UV space.
I quickly created the topology for my saddle by ZRemeshing to a low subdivision then removing final edge loops in Max.
I apparently had two of the same mesh ontop of eachother making the tri count for far more.
1,833 Tris Over budget and I haven't been able to add everything I wanted...
Saddle packed unwrap
Why is my plinth invisible like this? and how do I turn off the vertices?
My lecturer didn't know why or how my plinth was acting strange
To turn off vertex ticks go to tv looking Display icon and untick Vertex Ticks.
Plinth currently at 633 Tris. Leaving 4,000 tris left for Alphas. I'm still unsure why the plinth is invisible...
I FIGURED IT OUT!!! it's because in Display Properties, 'Display as Box' was ticked!!
1 x 2048 texture budget base & 1 x 512 for Alphas. Here I repacked the stones to make space for the plinth, taking Texel density into consideration.
These are the settings for when exporting 3ds Max settings to put it into Substance Painter.
Remember to make all parts of the asset you are exporting from ZBrush into one subgroup otherwise every subgroup is imported as a separate layer
I tried to unwrap and pack the rocks as quickly as possible, expecting them to be simple, but they turned out to be surprisingly stubborn. I focused on hiding seams in the back and where the rocks touch. On the right, the gridded pattern helps check if the textures will lay cleanly. Since they’re rocks, any irregularities or misaligned grids can easily be passed off as natural erosion, so it’s not a major issue if the grid doesn't line up perfectly.
I came across this really awesome texture artist who textured this fantastical bird. I liked how they used colour to create these aesthetic gradients on the feathers.
The silhouette of the low poly looks great. It looks a little odd withought the teeth and eyeballs though...
So that the uvs dont get seperated, tick UV tile so that you don't have to keep jumping back and forth between layers just to paint across a head and a body.
The only problem is, The baking icon isn't showing on my version of Substance painter.
Why isn't the icon showing up? How do I get it to show? Apparently it was because my version of Substance was older.
I had very very few baking errors. Only the interior of the mouth had issues. Th nails on it's paws also had issues but these only showed up once I bumped up the antialiasing. I didn't want to have to paint these out manually so I just exploded these instead - which worked very well!
For exploded bakes, duplicate the 3ds Max file so that the original placement of the assets can be accessed in their original places (just in case)
_EB = Exploded bake
To rearrange the assets back to normal after baking everything in it's exploded state, go to Edit > Project Config. The only issues I spotted once I brought everything back again was very minor errors in the gaps of the teeth, and on the back toes and nails. Since everything looks great from the front I am not too worried about these issues. The nails are going to be quite dark anyways. If I find that the baking error is too visible after painting I will fix it through photoshop.
Texturing
The symmetry wasn't centered on the racoons face so I was going to edit it but the whole panel decided to be greyed out so I wonder how I'm supposed to get it to work... In this session I blocked out rough colours and organised the layers. I didn't have a tablet with me so I didn't try to draw any details using a mouse.
You can change the background colours through: Edit > Settings > Interface Options: Background colours.
You can also edit the Environment Lighting by going to Display Settings (very far right of screen - Monitor & Cog icon). The default environment map is Panorama
^ I enjoyed using the Charcoal Brush to easily get that fur ^ strand texture.
Because I am using UDims, the eyeballs are positioned at a -1 X axis. Udims record things from 0<. This means I will either have to restart the entire painter file OR paint the eyeballs on a separate file. To save time I will make another file of the racoon to paint the eyeballs using the racoon sculpt for comparison. The struggle will now lie in making my textures cohesive - I will use Unreal to help me out with this.
Baking issue. Found out it was because there was no high poly information for the low poly to pick up. I simply forgot to import it into Max because there are so so many assets that make up the saddle.
went back and imported the part of the saddle that was missing in the high poly.
Perfect bake!!
But then I had to go back and edit the Material IDs of the fixed saddle because I had only fixed this error in the exploded polys. This means I am using two maps when I only have budget for one.
I hadn't unwrapped these??!!! I went back to 3ds Max to do that...
Base colours added.
Decided to explode bake the plinth and stones aswell
Went back into 3ds Max to plan out alphas and decals, like whiskers, duckweed, and Lilli pads. I also enlarged the plinth so that more lilipads can be placed.
POSING, LIGHTING & PRESENTATION
Week 7
I brought the racoon into Unreal to start lighting and setting up folders with their correct naming conventions. It was here where I realised I needed to go back and sort out my material IDs. On the right image you can see that the racoon's head is using the texture map of the body rather than it's own map. Since I was going to pose my racoon this wasn't a problem because ZBrush resets material IDs anyways so I will just sort that out once I have my racoon posed.
In ZBrush I used masking and the move gizmo to rotate, push and pull diferent parts of the racoon. This was a fiddly task and would often bring undesired areas of the topology along with the area I wanted. Despite it being tricky, I had a lot of fun getting to see the raccoon I had worked so long on finally start to come to life. Whilst posing I was aware of the minor issues in the pose which makes it feel a little off but with the time, experiece, and limitation of this workflow, I was happy with the outcome. To improve the pose I should have definitely used a real life reference. Instead I just posed it from imagination and what felt right but this obviously came withissues of it's own.
Giant Raccoon DEMON!
Material IDs have been sorted!!
Now to work on the alphas, Decal, hair cards, and eyeballs because the unwrap didn't allow me to paint on it directly since it was positioned outside of the UDIM threshold. So now I will do them along with the whiskers; both of which weren't possible to add to the substance painter file of the painted racoon textures.
I only realised the massive N-Gon infront of my face once I was in Substance painter because of the crazy triangulation. Went back and fixed that. I also used this time to sort out the smoothing groups and edit the topology of the lilipad because I ended up not liking the outcome so I am greatful for this momet to reinport the mesh.
Brushes I used to create the Duckweed:
Spray Skin - Creates large blotches. good for building irregular round shapes
Spray Coloured - Creates smaller blotches and removes / adds paint to keep the area 50 / 50
Swarm - Pixilates and adds noise to the texture's boarders which is perfect for the mossy texture I'm going for
Note to self: The second you realise the gizmo is in the wrong place, Go back and change it!!! It saves so much time to just start the scene with assets that have their gizmos int the right places, otherwise everything gets moved out of place when you reimport it
I scattered the alpha meshes around. I had forgotten how to make the alpha channel turn on so this image shows the RGB not the RGBA being put into work. A peer helped me locate the masked blend mode (seen below) and my lecturer helped me to locate the double sided button
Opacity mask
To make alphas work, change the Blending Mode to Masked
To make alphas double sided go to the alpha material and tick on Two Sided
My plan was to make the duckweed into a decal but I actually preferred it being a plane because it casts a shadow onto the water which makes it look like it's actually floating on its surface and casting a shadow within the water!!
Finding examples of well-presented 3D characters was surprisingly difficult—not because there isn’t great work out there, but because in most cases, the presentation isn’t doing much beyond what the model or sculpt already achieves on its own.
However, I did find two fantastic examples that take a more intentional approach. On the left, there’s a stylized piece by Dmytro Danylov. He uses a dark backdrop with vignette-like borders, reinforcing the character’s feelings of being trapped or doomed. Additionally, the painterly brushstrokes blend the character into the negative space, creating a seamless transition while emphasizing an unsettled, emotional atmosphere.
On the right is a piece by Robin Tran. The artist masterfully uses lighting, posing, and post-processing to create an elegant and whimsical composition. The warm lighting contrasts beautifully with the cool tones of the owl, enhancing its presence. The swirling motion of the owl’s wings follows a golden ratio-like flow, reinforcing the line of action and adding a dynamic sense of movement. Additionally, the blue hues of the feathers contrast strikingly with the glowing eye, drawing attention and adding depth to the piece.
I was indecisive on the backdrop colour. I liked the deep indigo backdrop because it compliments and contrasts the oranges of the saddle and racoon (since blue and orange or complimentary colours) but I also liked the turquoise green backdrop because it created a cohesive setting which put attention to the raccoon and grouped the background into one space. To get the best of both, I final chose to go for a turquoise backdrop with a deep indigo light hitting the wall to create this slow gradient transition between the two and also giving this gowing halo effect on the raccoons head.
I then added a Niagra System of petals. I reduced the quantity of petals and changed the colour to a green rather than white. I then added a post processing box to increase contrast just a bit and introduce a light bloom effect. I also decided to move the directional light so that it shines on the racoons face head on. This removes the cast shadow from the right side of the face and moves it to the raccoon's mouth and chin which help to frame it's face as the focal point.
To my disappointment, when I opened my Unreal scene later that day, I was met with a dark backdrop that wasn’t there before. I’m not sure what happened, but no matter what I tried, I couldn’t get the wall back to the nice light green color I had originally selected.
I even attempted to add an emissive, but for some reason, I can never get it to glow properly - it always just washes out into a dull grey, which isn’t appealing at all. It wasn’t the world settings, as I had unticked auto exposure when I first set up the scene. I also adjusted the post-processing box, but nothing seemed to change.
I would have loved to ask my lecturer why this happened, as I much prefer the lighter background. It gives the scene a more whimsical and magical feel. Unfortunately, this issue came up outside of Uni hours, so I had to troubleshoot on my own.
How to take High resolution photos of your work:
First, make a duplicate of your screen by going to Window > Viewports
In this new window, change your perspective to camera in Perspective> CineCameraActor
The lens can be blurry and unfocused which was the case for me because I wanted close up shots. You can change this by disabling the Focus Method in Focus Settings
POSTMORTEM
I am really proud of what I have achieved with this project and believe I have fulfilled my personal goals. The raccoon is clearly the main focus, evident through the sculpt, pose, and textures.
One of the biggest challenges was sculpting fur - I was surprised by how difficult it was. Raccoon fur tends to stick out like porcupine hairs, making it tricky to depict through flat texturing. Big, fluffy animals also make it harder to sculpt anatomical structure since there are no visible muscles. Despite this, I’m grateful for the challenge, as it helped me explore different sculpting techniques and develop a better understanding of fur manipulation. That said, if I were to advise a student taking on a similar brief, I would 100% recommend choosing a hairless or short-haired animal to avoid fur eating into valuable texturing time.
One area I could have improved significantly was the saddle. I was far more focused on perfecting the raccoon and was excited to work on the plinth, which led me to underestimate the saddle’s importance. If I had designed a unique saddle, I could have pushed this project to the next level by revealing more of the raccoon's story. The issue wasn’t a lack of effort or time management - it was simply a lack of planning and vision. I only thought about what items my raccoon could carry, not how they would be attached or how the seat would look. This is similar to how I approach clothing for characters - I focus so much on the face, hair, and anatomy that the design and narrative details of the clothing get overlooked. I want to change this by studying garments, bags, saddles, and other accessories that add storytelling elements to a character. If I were to redo this project, my main focus would be designing that damn saddle!
In terms of the plinth, I’m very happy with the colour palette. Since I was running out of time, much of it was left unfinished and under-detailed, but from a distance, it still provides the look I intended. The water lilies, lily pads, and duckweed hold up even under close inspection, though I can’t say the same for the rocks and bushes, which were meant to be placeholders but ended up in the final piece. However, I’m extremely happy with the lighting - every part of the model is clearly visible, and the colours help bring the oranges, greens, and blues to life.
Looking back, I also recognize that my peers relied more on references than I did, which led to them creating more accurate studies of their animals. My raccoon leans more stylized and doesn’t match a realistic raccoon as closely. While my goal was a semi-stylized sculpt, I do wish I had studied its musculature more to make it feel slightly more believable.
Overall, this project was a huge learning experience. I struggled with fur, but it taught me a lot. I overlooked the saddle, but it made me realize the importance of storytelling through accessories. I’m proud of what I created, and if I were to do it again, I’d focus on refining those aspects to make it even better.