For the last scene of the film I wanted to bring it back to the butterfly, however after I did the first render of the butterfly on the flower I thought that after witnessing the child being kidnapped into a life of war and slavery that the flower was the wrong choice. The butterfly on the gun brings together the innocent and free child at the beginning with the child slave kidnapped into war, it brings the juxtaposing representations of the child together in a tragic visual metaphor.
I played about with composition to see what might work best and so I settled on the lower angle.
After I had finished the first render (render 01), during feedback Gerard suggested that like the storyboard I stick to the fully silhouetted render and not have any light on the child at all, so I quickly corrected it.
Running Into More Problems
Originally I had animated the butterfly so that it pivoted at the the bottom of its wings and I simply key-framed the rotation of the wings and exported the file as a game export in maya and imported it into Unreal but unfortunately for some reason it wasn’t working. I followed the instructions on the forums and on YouTube videos explaining how to import key framed objects but it wasn’t working, there wasn’t many ways around it and I can be quite creative a lot of the time but alas I couldn’t get round the problem. I decided that rather than hope that if I keep trying it’ll work soon or I’ll find an answer that I could rig the butterfly quickly, animate it and import it that way.
After rigging the butterfly using the advanced skeleton bug rig, I added a couple extra limb and rigged the butterfly, I just flooded the wings for each side, key framed it and exported it to Unreal and thankfully it worked.
Ronan Mallon a 3rd year student helped me out a little by animating the hand with the butterfly on the tip of his finger, I just imported them both into Unreal with no fuss in the same environment and upped the saturation a little bit to match the storyboards and it was as easy as that thankfully.
For the Final Presentation I quickly rendered out a tif sequence on maya and put it over a orange sky in After Effects just as an example of what the shot would look like and what the animation would be. However I thought it looked too flat and 2-D which I believed would mislead and disorientate people when in the next frame it jumps to a fully 3D scene so I wanted to add some dimensionality to it so as not to confuse the audience when they mistake it for a 2D film.
Its a subtle difference but with the specular highlights on the cheekbone, head and tip of the ear it adds that 3-Dimensionality to it, his shirt is a different colour too which adds some variation in the colour scale.
The first shot was the original lighting after I had lit the last scene. The thing is however, its not silhouetted like in the storyboard, so I had to break the rules a little and take some liberties with the lighting to recreate that silhouette image, because the full immersiveness of the black still holds the implicit connotations that they did at the beginning of the project, nothing has changed. I played around with the light quite subtly to get different degrees of silhouette. For instance I had soldier submerged in the shadow and the boy only partly, as though the darkness has saturated most of whom he is in the frame, with still that bit of light that shines on him as a symbol of who he was and that purity. Light can also be used as a good metaphor for hope. In the end I decided to stay true to the original vision, I shined a little light on the boy to give it a little bit of 3-Dimensionality so that the silhouettes weren’t too flat and 2-D looking.
Firstly I imported, textured and arrangement the props and included some plants and flowers near and around the cabin as well as some branches, twigs and rocks etc. It was important that I had everything in the scene including the character as reference before I begin to light the scene.
I also like some of the last scene painted some grass textures onto the landscape so that I didn’t have to have thousands of planes of grass in my scene and I feel like it works, a lot better than the last couple of scenes that didn’t work so well at least.
Campfire
Sprites Sheet
Material
BP_Light_Flicker
BP_Light_Intensity
With the use of two textures, a flame sprite sheet and a fire gradient I was able to create a campfire using the ‘unreal engine flipbook’ in the material. Using the actor & ‘Fire Blueprints’ I was able to manipulate the intensity of the fire and using the event graph editor, have the fire flickering and always facing camera which was an issue that I had when rendering.
The video tutorial that I followed along is belong if you’re interested in the details of the process.
Sticking to those original concepts I had of the isolated cabin in the forest surrounded by the trees I tried to recreate that as best as possible, I didn’t want it on a hill because I found in my research that there’s not really many flat wide hills in the Rainforests so I thought it’d be best to keep it on flat ground on the edge of the Rainforest so the silhouettes can work well in the last scene without any trees out the front of the camera distracting or taking away from the characters in the shot. I wanted to isolate the cabin in this case, have it stand out, almost as though its a character in the story as well, I didn’t want it hiding in the forest behind trees obscuring the lodge in the scene.
Lighting Tests
Final Lighting Choice
After I the assets were imported, textured and arranged it was time for the lighting tests. For the last few scenes we are at this cabin and in the last shot we have the early evening sky, still quite bright and so it was important not to contradict the lighting in every scene and have the lighting consistent throughout. I also had to keep in mind were the sun was in the sky so that the shadows weren’t casting in opposing directions. I patiently played about with the lighting, and shadows, were they cast, what was being obstructed, what needed to be seen and what was important to the scene. The specular on the ground could be distracting, the shadows could be obtrusive, and equally the lack of shadows surrounded by tall trees could be contradictory and subconsciously take away from the scene so it was important to consider these small details.
Composition
After I was happy with the lighting I experimented a little with composition, what shots could work well and could I incorporate any implicit meaning behind the composition. I wanted to create as much distance between the audience and the characters that I could, similar to the research I’ve already conducted I wanted the audience to be a witness to an atrocity, he’s helpless, hopeless and and we can’t be near him or help him, instead we just watch a cruel and eerie moment come to pass.
When the camera looks down on the subject from a high angle he gets “swallowed up”. The High-angle shots make the child appear vulnerable or powerless.
Similarly I considered low angle shots in the scenes where the captor stands behind the boy, Psychologically, the effect of the low-angle shot is that it makes the subject look strong and powerful. And by towering over the boy it can appoint that power over the child, and still yet, take the power away from the boy.
Again I’ve already storyboarded my compositions but its sometimes worth a try experimenting with the scene when its completed and almost ready to be rendered.
After Feedback
After the feedback and exactly like the other scenes we’ve discussed I went back and changed the environment around the house to imitated something closer to a Rainforest, personally I’m happier with the redesign and think it looks at least a little more appropriate.
First of all I thought I would paint the grass onto the landscape just to see how it would look, I just blended the Diffuse/Spec/Normals of different textures together in a material using the blend node and a texture coordinate and multiply to manipulate the UVs and then I would simply paint directly onto the landscape.
I didn’t like how it turned out, it was worth a shot but I thought it looked too unrealistic, perhaps a good technique for things in the distance like rocky/snowy/grassy mountains or hills etc but for a close up shot I wanted to use real grass. I was however able to paint out a natural path as a stencil for painting on the foliage.
Running Into Problems
When I started painting in the foliage it wasn’t coming up in the rendered video. I had no idea what the issue was, I thought after some consideration it was perhaps a scale issue but when I tested that out it wasn’t the issue, I also thought perhaps it was something to do with the camera but after playing about with the settings I assumed it wasn’t. So after searching on Youtube and going to the unreal forums to see what the issue was I found out it was to do with the cull distance when painting in the foliage. I had to extend the cull distance so the camera could pick it up in the render. Simple enough, I just had to play about with the numbers. That being said after a certain amount of distance it still wasn’t picking up the foliage in the background, so I scaled up the foliage a little and it seemed to do the trick.
After fixing the foliage issue I painted in some stones and sticks in the path just to give it a bit of extra something.
This I thought would be a challenging one and I was correct, I was having trouble with getting the cavern of trees that I had originally storyboarded. There was ways of bending the trees but they looked unnatural and it somehow worked even less, they let less light in from the top which was my main source of light but nonetheless I was patient and careful and did what I could under that the circumstances. Because the trees were so close together it was going to be dark and so I had to be creative with the lighting and the placement of the trees, as well as considering the space around the trees and not obstructing the light too much. If it was too dense in the scene it wouldn’t work but I wanted to give the allusion that we were still in the forest and that this wasn’t a tunnel of trees on a path in Ireland somewhere.
I did the best I could do under the somewhat restrictive circumstances, I too understand that its really more reminiscent of a forest, that problem that keeps reoccurring in my project but the tropical trees and palm trees and leaves didn’t quite work for this scene, but I think perhaps I might be able to get away with it.
Again a similar problem for this was originally I thought the shot was too bright, I wanted to be sure to keep that ominousness that I’ve been trying to capture in these last scenes. I was concerned that if it was too dark you couldn’t see whats happening or where the boy would be in the shot, however it was about achieving that gloominess and obscurity, a darkness that is evident but were you can still see whats happening.
Similarly to the last couple scenes I was making the mistake of creating a scene more akin to that of a woodland forest not a Rainforest so first of all I repainted the trees in the background with trees and foliage more affiliated with a Rainforest. I believe I was getting closer to something that I wanted, taking into consideration what Gerard said, flattening the colours of the trees and abating the detail of the trees in the background. I also pushed the trees further back to recreate that depth and enormity, the spaciousness of a Rainforest, the boy being something related to a needle in a haystack.
I wasn’t liking the heavy saturation of the moss on the fallen tree stump, for the life of me I couldn’t decrease its saturation so I considered replacing it with a similar asset
After replacing the tree and being happier with what I had now it was important to get the lighting nailed down. The 1st image above I thought was too bright and that the fallen tree in the background was a little too distracting perhaps, I wanted to really isolate the boy. Being alone in this big Rainforest running for his life I wanted to somewhat disengage him from his surroundings, its dark and unfamiliar, we don’t know what exactly is in the shadows and neither does he; everything is frightening.
Its really quite a subtle difference but I thought the second was just a tad too dark and so I increased the exposure and tilted the y axis of the camera a little bit. I think I’m happy with how it turned out and I guess I’ll just wait to here some feedback on it at the next tutorial.
Rendering Issues
After doing a test render I realised that the exposure would increase at the beginning. It was obviously very apparent and very distracting but I didn’t know how to fix it. It took me a while before I found the answer online. It was a post processing volume issue, I just had to check the infinite (unbound) box and put the exposure to min and max brightness 1.
Another problem I ran into was that the boy wouldn’t render when he was off screen. The camera would render him when he was on screen but not off. I couldn’t for the life of me find out why and I searched the highs and lows of the internet and I couldn’t find an answer anywhere. For hours I was finding hack ways around it but I obviously wanted to know what the issue and then embarrassingly I realised that you have to add the character and animation to the track editor.
Problem solved. And at least I know what the issue will be in the future and can solve it immediately.
My first attempt at the composition of the shot appeared a bit cluttered and the flat lighting and dark background made it a bit difficult to see what was happening in the scene. But I had the right idea for the closeup, composition and formation of the assets.
Puddle Material
To make the scene a little more interesting I added in a puddle, it adds to the wet and humid atmosphere of the Rainforest. I also changed the colour of the sky dome to something a little brighter and increased the exposure a little more just to brighten the scene a bit. I
In the background I removed the trees to create more depth to the scene to give the allusion to the large scale of the forest. I also replaced the ferns with something a little more tropical a using a blending node in the editor I was able to take away that sharp line at the edge of the puddle and get something much more realistic. I also took down the brightness down a little to match the colour and feel of the scenes on either side of this one, so that there’s a flow and you get the feeling hes on a continuitous journey through the forest and not hopping all over the place. The lighting and colour tells a story too and its important to keep that consistent.
Initially I was just getting the lay of the land, experimenting with the foliage tool, the different assets that I could use, what trees looked the part and how the lighting interacts with the environment, the lighting will help a lot with the colour of the scene and considering colour is an important part of the story, so is the lighting in this case.
Experimenting with composition and assets I patiently composed a number of different shots with different arrangements and formations to make the scene interesting as a setting for the physical encounter between the rebel and the boy. I wanted something reminiscent of a photograph, looking at different photos online and different photographers portfolios of their journeys through rain forest, how they composed some of their shots.
I played around with the sky dome, to recreate skies more reminiscent of an early morning, that being said however, the skies in the scene behind the forest dont recreate the same colour of the storyboards or create that ominous setting evocative of the physical altercation between the boy and the rebel. The scene has to set the mood so the colour needs to be more akin to that of my storyboards, I need to be sure to keep coming back to the source material and not stray from the original vision and foundations of that. The forest has to be denser, foggier, more atmospheric.
Contextual Research
Border Ranges National Park
Initially the environments that I was creating were looking too much like a forest, I wasn’t getting those Rainforest vibes that I needed, and so I was doing the best I could with the resources and models that I had, I was looking at my research and trying to characterise what makes a Rainforest a Rainforest, what plants and trees were there, what the ground was like etc.
In the research its apparent that there’s a lot more foliage, tall foliage and twisting, crooked huge trees, leaves covering the ground, mossy rocks and rivers and ponds.
Like the greats do when they characterise a person and include and accentuate their most noticeable and prominent, important features, I had to do the same with the environment. Include the essence of the Rainforest into my environment. The things that made a Rainforest.
Second Attempt After Feedback
Quixel
Using quixel megascans and the bridge application I was able to quickly and conveniently acquire more Rainforest type assets and really start to create something like an African Rainforest. Unfortunately because there wasn’t many rain forest like assets available on the store and really just more woodland forest type assets it was rather difficult to get to the point were it was believable but I believe I did a rather good job considering the circumstances, it looks a lot more akin to a Rainforest than the previous renders.
Gerard said that it was important that I don’t stray too far between the style of the silhouettes and the rather simplistic aesthetic at the beginning of the film and wander too far into recreating something too realistic. He suggested that I flatten the colour a little, so that you don’t see every fine detail on the trees, rather to have the shapes and silhouettes of most of the nature and keep the focus on the characters.
I completely agreed and remembered to refer back to my storyboards as reference so that I don’t start straying again from the original idea. The flattening of the colours and the silhouetted trees around the characters was more consistent with the style of my story-boarded concept.
After playing about with the composition a little, I wanted to set a stage for the altercation. I didn’t want anything obstructing the view of the characters or being a distraction in the scene, so I decided to trim down the foliage a little and remove some of the trees. It still maintains the look of a Rainforest though and is a clearer stage for the most important scene in the film.
Animating has never been my strength, I always enjoyed it but it can be very tedious and a lengthy process, especially with the amount of animating I had to do and as much of an amateur as I am at the process. So I decided that I could and would use a combination of mixamo animation with my own, and blend them together to create the animations I needed. I wont be able to use mixamo for the entire process but it’ll certainly save me time in the long run, I just have to hope that I can blend the animations together without any hiccups on Cinema 4D (which I’m totally unfamiliar with).
Cinema 4D
So I had to get a little familiar with Cinema 4D first but thankfully it didn’t take me that long, it is quite similar to all the other computer graphics softwares so I had a fair idea of what I was doing.
When importing the separate animations into Cinema 4D I learned that it was important first of all to begin with a TPose, Unfortunately I learned that the hard way when I didn’t think it would be necessary but instead my animation would glitch out uncontrollably. It important as well that I chose animations that I thought would run together quite seamlessly, sometimes when legs and arms and posture are in different positions the the model can jolt about and look quite robotic, and the characters limbs would slide about. It was also important for me to keep this in mind when I was Animating my own parts, I would regularly use the pose from the last frame of the animation to blend in to my own. I’d simply export it as a game exporter file and import it into C4D.
After importing each animation separately and grouping them together, you simply add them to the timeline as you can see above and drag them on in the right order. Afterwards using the 3D viewport you sync the animations up as seamlessly as you can, it can be tricky at times and tedious but overall it saved me a lot of time. I was able to complete all my animations in about 2 and half weeks, which I believe is at least a 3rd of the time it would have taken me to animate key every frame myself in maya.
I’m currently still on track and can continue on now to modelling the environments in Unreal Engine, it gives me a bit of extra time to familiarise myself with Unreal as well.
Below is some of the animations as examples, I just included a few, rather than all of them.
After Feedback
Thankfully there wasn’t much improvements or corrections I had to do at all, Henry and Gerard only had 2. During the fight scene Henry suggested that I have the boy taunt the soldier and throw something at him so I added that in. Gerard suggested that at the end when the soldier is walking out onto the porch that he lean on the frame so I included that as well.
I modelled some props that I wanted to have on the porch of the cabin and around it in the environment, just to make it a little more interesting and give the environment and the cabin more of a personality and “lived-in-ness”. Just simple, rather low poly models that I could add more detail to in substance.
With Substance I’m always able to get some really nice results. I just baked the meshes, applied a simple wooden smart material, some layers of dirt etc and using the paint layers I attentively painted in some personality to the props, dirt and scuff marks where there would be. I also used some Alphas and playing about with the height was able to create some chips, scratches and carvings into the wood. Also using the Alphas I marked some graffiti and doodles onto the chair and stool, to give it a little something extra. It reminded me of the chairs in school, I just thought it added more realism to it.
Texturing AK_47
Similarly with the AK I added the appropriate textures to the gun, the aged steel and wood and then on-top of that added some wear and tear to the edges and a little but of rust at the top and bottom where I believed some rust would accumulate. I added some bloody finger prints and dirt as well just to make it look more like it belonged to somebody, like it had been used and definitely has some stories. It simply makes it more interesting to the audience. I figured the rebel militias wouldn’t be getting brand spanking new weapons sent to them from anywhere really, so they would all be old and a little damaged.
Texturing Cabin
Again for the Cabin I just did the same thing, adding scratches and carvings on the posts and the door frames, adding blood on the porch and dirt, muddy footprints and wear and tear on the edges of the wood. It just gives it more of a realistic, authentic look to the model and again adds depth to it, a story to it, a past that the audience didn’t witness, wear did the blood come from? How longs the cabin been there? Who lives in it etc. Like I said before, textures can evoke a depth and provocation to a model, an extra bit of information, history and narrative.
Similar to the other characters I textured I wanted to keep in theme with the old donated, hand me down clothes prevailing in the Congo, especially amongst the displaced and vagabond child soldiers abducted into these militias. Although its all very sombre, I can still have creative freedom and expression in my Art and so decided to continue with the theme of the 90s clothing, really not as unbelievable as one might think anyway. So I continued my 90s aesthetic sports attire research and came across a few t shirts I thought could work well and were still rather appropriate, which is something I’m always mindful of; “Is this choice appropriate?”.
I wasn’t able to find the Adidas team logo or conveniently trace over any image of it on a t shirt because it was usually folded and distorted so I sketched out the logo on Photoshop as an Alpha for Substance. I did the same with the blue Adidas logo. I also stamped the moderately well known “Punk is not ded” from Persepolis, one of my favourite “mature themed animations” as a homage to an inspiration of mine and another example of an animation that tackles more serious themes.
With the main character I knew exactly how I wanted him to look. Although the clothes aren’t considerably interesting, it keeps to the authenticity of the realistic aesthetic I’m going for. Using references from Rebel Militia leaders, Sgt. and Commandments I combined the realism and the reoccurring motifs of the uniforms with the concepts I had in my head and had already sketched out.
Texture_Maps
Because of the lack of close ups of the main character and and the obscuring shadows I wasn’t concerned with the smaller more, particular details. I did however add in some personality to the clothes, like dirt, small blood stains, splashes of dirt and dirty water etc. It gives the character more personality, makes the audience wonder what the character has been doing, who he is, where he’s been and what has happened to him, it adds more authenticity and it makes it just that little more real for the audience. Seems strange that dirt on the bottom of his green cargo’s and the blood on his boots would evoke such thought provocation, but it helps add layers to the characters, and shows that as an artist and a texture artist have considered who my character is and what he might have done, it shows I’m invested in my characters. How can I expect my audience to be invested in my characters and my world if I’m not?
Particularity with the cargo trousers I played around with different layers of detail, like the intensity of the folds and creases in particular and the fibres of the material. I’ve 2 separate files so that I can do lighting tests and see what one works better in the lit environment in unreal engine.
Above are some references images of Rebel Militia Soldiers in Africa. As you can see the children particularly aren’t wearing a full uniform, the full camo and the jackets etc are reserved for leaders and Commandments. So usually the children and the newly kidnapped or inducted members of the Militia wear hand-me-down t shirts, vests, shirts etc, or clothes that they have stolen from villages or dead bodies.
Substance Painter
As usual the first one textured is the background character. I wanted the commandment character in my animation to stand out as more powerful than the background characters as I’ve stated time again and clothes are another good way in this case to establish rank or prestige. So for the background characters I dressed them in casual retro sports attire with camo cargo’s. There was no particular reason for the choice of torso attire, I just liked the design and thought after looking at a lot retro and old school athletic apparel online, I thought the design looked well and was interesting but not too obnoxious.
Because they’re background characters I didn’t worry about having to model another character. The plan was to give one a bandanna and then have each wear different clothes. I can remove the bandanna on the other in Maya before I animate. For the second character because it was quite common in the footage, documentaries and photographs I saw, I decided to have the soldier clothed in a football jersey. I made it an Henry jersey to keep in theme with the old clothes even though you almost certainly won’t see the back of the character, I can use it for online portfolios or show-reels. I made it an Arsenal jersey just a homage to my late grandfather.
I also tried some variation of colours etc and wrestling with the idea to just keep it simple but I thought most of the colours like the blues, turquoises, purples, greens etc clashed with the colours of the skies and environments. This is also the reason I avoided the use of Oranges, Pinks and Yellows as well.
So after I had my Characters Modelled, Clothed and UV Mapped, it was time to Rig them. Alec suggested that I use Advanced Skeleton 5 plug in for Maya and so I did. I downloaded the plug in and got on it.
First after you import the bipedal rig, you have to conform it to the form of the model. You have to precise when doing this, making sure you place each joint in the right position so that you don’t have any problems later down the line. Its a rather tedious process, especially when it comes to the fingers; especially at the angle my characters arms are at. Hes in an A pose rather than a T pose so I wasn’t able to use the orthographic views to easily place them, I had to use the perspective view and just be patient and meticulous.
After you have placed the skeleton you simply hit build advanced skeleton and it does in seconds what would take an amateur like me maybe a week. With IK/FK joints as well. A beautiful thing. I recommend the rigging tool to anyone.
I had some issues with the skinning in the beginning, I have to say the tool does a great job at skinning as much as it can but I had to go in with paint weights tool and fix up a few stretching issues with the shoulder joint and its influence on the T shirt.
Main Antagonist Character
So I did the exact same process as the other model and was getting the same issues basically. So I skinned the model as much as I could and used the component editor to revoke the majority of the influence from the effecting joints. But it just didn’t seem to be working. I wasn’t sure why. Alec had showed me some of the mistakes I was making like, not having imported the “reference model” into to the scene, using the component editor and resetting the paint weights tool. It still didn’t seem to be working.
I attempted to get around the problem but doing things that I thought might work; for instance, separating the t shirt and and jacket geometry. Didn’t work. Separating and deleting the hidden inside mesh of the jacket that couldn’t be seen, I thought this was my final and most painful thing to do and after I did it and re rigged t he character it didn’t work. I was still facing the same issue.
Time was creeping onward and I needed the characters to be rigged so I could animate them. I need to start animating soon so they had to be finished. I spoke to Jack a classmate of mine and he said that I could rig him without the jacket and animate the character then import him into Marvelous designer to simulate the jacket in that then bake out the animation from Marvelous. I thought it would probably have to be what I needed to do because I felt as though I couldn’t spend anymore time on this problem. Another character needed rigged.
I thought that because it went smoothly with the other character that it would with this one but I just couldn’t catch a break unfortunately. I was beginning to become concerned, I couldn’t figure out the problem and I felt as though I had ran out of options. I left it a couple days and continued on with some other work that I had to do in the mean time until my tutorial that Tuesday. I thought perhaps somebody could help me out and I could maybe resolve the problem if I was lucky.
I consulted with Alec about the problem and he made a few adjustments with the component editor, assigning 100% of the influence in the affected areas to the spine joint. It worked. Praise be, it worked. And after he showed me how to resolve the issue, with a little patience and understanding I was able to immensely improve the skinning of the model. I can at least say I’ve learned a lot from this experience and will know how to resolve similar problems in the future, so it wasn’t a complete waste of time I guess.
Rigging the Boy
Again just the same process as before and the same problems but thankfully after the mistakes I made on the last model I was able to quickly correct and skin this model.
I rigged the boy with the helmet and sandals so that I can just delete that geometry when I’m animating him at the beginning of the film were hes in just shorts, TShirt and bare feet. Saves me having to rig him twice. I can change the colour or pattern of his T Shirt and shorts. The references of the Congolese child soldiers that I looked at all have shorts and sandals anyway. Rarely do the militias acquire standard militarily uniforms so the kids can be seen in casual camo wear.
Below is the tutorial that I followed along with if you’re interesting in the full process.
Originally for the boy I was struggling to get the form correct. I just couldn’t seem to get the shape of a child, rather most attempts I had done just looked like a little man. So I showed Henry and Brian my issue and Henry very kindly gave me a base mesh of character of his and recommended that I import a reference image into ZBrush and shape the character like that. This helped tremendously and I was able to come up with a better looking character concept for the child.
Uldis Zarins
Similarly to the last character I referred to Uldis Zarins book on anatomy to characterise how a child’s skull differs from an adults skull.
Second & Successful Attempt
Marvelous Designer
Like the other 2 models I brought the boy into Marvelous to simulate his clothing. I did the T Shirt the same way I did the others, you can revert back to the older posts to see how I did it. With the shorts it was similar to the trousers except I just took the length off the bottoms a little. Because hes a main character and will be predominately focused on in the animation I decreased the particle size to get more wrinkles and interesting “more realistic” folds and creases in the clothing.
I downloaded a helmet from free3d.com to save myself more time and I fixed a couple of things with the clothing in ZBrush, using the Zremesher function and Zremesher brush tool to correct the triangular mesh and make UV mapping the clothing easier for myself in maya.
Using my original model as a base I wanted to sculpt more details on it; so that when I bring it into Marvellous Designer I can achieve that more realistic appearance and fit to the body that I’m going for. With Marvelous it simulates the textures of real clothing so I wanted the pectorals and biceps to be visible under the shirt. It will help create that buff, powerful look that will contribute to the frightening and intimidating characteristics of the main antagonist.
I perhaps spent longer on the model than necessary, fixing small issues but as well as the challenge I also wanted to consider what I can use for my portfolio and showreel. It means its closer to a finished sculpt now; I’ll not need to fix it up as much later down the line.
Uldis Zarins’ book on Anatomy of Facial Expression
Henry recommended to me that I go to the Library and check out Anatomy of Facial Expression, so I did. He had explained to me as you can see below that people in different parts of the world have different shaped skulls etc and that if I am to properly and accurately sculpt my characters I should consider what attributes an African skull and face include.
African skulls have more rectangular shaped orbits when compared to European and Asian skulls. The nasal aperture is wide and the nasal bridge is very flat. The jaw protrudes significantly from the rest of the face so I kept this in mind when sculpting my head model. That being said I could take artistic liberties and I had my reference photos of African males to work from too.
I modelled the head separately before I put it on the model. With the head combined with the rest of the body (so that there was no seams) it meant that I was losing definition in the face. So I wanted to separate the head from the body to get a more refined and detailed face so that after I clothe the model I can delete the body underneath to reduce the poly count and have a more detailed and refined face for the animation.
FiberMesh
After the head was sculpted I was just playing around with the fibermesh hair, for fun really. I thought that it was perhaps unnecessary and really rather more of an inconvenience to have to export from ZBrush to maya, exporting the curves recreating it on maya and then importing it to unreal. Its still a possibility if I have time and it looks really cool but I have prioritise my tasks and do what it is necessary first before experimenting with inessential details. Still a cool feature and one that I’m comfortable with, so I’d love to come back to it at some point.
Another thing I was curious about and experimented with a little was the AI Skin shader on ZBrush just to get a insight into what the character will look like with the skin texture before I have to retopo, UV Map it and texture it in substance.
Retopology with 3DWrap
I imported the high density model of the head into Wrap3D and using the node editor I synced up the imported “Retoped base mesh” and my high density one. I plotted the numbered points on both heads and afterwards I simply hit compute and the software Retopologised my model.
Marvelous Designer
After I had sculpted the character I brought the combined model into Marvelous to simulate the clothing. I hadn’t used Marvelous before so it was a bit of a learning curve for the first few hours but I quickly became accustomed to the way it worked, like most things patience was the key.
I originally sketched out the dimensions of the T-Shirt and sleeves to use as a template to help me out a little, rather than just eyeballing it. I’ve attached the video tutorial I followed below if you would like to follow my process, rather than me going through each meticulous step.
For the Cargo trousers it was a lot more difficult and time consuming, I wanted to make them look like actual cargo trousers and not sweat bottoms or jeans, so I had to humour the details a little like the pleating on the pants, the hem, the side pockets and most annoying and time consuming of all, the pleated pockets. Which was challenging but rewarding once I had finished them, I think they look pleated and I’m glad I took the extra bit of time to include the smaller details of the trousers. Again below I’ve attached the video that I used to follow along.
I did a variety of different T shirts for the model just to compare and contrast and see what would look the best rather than just going for the first one that I create. I also did the same for the trousers, I just had 2 different versions because I knew how I wanted the trousers to look but after playing around with the particle distance I was able to increase and decrease the the amount of creases in the trouser leg.
I did the jacket the same way I did the T shirt, I just increased the length of the sleeves and changed the fabric to Denim for more strength similar to a jacket. I also mistakenly omitted the trousers from the model the first time, so I had to go back in and re-simulate the jacket with the trousers on the model so it would rest over the trousers authentically, without having to manipulate anything in ZBrush.
The collar was tricky, there was no tutorials on how to do it so I had to figure out how to do it myself but I feel as though I was successful and I’m happy with the finished result.
Head Remodelling
After I sculpted and retopologised the head I decided I just wasn’t feeling the look of the Main Antagonist character anymore. Probably a silly thing to do considering the characters face will most likely be shrouded in shadows through most of the film I still decided to sculpt and retopologise another head. It only took me a day so I it didn’t hinder my progress too much.
Doing the exact same process as the other head model, I wrapped it on 3DWrap and after consulting Henry about the mesh, I was concerned it was maybe still too high and I might encounter some issues when I’m rigging. Henry suggested that I just deleted a few of the edge loops and it should be fine, so I just deleted every other edge loop on the face, it should help a lot with the re topology. And I can just smooth it out on maya before baking the animation.
I used Zremesher and the re-topological brush to retopologise the clothes. I can UV Map it easily on Maya.
I also uploaded the Model up to the Mixamo website to view the character in some different action poses before I rig him, just to get more of a look and feel to the character.
Originally I had a scene were the boy is hugging his mother but after I had started this model I was told that I probably could omit the scene from the animation. It would still make sense, be less confusing for the audience who might misunderstand the scene and it would ultimately be less work for me.
I had originally just modelled the body using a sphere for the torso and limbs. I used the “Default Female Head” as a base for the head in the mean time before I was ultimately going to do my own head model. Although I didn’t get to finish the model its still a good start, it diversifies my portfolio and it is a good foundation in my asset library now if I ever need a female body. It was interesting sculpting the female body and surprisingly more different from sculpting the male form than you might think. Its ultimately a development of my skills, an addition to my anatomy sculpts and another addition of diversification of form in my portfolio, not a total waste of time in the end.
Originally I intended on doing the background character quite emaciated looking, flat, retreating, stomach. Prominent ribcage, hip bones jutting out and long boney arms. Again the idea was to accentuate the prominent, distinguishing characteristics of the character but as I continued modelling I just wasn’t feeling that it was the right direction, that it looked good or that it was even how I was feeling about my character anymore.
So using the the foundations of what I already had, instead of starting from scratch I just built on the model and immediately it began to take shape, more realistically, appropriately and just better overall. I built the chest out more, gave him more of a stronger physique but was still able to attain those skinny attributes, like the ribs, the shoulder blades the thin arms, and the visible spine.
Again this is a background character and the details weren’t necessary but I was again thinking about portfolio and showreel and wanted to display a diverse catalogue of anatomical studies and models. Its what I want to do after I graduate so I’m keeping that in mind as well as the project. Under other circumstances I wouldn’t have entertained the details of the anatomy as much as I did in order to save time.
I again did the head separately to reserve more of the detail in the face and delete the body underneath the clothes like the other models to reduce the file size. I kept the retopology low because the character is in the background and his face will be far less visible. Again, the models, if not visible in the animation can be presented on my portfolios and showreel.
Staying consistent, I simulated the clothing on Marvelous and enabled the elasticated tool on the tshirt at the bottom so it would create a more “tucked into the trousers” look. When I simulated the trousers over the tshirt and used the pinching tool to try and recreate the “tucked in” look, it wasn’t working. So I had to improvise and be patient; patience is key a lot of the time with Marvelous Designer.
Thankfully with the trousers I didn’t have to worry about pockets or pleats because he’s a BGC. It was a basic simulation and I cranked down the particle distance to get a better, smoother simulation that created more wrinkles in the trousers, just to make it a bit more interesting.
I used Marvelous to simulate a bandana for the character as well. There are subtle variations in the two just for some variety. I thought it would be an interesting addition to the character and would also be more akin to the research I did. A lot of the soldiers cover their faces, similar to many rebel militias around the world so to give the animation and the character more faithfulness to my research and reality I would try include a bandana and a ski mask. I also modelled a beanie just for kicks, I’ll not use it but it was fun to experiment with different shapes and methods of creation on Marvelous. I thought I may as well get my “free trials” worth.
I also did a couple variations of Beanies, really just because I thought it would be fun. Its probably a little less appropriate for the African militia soldier but it was a fun experiment. If you would like to know what my process for creating these items of clothing I attached the video tutorials below.
After the feedback and the reconsideration of the style I went back to modelling the background character from scratch. Again I didn’t have to focus too much on the details and rather just work on getting those realistic body proportions correct. Its the same organic process as the last, modelling the torso, head and limbs with separate spheres, dynamesh it together and use use ZRemesher to really nail down that solid, anatmocially correct form. I didn’t need to re dynamesh because the finer details aren’t necessary.
Similar to the clothes on the last model I just used the lasso tool as a stencil, extruded and separated as a subtool, then used the standard and clay buildup to add the details of the folds and wrinkles. As I said the character will feature predominately in the background and will be shrouded in ominous shadows so I didn’t pay to much attention to the details on this model. I want to focus primarily on the child and the main antagonist. For the boots, I modelled the basic shape on maya, then on ZBrush used the lasso tool to extract the “blutcher” part of the boot and used ZSpheres to create the laces. I was more attentive to detail because I can use the boots for the other characters too.
duckinthesobriety 