Personal Projects during Placement

by

 

I haven’t found much time to look at personal projects during this placement year, but I’ve had a few experiments and works in progress. I started with a likeness sculpt of Anya Taylor Joy and then decided to plug it into mesh to metahuman to see how well it translates and to play around with this software. There would need to be further adjustments made to the metahuman achieve a better likeness but I saw some similarities between the two, I thought perhaps recreating the hair in xgen would help but I didn’t get around to this.

I did manage to delve into Marvelous Designer and achieved something really quite similar to her outfit in the movie, I would need to sculpt on the leather belt strap but I was happy with how this came out and the texture process in substance painter was fun.

I did also play around more in MD since I really enjoyed the software, and hope to use it in my final year project.

 

I wanted to work on digital painting some more since I hadn’t really done much in my free time since starting the course but I didn’t stick with it. I would like to revisit it and improve my skills a bit since I do find the process quite relaxing.

 

 

 

 

I started a sculpt inspired by Arcane’s Jinx. Looking at more stylized and planar forms, and hand painted textures (eventually). I want to incorporate something stylized in my final year project over realism, but I’m not sure of the style I wish to achieve yet. I am Inspired by Arcane but unsure if this is the route I want to go or try something a little bit less derivative.

Vertical Slice – Portfolio

by

Concept art and test model/anim

 

 

 

 

 

 

 

 

 

 

 

Grass

 

 

 

 

 

 

 

 

 

 

 

 

Mushroom

 

 

 

 

 

 

Landscape Materials (and textures)

 

 

 

 

 

 

 

Smart Material/Texture guide (not my assets, just showing material)

 

 

 

 

 

 

 

 

 

 

 

FaultE (everything)

 

17 unique FaultE anims (used in final)

 

3 climb anims

 

8 walks and 3 beams (blended overtop)

 

2 pushes

 

Scans

 

one by one

 

 

Carniviper and Evil Mushroom remodel and rigs

 

 

 

 

 

 

 

 

 

 

Carniviper and Evil Mushroom’s 4 animations

 

All character animbp logic and states (and montages + blendspace setup)

 

Niagara Particle FX (smoke, leaves, spores, bubbles, scan, beam and thrusters)  (beam/scan/thrusters spawn – scripting)

 

Cloth Simulation

 

Level Design for first two playtests, final lighting and atmospheric fog set up in finalgame (0:15 onwards)

 

Textured Posi

 

Vertex Painting Material setup

 

 

 

https://www.youtube.com/watch?v=3Awzd32S188&ab_channel=TheCondorgaming

Vertical Slice Group Project

by

Pre-Production: Paintings and test model/animation

We started out by using the wall in class, writing up genres of stories and types of games we all wanted to work on. This was followed by a vote, where we all voted out the least favourites. We ended up going for a Sci-Fi Fantasy Walking Simulator, didn’t sound like the most interesting game but definitely had some scope for cool art. Then we brainstormed ideas out from what the storyline could be since this was the main factor of walking simulators. This was all recorded on our miro board by Sam.

(This blog is less detailed than I hoped in terms of documenting my development process because I just lost 240 screencaps </3)

 

 

 

 

Then we organised what role each member would take on, with me being given character modelling and art direction. I was happy enough with this since I was hoping to work on either character modelling or character animation. Using the brainstormed ideas from Miro I then made some concepts. In terms of art direction, I was hoping to go for hand painted and stylised assets – so I painted my concepts over sketching.

 

 

 

 

 

I really like how the coloured thumbnails came out (first two). I wasn’t 100% on what a walking simulator entailed but I imagined we could play into loneliness, maybe an archaeological robot alone on an alien planet that he was sent to for documentation. I was thinking of games like ICO and shadow of colossus, playing a lot with the scale of the forest to help convey that theme of isolation. The concept on the right is looking at some potential assets, some fungi, ruined technology (perhaps this add a mystery element) as well as a strange alien plant creature. This was a little muddy but I still like how they came out.

 

Another game I thought would be good for environment inspiration was the last guardian, from the same team as ico and shadow of the colossus, with similar themes.

 

 

 

 

 

 

 

I loved this design I found on Sketchfab while I was looking at design incorporating robotics with cloth, so I thought this would be a good model to learn from. We wanted a medium poly style so I used this model as inspiration, but obviously not recreating it. I approached it in a weird way, placing different cubes and using quad draw to create the faces with cleaner topology. I realised I could model it in pieces easily since the character wouldn’t require organic deformations, so I modelled it similar to how it would be structured in ‘reality’, with the helmet as a separate piece.

 

 

 

 

 

 

 

 

I also hadn’t modelled a character in a while and never modelled a robot character, so using this model as inspiration was useful, but as you can see I started to branch from it. The group were mentioning Ghibli and breath of the wild as inspirations, so I added a Ghibli screenshot as I modelled to see how he would fit into that world. I wanted him to look quite peaceful, so I gave him weaker “features” (chin) as well as bad posture.

 

 

 

 

 

As the ideas progressed from the first three concepts, I added a bit more colour to better match the whimsical themes the group were heading towards. I also played with the idea of a companion in game, a native butterfly or a fairy creature, as well as a hard to make out mushroom creature across the river (to tie into the fantasy creatures). This painting wasn’t as strong as the previous few, I’m still trying to figure out painting but I think the simplification in the previous environment shots was what I liked so much about them, this more detailed attempt ended up much muddier.

As mentioned I was hoping to work on character animation too, so I also mocked up some idle animations to add some life to the character. This wasn’t rigged or anything, more just having fun seeing how this guy might act if we went down this route.

 

I liked the idea of having two binocular eyes and a little antenna, I felt like it made him feel a bit more animalistic. I thought it might’ve been cool for him to have no real arms, just a long cloak and legs, thinking we could maybe play into birdlike movements for him as he trotted along and inspected his surroundings.

At this point Jordan wanted to draw up final concepts for me to model from, his concepts were going a different route from what I was doing so far so I moved away from the character for a while and decided to make some assets we could maybe use as guides for environment art style. I still found this really useful as research and practice into hard surface characters though.

 

Environmental: Grass, Landscape and Mushroom assets

 

 

 

 

Since we were looking at things like Ghibli and BOTW for art inspiration now, I thought we needed a grassy plains, so I found some grass in games I thought fit this style, BOTW, Genshin and The First Tree. They were low poly, tall and felt fluffy.

 

 

https://www.youtube.com/shorts/8z645zPawhY?&ab_channel=Financian

This 30 second video from Stylized Station was what I really based the grass on, as well as the video above from Marpetak. Marpetak went into more detail, teaching me about vertex painting in order to change how the wind in unreal effects the mesh, as well as aligning the normals to point up, so the shading works correct in unreal engine.

 

 

 

 

 

 

 

 

 

 

 

This is the grass I created. From my research I read that Unity is better with alpha transparency while Unreal Engine can handle polygon count better for large foliage instances, so I started by modelling this simple grass mesh in Maya and duplicating it, making a small bunch of grass. I also eventually made a much lower poly version with less blades as an LOD, this would spawn in place of the original when you are further away, optimisation like this is important for the game to have better performance.

I have a fairly surface level understanding of Unreal Materials, but this setup wasn’t crazy, just some textures to add variation in colour and simplegrasswind, an in engine preset for adding “wind” to assets, affecting the world position offset. The above videos use very similar techniques, while my final grass ended up a little different the logic behind them is very similar.

 

 

Loved how the grass came out even on the first pass, think I managed to achieve similar enough to the references I looked at and as time progressed it deviated and changed to fit our final style easily since it was set up in way that’s easy to edit. I did notice the gaps without grass were noticeable when I just applied a green colour to the surface, leading to me working on the landscape materials we would use to paint our level with.

 

 

 

 

 

 

I attempted to make a landscape material for this grass, with similar colours and texture variation in it, as well as normal information faking blades of grass. I set the grass up with virtual runtime textures (wasn’t used in final level). This would let the grass sample the colour the landscape was painted, then apply a gradient between the two, making it look like a coherent part of the landscape. I eventually developed a much stronger landscape material with a kind of with a modular set up, but this was useful for learning how to approach this and worked well for the time being. I was a big fan of where the grass was going at this point and was getting good feedback from the tutors, but I’m not sure why the group didn’t go the grassy plain route in the level.

 

 

This is a video from much later with my more finalised grass, it was set up 100% parameterised, allowing our level designer to customise it to her liking. It was set up with LOD’s, automatic decimation of tris based on camera distance as well as two meshes in foliage mode, one less dense that culls at a further distance, and more dense version with a smaller cull distance, helping blend the grass as it despawns. This in addition to the landscape I eventually developed made it very hard to differentiate grass from landscape material at a distance, as seen in the video above the cliffs, I included some empty areas with no grass in the video to compare the two when close up, if the grass gradually decreased in height in these points it would be very hard to notice.

 

 

 

 

 

 

 

 

Along with making grass early on, I also made some mushrooms. I modelled them from a cylinder in Maya, I cleaned up triangles even though they were going into a game engine to be triangulated, wasn’t sure if I had to but I guess it’s still good practice maybe. For textures I eventually used my smart material which provided a pretty solid base, then pushed the highlights and ambient occlusion using dirt and metal edge generators, only taking the colour information and applying a blur slope filter. I then used a curvature generator to add the emissive elements. One thing that caused problems I had in my last project was with the lightmap density, so I worked on this early on so I didn’t run into any issues, the one on the left in the last picture is what I managed to get.

 

 

 

 

 

The last image is a screencap of them in game, while I like how they came out I think they could fit in a little better, I’m guessing it’s just since they were made so early on in production, they don’t hit all the beats of the style we ended up going with, but I didn’t have time to come back to them or play with their textures anymore. I do like how they look but there’s something missing.

 

I loved the materials and their setup shown in this video (stylized station again). Substance Designer is very powerful, I didn’t end up using it since I struggled to figure it out but some of the ideas from this video definitely translated, while none of my materials ended up similar to his,  some material set up ideas did and the vertex painting he shows makes an appearance later on in the project when I was working with Ben’s cliff’s. Another great video I found with really nice landscape materials were from this youtuber, who doesn’t have any breakdowns but has some really pretty environments. https://www.youtube.com/channel/UCV2yUfvVDuJtEDMDzEgkQ0g

The Dreamscape packs in the Unreal Marketplace also had some very nice environments, but you have to buy their packs. https://www.unrealengine.com/marketplace/en-US/product/dreamscape-nature-mountains They were great inspirations though.

 

 

 

 

 

 

 

My first few attempts at creating seamless textures didn’t work out well, I thought they were seamless when making them but obviously not, definitely took a bit of trial and error to get something that worked well, and I did change my approach up quite a bit after figuring it out. I feel like I really captured the aesthetic of the first image with the two photoshop files though, but changed what I was going for with the textures before the final landscape material was made.

 

 

 

 

 

You basically had to offset the image so the corners would be at the middle, then make it manually seamless that way, one method I found useful was using the content aware fill as this would take information from there you highlight and it somehow fills in the gaps, a really powerful tool in Photoshop. I know (if I understood the software) I could’ve made a much more efficient and probably better looking material in Substance Designer but the landscape material was made when I had a lot of work going on and Photoshop felt safe.

 

 

 

 

 

The last three images are my final landscape material set up and the textures I needed for it, seamless textures created in black and white in Photoshop, allowing me to make them any colour in Unreal Engine. To create Normals from this I used this tutorial: https://www.youtube.com/watch?v=YJqWHsllczY&ab_channel=SaeedMandegarTutorials I wanted to learn Substance Designer but I barely scratched the surface with it, I am much more comfortable in Photoshop, especially since I didn’t have a lot of time to learn a new software at this point. In order to add further variation I used a texture file within Unreal (macrovariation) which would blend overtop what was painted.

Despite being pretty simple textures I think they served their purpose really well, after making the grass, the sand/dirt was simple since there was much more empty space. In retrospect I do think I lost a little bit of the hand painted touch I had in the first few attempts, probably because there were weeks between the different landscape versions with many different attempts. I did learn alot when it came to this material in particular though, and think the way its setup makes it very intuitive when editing the parameters, allowing colour and tiling customisation for both the standard texture and the variation overtop.

 

I’m really happy with how this material came out and the way I structured the variations let you actually paint if you lower the opacity, creating really pretty and natural gradients. I think the grass and dirt fit well in with the environment and isn’t too simple, but not too realistic to no longer fit the style. The macro and micro variation i set up within the material adds a nice painted effect as seen in Ghibli environments, with the darker blues etc. – while also helping to make the texture a little more interesting.

 

Art direction: Texture guide and Smart material, ended up used on majority of assets

 

 

 

 

We had a few comments regarding the consistency of our assets being made very early on so I thought making a guide when it came to texturing would be a good idea. I developed a smart material that would create a hand painted effect with some normal stylized rendering techniques like gradients to guide the eye upwards, this was all procedural and only required a decent bake. The first image is a file I typed up based on my research, breaking down the different aspects of the style as well as how to achieve this within Substance. Then you can see how the material looks on a rock asset, as well as the example JadeToad that comes with Substance Painter.

 

This is how I set up the smart material, as you can see its all based on procedurals and generators meaning it would apply to any model based on its baked information. It definitely isn’t a final look, more a base to work from since trees would be reacting differently to light than rocks, so you would push highlights further/less etc. I hoped it would keep a level of consistency among all our environmental assets. These are all based on techniques I use normally when texturing, making heavy use of folders with masks, as well as when going for this hand painted look which would normally be used in conjunction with actual hand painting for further detail – this non destructive and procedural workflow made it easy to transition into making it a smart material that could be easily manipulated to change colours and push aspects.

These are few of Ben McCullough’s models with my smart material applied to them, I think they worked really well with his  work since it would often be modelled well/baked from a sculpt so the smart material had alot of information to work with. (Ben’s assets !!!)

 

 

 

 

 

 

 

FaultE

This is Jordan’s concept work as we neared a more final design for FaultE. The group were wanting the model to be done much earlier than when design was finalised, so I had to start modelling without concept, then merge my work with his design once I received it. I do wish I had a final concept to work from from the start because now when you look at the concept art, certain aspects aren’t in the final model. This is also because there were discussions between me and Jordan involving the cloth aspects, leading them to be much more simplified. We were worried about how we would handle the cloth and didn’t want to keyframe animate it, so we kept it as a cape not interacting really with his shoulder, so there would be minimal complications with cloth simulation, while also letting us have the option to easily hand animate if our plans to simulate didn’t pan through.

 

 

 

 

I lost my documentation screenshots but I was able to salvage some obs footage while I modelled. I just modelled it in pieces, before working from Jordan’s concept I used my Knight armour set from last module to estimate proportions. There were alot of great references on sketchfab too when it came to modelling a robot, I wanted it to seem like it was a real robot but I wasn’t 100% on how the machinery and pivots would ‘function’ but they helped me figure out the leg and elbow/shoulder joints. You can see in the video there was a little bit of experimentation required for the ‘bicep’ and the legs before figuring them out. I was trying to focus on having clean topology throughout so I could hopefully use this as a portfolio piece, and I think it came out pretty nice.

 

 

 

 

 

This is my collection on sketchfab for this project, it had a lot of robotic stuff but also some environmental assets and creatures. Not 100% sure what in particular I was looking at for topology references or as inspiration but whatever I have used is definitely in there https://sketchfab.com/matthewshannon/collections/coollg

 

 

 

 

 

 

 

 

Some of my research into cloth simulation and approaches to model clothes and cloth assets, as well as then my own (early) experimentation in Marvellous Designer in creating a cloth poncho/cloak. Learning about the patterns was quite fun but I don’t think I scratched the surface, I was going to approach some fashion design students regarding patterns but decided to drop using Marvellous eventually, which I detail below.

 

I first was using Marvelous Designer to try and make the clothing but the effect was way too realistic and didn’t fit well with what we wanted, I do think I’ll come back to the software in the future because it’s really cool and the cloth simulation in it is very pretty. We were considering using it’s cloth simulation on our animations baked as an alembic file but this is very performance heavy, better for cinematics, and wouldn’t respond to player inputs like rotation, so I was researching Unreal Cloth sim at this point. To model the clothes however I ended up using ncloth simulation in Maya, then adjusting this using the soft selection tool to get less realistic folds, building a stylised scarf and cape. I think it came out really well and the scarf is almost 1:1 from the concept art, we just changed the cape as discussed. I also added a few cables to the head above the right shoulder for some more asymmetry and to better fit the concept, the last clip in Unreal engine looked great so I was happy with the model. I then got started on the rig, since (while extremely limited and with alot of help) I was the only person that had rigged previously.

 

 

 

 

 

There were plenty of resources given to use for rigging, skin weight painting from alec and some general rigging theory. The most useful video to me was probably the series from Mike, rigging a humanoid character from start to finish. I pretty much followed this for creating my basic controls and skeleton, while I didn’t fully understand the orientation of joints part I just copied what Mike had and eventually figured it out. Creating the Null groups and naming everything was probably the most tedious part, while the align script given to us by Mike made this easier, it stopped working for me after a while I had to do it align manually. The theory behind what Mike was doing all made sense and working through the videos, creating a character alongside him, was very helpful. Since I had something a little more complex with my character going on though, I had to figure things out by myself.

 

 

My main struggle came from figuring out how to make an IK/FK switch. With what I now understand this is actually pretty easily set up, but for some reason I thought it should have automatic IK/FK matching while swapping between the two versions, which caused a lot of issues. Once I stopped trying this however it all worked fine, I just had to figure out a way to keep the hand controls visible when swapping between modes. Again, this is a pretty simple fix, either having one control or, instead of having the controls grouped under the wrist, they would be constrained to both the IK and FK wrist, with their influences being controlled via the connection editor connected to the IK/FK switch. Instead of doing this to every control, I added a locator that would control them all (acting as either the ik/fk wrist).

 

 

 

 

 

 

 

This details my full IK/FK setup once I got it figured out. There are 3 arms (IK, FK, Bound). The 2 modes would be constrained to the bound arm, with their influences (in channel box) being controlled by the IK/FK switch custom attributes I added (IK,FK) using the connection editor. I used the expression editor to have a simple line of code, saying if IK equals 1 in this switch, then FK will be 0. I then hid the FK attribute in the switch to simplify. Then, to make it nicer to use, I set their visibility to be influenced by the switch, hiding the FK switches in IK mode and vice versa – this was done using set driven key in the animation tab. This same general method was used on the locator influencing the hand controls mentioned above.

I wouldn’t have figured this all out without some help from Daryl, who sent me over a robot rig he made that had an IK/FK switch. This, along with many YouTube videos, feedback from Mike and this video in particular https://www.youtube.com/watch?v=8HX7Np8eeo8&ab_channel=StanAbraham all helped me get my head around it.

 

 

Rig came out great and Jordan said he had a good experience when animating with it, so I’m pretty proud of it, especially for my first real rig and for figuring out that FK/IK switch. It isn’t the most complex rig, as seen in the video its a fairly standard character, but it served its purpose well and we never ran into any issues.

We both decided we really didn’t want to hand animate cloth and thought simulating would work best in game, so I looked into our options. I was planning on Unreal Cloth sim, and if this didn’t pan out my back up plan was to bake the deformations from a cloth simulation (maya or marvellous) to a dynamic joint system within Maya, which would then drive the animation on the cloak in game.

 

 

 

 

 

Many tutorials for cloth didn’t involve character’s, but I managed to find a few videos going into a bit more detail. The videos were very long and experimental though, so it was a lot of information to navigate – they helped me get the core setup down though.

 

 

 

 

 

Cloth simulation wasn’t that complicated once I figured it out, its just very heavily under documented for Unreal, and it’s not the prettiest simulation out there. You need a separate uv map for the areas that will be simulated. Then you basically paint the influence, similar to painting skin weights. Next you set up the fake collision, since it doesn’t collide with the actual mesh in game, but instead simplified capsules. You can edit a few settings to create slightly different simulations, but it isn’t that powerful in this version. I found having the influence of simulation just below 1 the best for keeping the stylised shape of the cape. It does get a little more detailed when going into the cloth config, this video https://www.youtube.com/watch?v=NIwU2WJeco0&ab_channel=JohnConnor breaks it down pretty well, but you mostly influence the stiffness as well as self collision, with the self collision not being amazing most of the time – its kind of hit or miss. The base cloth sim isn’t great but with a few tweaks I managed to get something the group were happy with.

I feel like I could’ve maybe spent some more time going into the config and figuring out how to get it behaving even more realistic, but the group were happy with what I achieved and this was more of an accessory to the game anyways, so since I had more important stuff to work on I left it.

 

 

I think for real time simulating this is pretty effective though. It doesn’t behave 100% accurate when jumping but the rotations and collision all look great. I would’ve liked more time to mess about with the self collision though and try and maybe make it a little less elastic. One downside that is with lower fps in game, which my pc would have in our level, it tends to break or not function as well since its calculated in real time.

 

 

 

 

 

Texturing was done with the same methods as before, detailed in the texture direction file. I used my smart material as a base and used masks to have different coloured areas, mainly red, blue and a darker deeper blue. There is also a lighter shade on the plated areas but this is subtle, as well as brown for dirt. I added some details with normal information like the stencil on his shoulder and fabric to his scarf and cape (separate file I can no longer find). To keep consistency I created a sperate smart material from his scarf and used this on the cape too. One effect that created a nice subtle fabric detail was the watercolorfx in substance painter, adding normal information where the water colour would “clot on the paper”. I also added emissiveness to his eye and hand.

 

Happy enough with the textures, I was going to add more obvious differences in the shades of blue for his armour, metallic gears and plating, but kept it subtle since a lot of our assets so far where kind of sticking to one colour. The dirt on the feet and legs, as well as in any cavities adds a subtle touch which looks better while in engine. The paint chippings and metal edges looked pretty nice too.

 

 

This is the final version of FaultE, the model and the rig in Maya. I think it came out great and I’m really happy with the rig too, managing to figure out an ik/fk switch that was simple to use and worked well, I went in having no idea how to rig and, while I don’t really love doing it, I feel comfortable creating custom rigs for character’s that are also user friendly to animators other than myself.

 

Carniviper and Evil Mushroom: Remodel/Retop and Rigs

 

 

 

 

 

 

Due to some other group member’s changing what they were working on, Gina (game design) ended up making the organic creatures. The models came out cool but had some topology issues, I’m guessing this wasn’t covered much in the game design course. Since I was rigging the character’s I was asked to also take over the model to make it better suited for animation. I wanted to keep the design as close to hers as possible so she still had her creative input, but I had to remodel some aspects completely like the legs to be modelled with animation and rigging in mind. The first image shows what I was sent (left) and my retopology/ remodel (right).

 

 

The rig was pretty simple, just rigging some IK legs, and a few joints to influence the mouth and spine/ mushroom flaps, then painting skin weights. I like how it came out and the rig was easy enough to use, facilitating all animations we needed to make.

 

 

 

 

 

 

This model was sent a little unfinished since Gina was struggling with the topology again and thought it would be better if someone from animation took over. The first picture shows what I was sent and what I ended up with after remodelling. Again I wanted to keep it as similar to what Gina was going for so she still had her contribution. This rig was a little more experimental than the above mushroom, since I thought hand animating a slither wouldn’t be amazing.

 

 

 

 

 

This rig was set up as an IK Spline, I then created clusters for each vertex control point which would be influenced by our controls (just parented beneath them) – they are also constrained (their null groups) to the yellow control, made to influence height easily in both automated slither and manual animation mode. To create the automated slither effect, I used a nonlinear deformer on these controls with a sine control. Manipulating this sine’s attributes would allow us to influence the slither, changing the wavelength. This could just be assigned to the mesh, but since we want to export to a game engine I needed a rig to get the animation in game. https://www.youtube.com/watch?v=58F15mLA6Uo&ab_channel=CharlieVelazquez this video outlines this technique in detail.

To make it more user friendly, I added a control using similar logic to the IK control I made for FaultE, which was connected (via connection editor) to this sine and its attributes. Activating automated slither mode, then moving the slither along the snake would create this animation effect for you.

 

 

I enjoyed making this rig, it was a bit more experimental but I think it gives a really effective technique for in place snake animations, and made animation much less time expensive. The controls are pretty intuitive but if I had more time I would eliminate the yellow and red controls, having one that would influence both, I was the only person using this rig though so I didn’t mind. I love the slither control and how I set that up though.

 

Animations

 

 

This is the final version of FaultE’s animations in game that I worked on. I had 8 walks (when camera orientation was locked, like when using beam), 3 animations for climb (idle, climb, root animation at top), 3 animations for the beam (start, loop, end), 1 scan animation and 2 for pushing (idle and moving). I worked on these in Maya then just exported the skeleton, I only really had one pass of each animation due to the workload we had to get through, so it was definitely quantity over quality. I know I can achieve a much higher standard of body mechanics in many of these animations but I was told to just pump out animations and not to be perfectionist with them, since we had alot of animations required by game design.

3 climb anims

 

The final mantle animation at top was our first root animation, done day of submission since we were told we needed it to progress through the level, not really happy with the weight at the top or the awkward hand movement at the end of it but it is what it is. The blendspace blends the idle and the moving well though, I like how they came out and the climbing has a nice loop.

 

8 walks and 3 beams (blended overtop)

 

Pretty content with how these turned out, obviously not perfect. Backwards and forwards walk are the strongest with the strafe being a little awkward, again we only got one pass at these animations really in order to stay on top of the workload. This sounds like a lot of work (8 looping animations) but once I had the forwards, backwards and strafes, I went back and manipulated those via graphs and some clean up to create the others. The 3 beam animations were required since we don’t know how long the player would use the beam, so one for the start, middle(looping) and end.

 

2 pushes

 

They came out pretty good too, the push has a nice weight to it and I like how the arm extends to the rock to push it. The collision is faked with the rock since its an in place animation, so any clipping is dependant on the asset, just how the push mechanic was set up.

 

Scans

 

Scan works well and slowing the movement speed with it feels cool in game, I like how you can scan while walking, or scan while standing still, starting to walk while scanning works well since I played it overtop our normal walking blendspace.

 

 

This is the final version of FaultE’s animations in game that I worked on, one by one in Unreal. I am really happy with the quantity I managed to make, but there are definitely improvements to be made on the majority of the animations. They were ‘placeholders’ that I never got time to replace with the finalised versions, so they all had one pass at animation.

 

 

This is the final version of Evil mushroom and Carvniviper animations I made in game, just an idle and a walk animation for both so only 4 animations total. I really like how goofy the mushroom walk cycle is and the snake slither looks well too. (I had to go into unlit mode because my pc struggles with fps)

 

 

I also had to set up all the logic and the states/montages responsible for the animation, which was alot to learn and take on on top of everything else. I definitely am interested in it now, despite hating every minute of creating this. It got pretty complicated with how I had to set certain things, such as the timed idle states leading into idle actions or the montages that would have different slots (in skeleton) to separate the upper and lower body, overlaying animations. This video was used initially https://www.youtube.com/watch?v=1K-Hyu4Xn3g&ab_channel=MattAspland

As it got more complicated, separating upper from lower body, having timed idle states, playing montages and root animations etc. I relied on Prismatica Dev’s Advanced Animation Theory UE4 course/series? linked here https://www.youtube.com/watch?v=flHL3qJB3_I&ab_channel=PrismaticaDev with a lot of videos going into detail with topics like; root motion basics, blendspaces, montages, layerd blend per bone, anim states, anim bp and additive animations.

 

Particle effects (smoke, spores, falling leaves, swamp bubbles, scan, beam, thrusters)

 

Last module I enjoyed working on Niagara particle effects in Unreal, they aren’t as scary as real particle effects and its still a really powerful system. this video shows some of the stuff I worked on, they were a lot of fun to play around with whenever I had free time and added a lot (I think) to the game, even the subtle ones like the thrusters made that double jump animation more interesting.

For the smoke I used a tutorial, incorporating flipbook animation from present files within Unreal. https://www.youtube.com/watch?v=tTTL_bzQLuY&ab_channel=Sir_FansiGamedev . The other effects were done with the knowledge I learned last year when making dust and flames. The beam when picking up objects took a little more effort, having to add a custom parameter for the end location, which would be set to the object being picked up on an event tick.

 

 

 

 

This is some of the scripting I set up for the particles, spawning on child actors (last pic) that I placed on our character in different spots. In order to prevent them from being spammed, I had to tie them to variables for events, such as when jump value is 2 or when hand scan is true. This also has some of the scripting set up for the animation montages that blended upper body and lower body, since they were tied to the same events. I normally hopped in a call with Scott to help figure out any problems I was having with scripting.

 

 

These are the different systems I created, the beam and scan use the same set up. The smoke uses a flipbook animation within the system. The bubbles, thruster, smoke and falling leaves use similar set up, with added effects like gravity or different spawn rate and locations, or different meshes. The meshes I had to make were very simple, just a leaf model similar to Jordan’s leaves and a sphere for the other emitters.

 

Some Level Design (for our playtests and final lighting and atmospheric fog in mainlevel)

 

I had to take this over while Sam was away for a while, just for the first two playtests. My playtest levels looked okay (2nd one was in collaboration with scott) but didn’t have the theory that game design students do, so the common feedback was they looked nice but lacked playability/gameplay. Sam came back and she took over from me, but I came back at the end to try and pretty some things up with lighting, fog and godrays in our final level too. Godrays are cool and fog adds nice depth to the forest, making our final level look much nicer than it did (at end of video ^).

 

 

Lighting/fog before and after

 

 

 

Lighting/fog before and after

 

 

 

 

 

 

 

 

 

 

Wasn’t 100% on what to do with the lighting of the environment but it definitely needed changed from what we had previously. I think I got some really cool effects with the god rays coming though the trees and past the asteroids though, and the scene felt more saturated and nice to walk through. The group all seemed to think it was a big improvement too. Really like the 4th screenshot and the lighting coming over that small cliff.

 

 

Vertex Painting Material setup

 

I also set Ben McC’s cliff material up with my grass texture that we painted the landscape with. Ben had the tops painted green but we thought it would blend better with the landscape if we could have the same material on top. I set up vertex painting, which influenced the alpha channel of lerp nodes, switching between bens cliff material and my grass material, allowing the level designer to paint this to better blend the cliff with the landscape. https://www.youtube.com/watch?v=Lz_tqHAS-Kk&ab_channel=3DAssetLibrary

 

Texturing Posi

 

We were getting ready to submit and still hadn’t received Posi’s final textures so I also quickly threw this together, similar texturing workflow to FaultE. I added normal details for nuts/screws on certain panels as well as the same symbol featured on FaultE’s shoulder, to add some connection between the two. The green emissive ‘eye’ was added to try to incorporate something similar to Scott’s UI when Posi followed you in level and to better link with FaultE, this would switch off when the actual UI appeared.

 

 

 

 

 

 

Overall super happy with the work I done, especially the landscape materials, grass and character. I can now also say I am able to set up an animation blueprint with many different states, state machines within states, blendspaces and layered blending through animation montages. I’m happy with the rigs but I would’ve like to clean up all my animations, quantity was prioritised over quality. The group work was pretty good (disregarding the members who didn’t contribute work or communication on both the animation and game design sides) but I wish our final game was a bit more impressive and had some more gameplay, since we put in a lot of work. This project wasn’t great for my portfolio, hoping to focus on solid animation and organic modelling, but modelling a hard surface character and rigging were still useful skills to practice. I did get to learn more about Unreal materials which I wanted to do, and got to play with particle effects which is always fun. The (limited) scripting I did was pretty stressful but I started getting the hang off it by the end and the two scripters in game design were always helpful. If we didn’t have some group members not contributing, we could’ve ended up with a really solid final game because we made something pretty decent with the smaller group, especially from the art side, making some really cool assets and materials.

Personal Development

by

In support of my portfolio / showreel, I animated a 3D scene to continue my personal practice & development. I animated a 3D shot in Maya focusing on dialogue and acting, for my last project I used a Miles Morales rig from the Agora Community, so I went there first to look for this project. I wasn’t sure what audio I wanted to animate, but I thought choosing a rig first would help narrow it down.

 

 

 

 

 

I found a Thor rig which seemed pretty cool, not too simple so I’d get experience with little bit more of a complex rig. I set up MGpicker, a very intuitive and animator friendly Maya picker tool, with the a powerful feature that enables you to create your own customised picker without any prior knowledge in coding and hooked this up to the Thor rig with the file included by Agora. I tested some facial expressions and liked the range I could achieve with the rig.

 

 

 

 

 

I also looked at an Aang rig, which was a little more simple and was advertised by Agora as being for facial animation. I attempted to recreate the same expression as before to see how I liked it. I didn’t like the blue eyes in the Thor rig, these eyes were much more expressive.

 

 

 

 

 

Aang had a lot more automated features in the rig, with sliders allowing you to pucker the lips or puff the cheeks out. It also had a studio library file included, with a bunch of pre-set facial expressions already set up for you. I decided to go for the Thor rig since I thought it would be more of a challenge, since it didn’t have any of this automation. Also because I haven’t seen it used much online in other showreels, so it’s a little bit more original.

 

 

 

 

 

I also noticed the Thor rig was game-ready so I tested it out in Unreal, I thought it might’ve been cool to render this shot in Unreal for my showreel, since it wouldn’t take long and you can make a pretty enough scene easily. I built a small scene using assets from the Vertical Slice project and brought Thor into Unreal, but I noticed the UV seams were pretty visible in Unreal and thought. if this was going in my portfolio, having scenes from my Vertical Slice project and my Lip-sync animation in the same environment might be a bit repetitive, so I decided against rendering in Unreal.

 

 

 

 

 

Sir Wade had a lot of useful videos online, going through how to find an audio and what to avoid/look for. His ‘secret’ workflow when it came to dialogue shots and lip-sync tips, like acting towards the camera, having the mouth shape before the audio happens (light is faster than sound) and having asymmetry/imperfections.

 

 

 

 

I found an interview Sir Wade did with the head character animator on How to Train Your Dragon, it gave an insight into acting through animation, his workflow and general tips on animating. This video from James Baxter (worked on a lot of Disney titles) was more useful and much shorter, mostly going over the mouth opening and closing as well as the timing for this, when to hold and when to close suddenly etc. in order to avoid a boring linear up and down. He also simplified the mouth shapes and combined shapes into one, so I guess you can get away with combining sounds/mouth shapes especially when the character is talking fast or slurring words, limiting these mouth shapes migth be easier in 2D animation though which is what he was covering. I also looked at a live action analysis of the over the shoulder shot, which I though would be useful to me since I had two characters talking but not enough time to animate both of them fully.

 

 

 

 

I also collected some material to look at in terms of the actual mouth shapes for sounds, finding a YouTube video breaking it down a bit and some pictures I could look at while working.

 

 

I watched a lot of interviews and clips from Chris Hemsworth since I thought it would be cool to have Thor’s actual voice in the dialogue, but not just taking from a scene in the MCU, since it would be recognizable and the watcher might compare my shot to that scene. Sir Wade mentioned have some sort of change in emotion/tempo or interesting sounds in the audio and I thought this clip was good work with and had an interesting popping sound in it.

 

 

 

 

 

I found an interview with a Dreamwork’s animator that was more focused on the lip-sync and acting than the previous interview. He went into this idea of subtext, going beyond just animating what is said and going into the thought process of the character, a useful exercise he mentioned was animating the same character and audio in 6 different ways (different emotions, gestures, emphasis etc.) I obviously don’t have time for this but it was useful just to think about. I took some notes similar to how he did in his example to help inform my animation, typing out a simple script and what I thought the character of Thor would be thinking. I also checked out twitter to see how other animators approach this, alot using reference.

 

 

 

 

 

 

 

I did some further research into the two rigs for examples of dialogue shots and manged to find some on YouTube and Art station. I really liked the first one, showing the capabilities of the Thor rig and how much personality can be added.  I wanted to do something similar with the style of my animation, the cartoony/stylised 3D style in these clips were very engaging. From comparing the first and second rig, I thought the eyes made a big difference on the animations, with the first one (edited from original rig) being more appealing.

 

 

 

 

 

 

 

 

 

I recorded some reference footage (bunch of clips on timeline, maybe 12 takes) and reviewed what I liked and didn’t like about them. I had the idea to have the pop sound be from him removing his finger from his mouth after eating, seeing something similar in Spider-Verse. I used this as reference for the style, as well as when building my little scene too. I also changed the eyes, just applying different lambert materials to parts of the eye to humanise the character a bit.

 

 

Some embarrassing reference footage, while I didn’t like the acting much in this one (kind of boring) the timing and actions felt nice and I had a bunch of other clips I could reference that had more personality in them.

 

 

 

 

 

 

 

 

I started blocking out the scene and testing camera angles, if I established the framing it would help with the rest of the animation, since I could kind of cheat, animating towards the camera. I wanted to focus on the pop first since it was the main reason for choosing this audio.

 

I went straight in on the pop, blocking out the main poses and timing for it. I was happy with how it was looking already, it felt very satisfying and had a bunch of personality with the cocky/playful eyeroll.

 

 

 

 

 

Then I blocked out the poses leading up to the pop, where he would bite the food, I liked pushing the anticipation for the bite, it was pretty exaggerated but still felt believable in the animation. I’d push this a bit further later on too when adding more squash and stretch to the head. Animating with the animpicker was very useful since I could animate with no controls on screen.

 

 

 

 

 

 

 

 

 

Next I went in and added all the key poses I wanted to hit throughout the animation, I was worried about the flexing pose, and the pose after it, it felt like a lot of movement for this scene, but I didn’t want to stay with the same silhouette throughout, since he was already resting on a table so had a restricted enough range.

 

 

This is the rough blockout at the end, once I was happy with the pop. I think it is a little awkward but it felt nice enough to move on and the individual poses were quite strong and fit the audio.

 

 

Taking from my reference footage, I added the head shakes/nods throughout and some follow through/anticipation for the poses, just cleaning up parts and adding keyframes throughout, trying to be mindful of secondary actions and offsetting a lot of movement, since Sir Wade mentioned asymmetry and imperfection in his video.

 

 

 

 

 

Looking at the 11secondclub, some online livestreams focused on lip-syncing and lip-sync tutorials in Maya helped refresh me on animating the lip movements in my animation.

 

This was the first pass(?) of lip-sync, I tried to simplify and combine shapes as much as possible to see what I could get away with, so the mouth wouldn’t just be going up and down and he wouldn’t be overenunciating. I needed to fix the timing towards the end and add more mouth shapes. The ending definitely had some questionable timing and needed more keyframes.

 

 

 

 

I noticed my graph was looking ugly so I used auto tangents, and spline tangents on certain peaks, adjusting where I though necessary. I also played around with the timing using the dope sheet but I’m still not 100% from when he says arrows onwards so I had a bunch of iterations of the file, letting me experiment without losing previous progress.

 

 

 

 

 

 

 

 

 

 

 

 

I then went into the details, adding finger animation and making adjustments on these grey controls to add ‘contact’ between the table and the fingers and elbows. I tried to clean up the graphs but it was getting confusing with the amount of keyframes, since I had a lot of offset happening and probably didn’t have the cleanest workflow over these few days. I decided to bring the Aang character in and animate his “lip-sync” which was just vague mouth movements since it was an over the shoulder shot. I thought he was too bright and took away from Thor so I adjusted his materials exposure levels.

 

 

This is showing the collision I faked with the table, as well as the little ‘set’ I built. I had some difficulties with the Aang rig but managed to get it in by copy and pasting it from another maya file.

 

I added some camera animation, with subtle rotations and very slight zoom in to give it a more natural feel, as well as adding depth of field. I thought the depth of field would help take away focus from Aang since I wanted the main focus to be on Thor, especially because I spent only a few minutes on Aang’s ‘animation’.

 

 

 

I’m quite happy with how it turned out, there’s some things that could definitely be improved but I rushed through this and have spent too long looking at this animation now to progress much more. It’s similar in style to the stuff I was looking at, and definitely can see the spiderverse inspiration just in the framing and sucking ketchup from his finger,  it’s not too derivative, I tried to keep it dissimilar in the rest of the scene, adding much more energy and movement. I’m not 100% on Aang being added, it feels a little weird and think it might be better without but I wanted the over the shoulder shot, hopefully people don’t actually look at Aang’s animation though. There’s some moments in Thor’s animation that are a little awkward but I’m happy enough, I might’ve pushed the fingers a little too much though, and had moments where they seemed quite stiff and lifeliss. I forgot to document but I added a sphere to his finger with a transparent blinn material to make ketchup, which had a null group constrained to the finger control, then I just keyframed it to shrink and hide once it went in his mouth. It definitely would’ve been useful to send it to one of the members of staff for some feedback but it was so rushed in the last few days I didn’t really have time to.

 

 

 

 

Animations – Portfolio

by

 

 

 

 

 

Animations

by

Body Mechanics

I did a whole bunch of tasks before starting any of the animations but it was on campus and I didn’t think to keep a record of their development for the blog. They were pretty simple back to basics stuff, covering the workspace and tools in Maya, and some animation principles. The first task was to block out a ball’s movement using the stepped mode in Maya changing the spacing to achieve an ease in and out. The second task was to spline a blockout using the animation graph editor. We had a task focusing on arcs, animating a head turning. We animated an arm to distinguish between FK and IK, as well as practice anticipation and follow through. There was a lot of stuff we covered/ knew already but it was a useful refresher since I hadn’t used Maya in a while.

 

 

 

 

We had a class focused on posing which I managed to recover, I used the Jill rig and tried a pose from Gwen in Spiderverse, I’m a big fan of the poses in this movie. The Jill rig’s proportions are a little different to Gwen’s but looking back on this I think I achieved something similar but there’s definitely room for improvement, the bent leg’s calf is phasing into the hamstring and the arms/hands are a little off. Readability is very important in animation so strong poses always help.

 

 

Managed to find the boxing exercise too, I think this is an older version though since it is pretty unfinished, this was to practice timing and adding in-betweens I believe, just to get us animating and get us used to the dope sheet again.

 

The first thing I did towards my body mechanics animation was collecting poses I liked or wanted to use, since Spider-man is known for strong poses there was a lot of great references to use, I had a rough idea of the actions I wanted to do so I collected some running, landings, jumps and vaults – these came from fan arts, comics, animations. I also collected a whole lot of footage I could use as reference, spending 3 hours in class a little distracted by Spiderman movies, animations and games.

 

 

 

 

 

 

Research into body mechanic animation techniques/tips. Focusing on head screen space over the root, having C shapes in spine for dynamic poses, centre of gravity etc. The breakdowns of body mechanics animations to see these techniques in practice was helpful in understanding what makes good/bad bodymechanics.

 

 

There’s too many links to share them all its ridiculous, but here’s a quick flythrough of just my currently opened tabs . Kinda went overboard, probably saved around 50 videos of Spiderman just on my likes on twitter, some of the animations guys I really liked though were; Sarp Serter, Marin Monserand, Arturo Garcia, Pierrick Picaut, BBWB Anim, Arsal and there were alot of great compilations on the Anim Challenge Youtube. I referenced Arsal, Arturo and BBWB Anim the most for the specific movements probably, but tried to use the stylised animation techniques I saw from Pierrick Picaut (no spider-man animations from him, just cool animations and courses).

https://www.artstation.com/artwork/DxY3YG

https://www.instagram.com/bbwb_anim/

https://vimeo.com/agvanimation

https://www.artstation.com/pieriko

 

 

I kinda struggle with imagining ‘choreography’ so I built out a little block city to figure it out. This was the night before class so I didn’t have a whole lot of time to use a character for the blockout, so I used a ball just to show the tutors my idea. I imagined him jumping from the first two building and landing on the third for the close up, then we’d have a good side view showing off the run and jump animation, followed by the vault over the obstacles and dive from the building.

I got some feedback from both Alec and Daryl here, mostly on camera movement and timing. I changed the camera at the start to have less movement and be a bit slower, as well as rotate upwards to follow the jump, instead of jumping the camera up with it. I also adjusted the speed of the second jump. Daryl said to imagine this in real life for how to film it, you’d probably be using a helicopter or crane – this helped me to think of the camera animation a bit better. He also set me up with the film gate/resolution gate as the viewport isn’t what we’d see in the play blast.

 

 

 

 

I saw one of the previous students used a rig from the Agora Community, so I checked this out too. I wanted to do some sort of cartoony action animation, something Spider-Verse related with the super stylised movements so I used this Miles Morales rig on Agora. It took a bit of setup, installing a plugin called mGear and changing some file paths to get it to work. I started by exploring the rig which was very complicated in comparison to the Jack and Jill rigs we worked with in class, there were so many controls. I stuck with it though – I thought it would be good to get experience using a more complex rig and provide variety between my animations.

https://agora.community/content/spider-verse#!

This video helped me with exploring the rig and finding out what all the different controls can do.

 

 

 

 

 

I wasn’t able to reference the rig into my previous blockout I was working on and I had to actually work on the rig file itself as I couldn’t import it to a new scene either, so I started fresh, the first thing I animated was a jump since my blockout had him jumping building to building. I was originally planning on recreating the block out since the timing and layout was good but I just focused on the animation itself and used that blockout as a vague guide.

 

 

 

 

 

I watched so many Spider-Man clips and fan animations to get some inspiration and really liked this fan animation, especially the extreme stretch during his landing here, so I did something similar. I went straight from this extreme stretch to an extreme squash since it was a big landing – I think it came out well its definitely not realistic but I like the exaggerated effect it gives. (following feedback this is changed toward the end, I guess it didn’t work in 3D as well as I thought lol)

 

 

I recorded some of my process, I was focusing on strong poses and getting all the action blocked out, I made sure to be considerate of the arcs of each movement since without strong arcs it will look too floaty and linear. The poses could definitely be pushed more but it was beginning to take shape. I did play with the timing and spacing a little too but I wasn’t too focused on it at this point.

 

 

I had a viewport that I would switch to to make sure I had a nice strong and readable silhouette in each frame. Without clarity in your posing, you can’t have clarity in movement. Having a strong silhouette is the best way to ensure your poses are clear. Spider verse had very strong and exaggerated poses in almost every frame so I tried to replicate this as best as I could. I would later come back to the posing to polish them all up.

 

 

 

 

 

I tried to be mindful of the legs and arms being right in each ‘run’ but at some point I did mix them up, having the right arm forward with the right leg and vice versa. That’s not pictured here, I’m not sure when it happened but it took me a while to notice since it was so easy to do with this model and I had to go back and redo a lot of work.

I noticed some jumps (mostly the falling part) looked a bit floaty and empty, so I’d animate the arms and legs to come up a bit as he falls and wave a little, since the air resistance would be pushing them up.

https://youtu.be/i0i36AlJciI – cloth is really good for this effect, but the same principles apply to loose limbs.

 

 

It’s important to constantly be mindful of momentum in animation otherwise it will start to look wrong pretty quickly, especially with a character like Spider-Man, who goes from running to flipping to swinging, I also tried to make the transition between movements seem as natural as possible, so I would have some weird poses halfway in-between a jump/crawl and a run, depending on how fast he was coming into these movements.

The second half of that video is more or less what I sent to Alec for some feedback, he liked the posing, the second jump and the snappy landing at the end, but noticed some issues with the timing and arcs of the first two jumps. He also pointed out some poses I could work on, like on the taller rooftop he turned very quickly over few frames.

 

I took on all this feedback –  used the dope sheet to play around with the timings a bit and the graph editor to clean up the arcs. Overall I worked on cleaning up what poses I had since I sent it off very rushed, as well as the poses he pointed out to me. I also added a bit more to the action with a sort of scramble from the floor into a run with very long strides, preparing to jump from this building. I’m happy with the run but it looks a little unnatural – but I’d come back to it later. I left this animation for a few days after improving it based on Alec’s feedback – I though this was a good point to start the run and walk cycles, since I had very little time left already since I started this assignment so late.

 

 

I’m a big fan of the animations directed by Alberto Mielgo as well as his storyboards, the characters always seem off balance and stumbling/ slipping during action scenes, nothings perfect and its very messy, it adds much more character and intensity to the action and even realism, I first thought a lot was motion captured or even rotoscoped (the witness) but its all normal keyframed animation. I especially focused on his short for Watchdogs (2nd video), I love the movement in this. I tried to incorporate some of this into my animation with the sloppy chaotic run/scramble from the tallest rooftop into a jump, and the stumble onto his hands when he hops of the ledge from the close up shot. Even though I didn’t achieve exactly what I was hoping for, I’m comparing myself to professional animations so I think what I ended up with is pretty cool too, definitely room for improvement with some poses and the timing though.

 

 

After animating both a run and walk cycle I came back with fresh eyes. Something Alec mentioned in passing was that he wish he had access to the Maya file as some poses were hard to see with the background I had, so I changed that to something that would contrast Miles a bit more.

I played around a bit more with that jump to the tall building but couldn’t figure out what was wrong with it, it felt weird to me but I wasn’t sure how to fix it yet. I tried cleaning up and adding some in-betweens here and there, as well as some camera animation to focus on the framing of the scene then sent this off for feedback so I knew what to needed fixed before moving on to the next action sequence. It was Christmas break and I didn’t want to hassle the tutors so all feedback from here on has come from friends and family, they aren’t as experienced as the tutors but fresh eyes always help.

Feedback –

My brother thought it was too smooth and flawless for Mile’s character, also noticed the arms being mixed up for the run/jump.

I sent it to some classmates who pointed out the extreme stretch and squash of the landing being a bit too much, as well as the big jump being a little weird and the final landing being too empty of animation for a close up shot.

 

In response to this I added a bit of struggle to the big jump, with his foot barely making it (I push this a little further later) as well as adding much more to the final landing. I added some struggle, making him a bit off balance as he lands, grasping the ledge for stability, as well as idle movement animation while he looks off. I liked the contrast of the long stretch and extreme squash (^seen in first clip) I was hoping it would act as a kind of smear frame and look very fluid but since it came across as a bit too much to some, I just decreased the stretch and moved it up in height, adding a pose in-between them.

 

 

Here I fixed the arms switching/getting mixed up, this resulted in some pretty fast arm movement but I think this looks alright and they make sense now. Some people didn’t get that the sudden snap to the right was jarring so I modelled a physical Spidey sense like seen in the comics and toned the snap down a bit. I also worked on the camera animation to better showcase the jumps, since I think the front angle coming towards the camera was one of the reasons the big jump looked a bit weird. I kept Daryl’s advice in mind for this camera animation and tried to keep it pretty grounded in how it moved. To add to the personality I had him awkwardly go for a Spiderman pose as he landed too before relaxing his arm, hopefully showing he’s still new / trying to be Spider-man – adding Miles’s character into the movements.

 

 

 

 

 

 

 

 

I went in to add details too, mostly on the hands and feet. I made sure all the feet were flush to the ground, and there wasn’t any unintentional sliding, I also bent the toes where I thought necessary. I thought animating the fingers was very subtle but added a lot to the scene without you really noticing, it made his hand go from resting on the ledge to actively gripping it using his powers at first, before relaxing his arm.

I switched the hand from FK to IK for a couple seconds when he lands, since he’s grabbing the ledge but his upper body is moving. IK makes it much easier to keep the hand in the right place. This was easy enough because the rig has a switch in the arm to swap between the two.

 

 

I was hoping to get a web swing/pull at the end there but unfortunately ran out of time. I was definitely too ambitious (especially looking back to the blockout with the ball) and next time I think I should probably try to go smaller scale and focus on one or two actions and perfecting them, quality over quantity. The final animation has parts that still feel very unfinished to me, especially everything after the close up. I wish I had a bit more time to clean up the little scramble/stumble from his hands into the run cycle, the poses and timing feel a little off, as well as have him pull himself off screen with some webs. A couple more polish passes with some more feedback and it could come out really cool though.

I tried to be mindful of the principles throughout. Each jump feels like they have weight, with follow through and overlapping action on the way up and down (on loose limbs), anticipation before each jump and a satisfying ease in and ease out. I did struggle with the level of anticipation on the third jump since he’s running into it but I think it works. I really pushed the squash and stretch since it was so easy to do with this rig, stretching before landing and squashing on it, as well as some squash and stretch on the head when his Spidey sense goes off. The timing and spacing all feels right to me and the arcs are smooth. The second jump’s arc was a bit hard to figure out since he lands on the apex of the jump, so it looked like he was floating horizontal for a while. I’ve definitely exaggerated the poses, sometimes even too much like on the close up landing (which I changed thanks to feedback) and most of the main actions are supported by multiple secondary actions. It was hard to think of the staging, I always struggle with this especially with a lot of movement like this but I think the camera animation I came up with isn’t too much I wasn’t trying to be that cinematic, I used the angles I thought showcased the animation itself best. To keep it as readable and appealing as possible I focused on strong poses, checking their silhouettes frame by frame they feel strong, although some awkward points are a little questionable like the transition from crawl to run.

 

Run Cycle

 

 

The first run I was trying to make was inspired by the anime/ninja run, very quick and stylised leg movement with the upper body bent almost 90 degrees and moving much less. normally holding a sword. I tried to tone it down a bit to better suit Azri and because I didn’t think it would translate to 3D being that extreme. I eventually moved away from this altogether because I didn’t think it fit the model too much and the cloth would be annoying to work with since the legs were so close to the body and I just wasn’t a fan of how it was looking.

 

 

 

 

 

You can get a copy of the Animator’s Survival Kit online so I used this to help me identify my keyframes as well as provide general tips for run/walk cycles like how the hips and shoulders behave in order to convey weight shift correctly. It mentions pose to pose and straight ahead animation techniques, I used a combination. I set the keys to get the overall picture, then went straight ahead from there. I’d separate it in ‘layers’ so I’d do this for the body, then cloth, then hair then I’d go in for the details like hands and feet . I was switching back and forth between ‘methods’ – these aren’t set rules so I was kind of just feeling it out once I had the key points done.

 

 

 

 

 

Most these videos on the run cycle just covered the same stuff in the above book (the poses) so I also had a look at the 12 principles again just to remind myself of what to be conscious of in the run cycle.

 

 

I kept the legs the same from this run and just changed the torso but it looked very awkward. I spent some time playing with the graphs and dope sheet trying to make it a bit smoother, then eventually just redone a lot of the poses since they weren’t the strongest anymore since rotating the body up. I was experimenting with having the right stride be longer/ more powerful than the left to try and avoid twinning and but it was too obvious so I kept them pretty similar later.

 

 

 

 

 

I added some of that downward rotation back into the run so it looked as if she was putting more force into it as well as rotating and moving her body left to right in response to the force from each stride. I wanted it to be quite a fluid and smooth run since she is an elf, but still pretty strong and forceful with each stride.

 

 

 

 

 

Using the graph editor I worked on the arcs of the run, wanting the feet to move in a smooth circular motion, I thought what I had in the previous video wasn’t great. The back legs seemed to drag behind too long and just moved a little awkward when pushing then transitioning to the front. I added some left to right movement of the feet to compliment the left to right bounce of the torso better.

I was looking at some run cycles using the Azri rig and I loved how this one came out. It’s definitely got that stylised anime style I was going to go for previously but even though I was moving away from that it still came in helpful with the seeing how far I could push poses and exaggerate the movements, as well as a good reference for the cloth and hair, although this is a lot higher energy than what I was working on so it would definitely need to be slower and floppier. I thought presenting a run cycle with the shadows and the turn table was very nice though, so since the run cycle was much more aggressive than mine, the staging was my main takeaway from this particular video.

 

 

I played into the lean a bit more and thought the poses were looking much stronger than before, each frame looked very strong and readable and the run was looking much better, so I started looking at the hair and the cloth. It was definitely a mistake to do the cloth at this point because in response to feedback I would make changes to the poses again, and change the timing of the legs – this messed up all the work I done on the skirt and was very annoying to fix. I actually had to use the outliner to find each hierarchy of controls for the skirt, since they would be hidden in the body and legs a lot.

 

I liked Aaron Blaise’s tutorials breaking down cloth for the skirt, it is for a jump but it definitely helped me with how the cloth should behave as Azri falls towards the ground.

 

 

 

 

Working on the cloth was so tedious and any change I made later meant I had to redo most of the frames for the skirt. It means there’s no need for cloth simulation, you just adjust the cloth as you see necessary but with the exaggeration I had in the legs this meant I had a lot of adjustments.

As well as working on the cloth, I added some overlapping action with belt jiggle and the satchel’s jostle, rotate and bounce, I really like how the movement of these props and the belt came out.

 

 

 

 

I also added overlap to the hands and feet, having them rotate a bit slower since they are at the end of their chains. The shoulders would rotate first, then elbows, then hands – all in response to one another and the air resistance (same concept for feet). I did keep it pretty subtle for the feet since she is in them giant boots, but since the hands were much looser they had more delay.

 

 

I had most movement happening in the hips and shoulders, in the y and z axis. I didn’t have a whole lot of movement in the torso to match my original idea, but also because there wasn’t an option to translate the spine, I was hopping to have some spine compression on the contact pose and especially the down pose. I have a subtle hand animation with the fingers opening and dragging behind at the back, and becoming a little more firm and closed at the front. You can see close to the final overlapping action in this clip above. I was struggling a bit with the hair but I’m also getting close to what I stuck with in this video.

The main feedback of the run cycle at this point was that the hair felt too heavy and that the run didn’t feel too fluid, her right arm felt a little snappy from the front angle and the run was a little generic in terms of personality.

 

 

 

 

 

 

 

 

 

 

This documents my final adaptation to feedback, I didn’t do a whole lot of iterations and passes of this run since it was so annoying to work with the cloth each time I made a tiny change, so I just relied on my own thoughts throughout, and my brother for a fresh set of eyes every so often. As seen in the video I double checked all my arcs to make it as fluid as possible, often using the auto tangent and making manual adjustments to this and I really pushed that bounce and left to right sway much more to add more energy and character to the run cycle.

Her right arm had one keyframe that was off (negative rotation instead of positive on one axis) so I made it mirror the animation on the left. I don’t know if I just spent too long looking at this run cycle but after fixing it and having identical animation on the left and right I still sometimes felt a snap on one side randomly. I thought maybe it had something to do with making it cycle in the graph editor but I checked this and it all seemed fine. I actually thought the left arm felt snappy sometimes so maybe its an illusion if you focus on one arm too much or something (like those videos of the ballerina spinning but if you look at her foot she changes direction). I was really struggling to figure it out but I did end up adjusting the arms poses again and ensuring the arcs were all clean just in case, as well as one final adjustment to the legs arcs which I felt pretty happy with.

(other angles available in portfolio post)

I think the final run came out well, good timing and sense of weight and I’m happy enough with the hair and cloth/belt accessories, little delayed from the rest of the body falling and gives a cool effect. I think the hair could be better from the back view maybe and the smaller stands at the front were hard to figure out so I’m sure they could be better. I couldn’t see the ability to translate the spine other than rotations, I wish I could really push the squash and stretch further with some spine compression but other than that I’m happy enough with the exaggeration, maybe could push the poses a little further but I was running out of time to do more iterations, especially with having to completely redo the skirt each time I changed the legs or body. Its presented really well in combination with the front and side views and is very readable. It wouldn’t be a very fast run but you can feel the force behind each stride kicking back and driving forward.

Walk Cycle

I wanted to vary the rig for each animation just to show some variety, and I thought I would use the Jack rig which is pretty simple, but add some complications by adding props. I wanted to utilise Azri’s sword in my run cycle but I decided to skip that and add my own props to the Jack rig instead. I made some props and using the tips in the animator’s survival kit again, laid out my walk cycle. I think I set my tempo at around 16 (on 8s).

 

 

 

 

 

Videos covering poses/phases of walking, weight shift. Some common mistakes in walk cycles like foot slipping or the level of exaggeration in the bounce etc. Most this stuff just repeats the same stuff mentioned in the Animator’s Survival Kit.

 

 

I decided to do an adventurer’s walk cycle, based on a hobbit. This mean it would be quite an open and kind of rhythmic walk, I imagined him a naïve and eager explorer walking as he hummed a song. I first blocked out my 6 key poses then set the tempo. I imported an image sequence of the hobbits leaving the shire to use as reference as I worked, it wasn’t meant to be a recreation but it helped me work out the timing and especially helped with the walking stick.

 

 

 

 

 

Something I really liked from my Azri run cycle was the satchel bouncing out and up as she came down, I thought it gave a good sense of weight and so I animated the backpack in the same way, my walk cycled on 32 frames so I kept this in the same tempo.

worst ref possible lmao

I like the novelty of recording your own reference and couldn’t really do it for the Spider-man or Azri animations, so I attempted to get some for the walk cycle. This wasn’t focused on the walk, but how the character would play with their walking stick as they walk – how high should they bring it, how often should it land etc. It’s not the most useful reference footage, just walking about my kitchen playing with a broom at 2am. It did actually provide me with an idea I didn’t consider though, the stick bouncing up in your hand a little as it lands on the floor.

 

 

I mostly used Frodo as reference for the timing on the walking stick and from the reference and my own experiments it seemed most natural to land every third step – this meant it would swap which foot it landed with each time which I thought was pretty cool. To do this I just animated the right arm on 1.5x my walk cycles tempo. The walk cycle lasted 32 frames so this lasted 48, meaning it would land as the foot hit the ground every third step. I liked this added complexity to the walk cycle since I was using a simple rig. I keyframed this out to the shape I wanted it to move in, then parented it under the wrist control to keep it in the hand.

 

 

 

 

 

 

 

I sent the above video (end clip) to my friend and he thought that the walk seemed very methodical and robotic, so I worked on the poses a bit and mostly focused on the arcs so that I would get that swinging feeling. I made the staff move in a circular path (up and out, then back in on the back swing) and added some more bounce to the walk. The photos show me adding the pop and throw of the stick so he’s gripping the top, then when his hand is passing backwards he pushes down and grips nearer the middle – based on my own recorded reference.

(more ref footage)

 

I added some rotation on the feet to open the walk up further, and some torso rotation to compliment this, as well as some head movement (animated on 64 cycle) to add a bit more variety, sometimes he’d look at stick, sometimes just look around in distance – I kept this very subtle. I tried adding a little head wobble as his feet hit the ground but got rid of it since it looked a bit weird. I made a few iterations of how he used the stick, based on my own judgement just. I thought the way it moved in space was good, but the way it moved in his hand could use a little work. I actually swapped the hand into IK mode for a few frames to make it look as if weight is being put on the stick (very little – I wanted it to be used playfully rather than as a walking stick). I also exaggerated the force he exerts downwards as he grips it lower. I showed this to my brother and he thought to push it even more.

 

 

After my brother’s feedback I came up with two versions, one a bit more deliberate and forceful as he throws it down and one more casual and loose, which kind of pops up and outwards as a result of a tiny throw. We both thought the little toss out was better and more fit the personality of the walk. I liked all the extra animation I could get in a simple walk cycle by implementing these props, even added more to the bag as the left arm moves back and nudges it. I added a slight opening of the hand to compliment the walking stick being thrown outwards and some thumb movement as he plays with the staff while walking. I also pushed the rotation of the shoulders and hips a bit more (counterbalancing one another as seen in Animator’s Survival Kit).

Since it was coming up to submission day and I was playing about with staging and lighting I thought this was a good stage to send to some peers for final changes before I submit it. I sent that second video off^.

 

The video above is what I ended up with following changes based on the final feedback. The main concern was the hand moving too much for a walk which after being pointed out I thought was obvious, so I decreased the movement quite a bit but not to the standard of a normal walk since this was meant to be an adventurer. I tried to make him seem a little more tired than before in his arm movement as before it seemed to perfect and smooth. Other stuff was accessory like raising the toes and adding some finger animation to the swinging hand. I liked raising the toe, it made the foot kind of slap down which felt more satisfying and I tried raising the heel of the back foot (contact pose) but I didn’t like the deformation to the foot and scrapped this.

I also noticed the walking stick was moving quite jittery just before I submitted so I quickly went into the graph editor and fixed the tangents causing this, it was at the end of the loop cycle so I just made sure it ended at a constant speed and started at a similar speed to make it smoother and more of a swing back up.

(more angles in portfolio post)

Haunted Mansion – Portfolio

by

Final Cinematic

Really hope this is as clear to see as it is on my monitor, I know it is lit very darkly. (full screen and close curtains lol)

Quick Diorama Fly Around

Quick fly around both unlit and lit, showcasing my final layout.

Thumbnail Composition Sketches

Original thumbnail concepts focused on composition and layout.

Concept Sketches

Original pages of sketches to help develop art style and share ideas on assets.

 

 

 

 

Candles

Models, Uv’s and textures. (some changes in ue4 material and SSS more obvious, as seen in video)

 

 

 

 

 

 

Books

Models, Uv’s and textures.

 

 

 

 

 

 

 

https://sketchfab.com/3d-models/books-c0917d442bdf40b9a457c7c8083e30c1

 

Haunted Armour + Halberd

Models, Uv’s and textures.

 

 

 

 

 

 

 

 

https://sketchfab.com/3d-models/haunted-armour-ce7b40e159a446abb1884355db461a2d

 

Candelabra + Chandelier

Models, Uv’s and textures.

 

 

 

 

 

 

 

 

 

Piano

Remodel, Uv’s and textures. (also did retopology and uv’s for gramophone and dining table, idk if should include)

 

 

 

 

 

 

Cobwebs

Texture and material set up on plane.

Flames

Flame made using Niagara particle system.

 

Dust

Dust made using Niagara particle system.

Embers/Sparks

Floating embers and random sparks made using Niagara particle system.

3D Environment – Haunted Mansion

by

Planning and Pre-Production

This assignment simulates a studio-like scenario where we are required to produce a Cinematic Short film using the environment as its storytelling medium. Building film sets and digital environments for film and animation involves multi-disciplinary team efforts. However, the director’s vision gives the final output its unique and creative aesthetic. We are required, as a group, to develop and share the main concept and key assets for a Haunted Mansion. Once these are developed, you are then required to use the assets to develop your own interpretation of the environment.

 

 

 

 

 

 

 

 

 

 

 

 

We started off this project by brainstorming in class, listing and sketching the assets we could make, the different rooms, different sources we could use for inspiration. Most of the group were interested in a Lugi’s Mansion vibe, for both the mood and the artstyle. We used Miro to organise and plan further. Throughout the weeks this is where we’d post references, articles, plans and our own concepts.

 

 

 

 

 

I got started on thumbnail sketches pretty quickly, still not confident with these, always spend too much time or too little time on them. Here you can see some of my process, which kind of turned more into digital painting than quick thumbnail sketches. I was focusing on different layouts of the room and mostly the composition and lighting, using candles to highlight areas and having a main focus be lit from a window by moonlight. I also decorated the room using the props we previously mentioned.

 

 

 

 

 

 

 

This is what I ended up with at the end of the first week. I think the compositional thumbnails came out well, it gave us a good idea of what sort of room layout we could go for, as well as how we would light these rooms. I particularly like number 4 and 6 both because I like the ideas and they were very rough and free, just trying to get ideas across. I also came up with this page of asset concepts, it wasn’t anything in particular it was more just trying to come up with a style and direction to go in, as well as get some ideas down that I liked. I wanted the furniture to be tall and off kilter, as well as deliberately scary. This is also where I came up with the idea of broken floorboards allowing light in, and messy disorganised books and candles laying about.

 

 

 

 

 

 

 

 

 

Luigi’s Mansion was our initial inspiration when deciding on the theme of our environment. Films like Tim Burton’s The Nightmare Before Christmas, and Beetlejuice, contain visuals and cinematography techniques that will influence our work when we get to the layout stage. We also looked at examples of German expressionism as this was something that influenced Burton’s work. The Addam’s family, and Crimson peak, have inspired our house set design, our 3d model ideas, and the lighting of the environment. Having both live-action and animated film references has been helpful to us. We were inspired by the use of warped objects and surroundings to create an unsettling atmosphere in the game Little nightmares. The student game ‘The Untold Story of Bernadette’ made by Giselle Valenzuela, has written a blog about the art of the game. We read about her techniques in creating lighting and the 3d models for the game. She also explains her challenge to implement shape language into her work.  ‘the exaggeration in shape language to represent madness and discord’

 

 

 

We hopped in a call following the feedback from our presentations to somewhat make a style guide to follow, more just buzzwords that each concept/prop should meet. Jennifer typed up a few sentences so we could get our ideas across clearly to the lecturer’s and to make sure everyone was on the same page/clear up any confusion.

 

We each picked one ‘hero’ asset that we wanted to make in the future and I chose a haunted suit of armour. I looked online for inspiration and different adaptations of this, and these are two I found very cool and implemented into my concepts – Alphonse from FMAB and a knight from A. Shipwright on Artstation. I particularly liked the concept of drawing from skeletons/bone from Shipwright, and the silhouette from Alphonse.

 

These are the different concepts I came up with, trying to vary them as much as possible. I like to come up with contrasting designs in group work just to get as many ideas across as possible and to let the members vote for their favourite concept/ pick and choose which aspects they like from each one.

 

 

 

 

 

 

 

This was just me kind of playing around blocking out shapes and testing things out in Maya, very early on. I wasn’t assigned these props or anything I just wanted to get my hands on Maya again early and get used to the workspace again, as well as test out what super simple assets might look like in 3D. I did end up working on the candles and helping with the phonogram in the future though.

 

 

 

 

 

The lecturer’s wanted a block out of our environments to show in the presentation, so I looked into this in unreal and how to go about it. I didn’t end up doing it in unreal, I tried but I didn’t have a lot of experience in the software and moved to Maya just to get something done for feedback, but I would come back to these videos.

 

 

 

 

 

 

I based my layout on the set from Crimson Peak, a great reference I found for gothic architecture and a haunted manor. I liked the tall ceilings and huge open plan staircase, this would allow me to have the moonlight I wanted from a skylight above. Having a room under these stairs in another, more enclosed room would allow me to have the fire lit room, contrasting the blue from the moonlit room. I used Maya, built a bunch more blockout props and sketched over this in photoshop for a potential second floor, which was probably too ambitious. I like how the block out came out but I definitely deviated from this since it wasn’t amazing.

 

Above is a video in Maya and a video in Unreal showcasing this, made for our presentation that week. In hindsight this is not a great blockout but it was good to get something done. The lighting in Unreal is pretty bad though, I had no idea what I was doing it was kind of rushed just placing coloured spotlights about. I liked the coloured aspect and the backlit piano, but the intensity and mood, as well as actual light sources are all wrong.

 

https://www.architecturaldigest.com/story/set-designer-david-korins-beetlejuice-on-broadway – environment artist ref

https://80.lv/articles/creating-a-haunted-house-animation-project/ – environment artist ref

https://80.lv/articles/crafting-haunted-chateau-in-ue4/ – environment artist ref

https://80.lv/articles/game-design-building-a-creepy-wonky-art-style/ – building an artstyle/modeling ref

 

These are some links I posted to the Miro at this time that I found useful, mostly from a site called 80.lv which is great for full in depth breakdowns of different 3D projects. One that was particularly interesting was this game design student’s wonky art style and how she went about achieving this., it was kind of perfect for our project.

Modelling and Texturing

 

 

 

 

I had created a sketchfab collection called Haunted Mansion, my source of different 3D modeling techniques used for similar assets to what I was making as well as models with  good topology/topology references, this was useful for things like the knights armour and candles. Some references I used aren’t in this collection, like the pipes I used as reference when making the candleabra. There are about 50 assets in here so I had a lot of variety, Sketchfab is useful because you can go in and look at the individual maps and wireframe, so I could use it as texture reference too. https://skfb.ly/o86JT

Candles

I first made a bunch of candles but I’ll try and focus on just one to save space on the blog, since I used the same workflow for each one and my blog posts always end up so long. I started by sculpting the candles in Blender just from cylinders, using dynotopo to add more topology to work from. I have practice sculpting in blender over the summer and knew I wanted to candles to have cool organic shapes and silhouettes so I thought sculpting was the way the best way to go about this. 

 

 

 

 

 

After sculpting I looked in the different retopology methods online I find using the quad draw tool in Maya to be kind of slow and boring. I learned about the auto retopology feature which wasn’t as useful a I hoped and did end up using the quad draw tool for the majority of the candles, the automatic retopology didn’t have the best edge flow and would complicate my uvs.

 

 

 

 

 

The top and bottoms of the candles had a lot more detail and curves than the middle, so I started with them. The bottom was more simple since I could delete the unseen faces and I approached the top of this one with similar edge flow to a hand with fingers coming out. When connecting the more detailed sections to the middle I had to redirect the edge flow using junctions as can be seen above. This wasn’t necessary but since candles would be so abundant in our scenes I wanted to minimise the poly count to be more efficient for UE4.

 

 

 

 

 

I am super happy with how the topology came out on this candle, I’m not even too sure if its the best way to approach a shape like this but it worked well for me and looked really clean. I thought the UVs were simple too however this would come back to cause me issues when it comes to texel density in UE4.

 

 

 

 

 

I baked details onto the candles from the high poly sculpts, which made the models look more detailed than they really were, but this caused some artefacts in complicated areas like the top underside. I looked online on how to fix this, tried anti-aliasing during the bake but ended up just bringing the baked maps into photoshop to clean them up manually. This method I found on Youtube off overriding the maps within Substance Painter was cool but didn’t work as well as just bringing them into photoshop.

 

 

 

 

 

I’ve used this method in the past but I like using black masks with procedurals and filters to create some randomness, I’ve highlighted this in red to demonstrate but it was actually for roughness variation. I normally use the black and white spot procedurals in conjunction with the blur slop and blur directional filters. Henry also mentioned Sub Surface Scattering since I was doing candles, the middle picture shows this – where the light scatters through the candle where the flame is.

 

 

 

 

 

I normally don’t move on to texturing so fast but if I get one candle textured how I want I could replicate this for all the others easily. eventually created a smart material for the candle wax texture I made to allow me to easily add this to the other candles. I got started on masks for the sub surface scattering since it only would be needed at the top of the candle, this had to be done manually for each candle. Here’s some more examples of the retopology for the candles – this 3rd candle comes out really well after texturing.

 

 

 

 

 

Some further research for working on the candles. Looking more into subsurface scattering in substance painter and in unreal engine, again it seemed pretty easy to implement. To save time on retopology I would use the conform tool like how this picture shows, or even on a cylinder with increased subdivisions to cover a whole, more simple, candle.

 

 

 

 

 

Some more retopology, this took up the majority of the time with the candles probably, since it was individual to each one – I could just texture one candle and replicate this. Showcasing the edgeflow for different parts, trying to keep it as clean and simple as possible since I didn’t want any issues with UVs or in unreal engine.

 

Here you can see the topology comparing the sculpt to the retopologised version – I managed to preserve a lot of detail but decrease the polycount by a significant amount.

The candles were definitely a challenge but I think they were worth it and came out very nice. Got some experience with ‘complicated’ topology with different extrusions and curves to deal with – using essential topology junctions for increasing resolution while maintaining quads, I saw people online refer to the techniques I used as ‘The Diamond’ and the ‘U-turn’.

 

 

 

 

 

I brought all the candles in and got to work on them, as you can see more artefacts in the Normal and Ambient Occlusion baked maps to deal with, just fixing in photoshop again. Then I started making the final masks for Sub Surface Scattering.

 

 

 

 

 

This is what I ended up with before I moved on to the books, I wasn’t sure how to handle the flames at this point. Michael mentioned exporting a simple animation in Maya as an allembic to UE4, but I looked into this a little and saw some people having issues with allembic animations, I also wasn’t sure if an animated mesh would work well or if you’d notice the looping/repetitive animation among the candles.

 

This is what the textures looked like before importing them to unreal. I’m happy with how they came out, there’s a bunch of cool height, roughness and colour variation as well as all the detail from the sculpts. I had a gradient from top to bottom, going from smooth to rough since the top part would be melting while the bottom is more solid. I had height scratches and little extra details in the surface, and yellow shadows with white highlights, using the generators for ambient occlusion. I really like how they came out with the long drops of wax, makes for cool silhouettes, but if I knew the complications they’d cause in unreal engine I might’ve went a little more simple with them. The Subs surface scattering is also super subtle but that’ll be different in UE4.

 

 

Books

Next I wanted to work on the books, we were starting with our simple assets important to the scenes, then we’d do as many assets as we could from the ‘Key assets’. We had an asset kanban set up to organise this and to keep everyone on the same page, it was a little disorganised but it helped. You can see there were some blank spaces and question marks beside assets, these were mostly assets we cut or didn’t get made in time.

 

 

 

 

 

 

 

As well as using references from Sketchfab I would also look at real life examples, although this helped more with the texturing. I did take some aspects from these though, like the sloping books seen in the 2nd picture and the general silhouettes.

 

 

 

 

 

I started by modelling these two symmetrical simple books, fairly standard stuff just starting with a normal cube, adding some subdivisions, extruding faces. They were simple but they’re pretty solid books and a great place to work from.

 

 

 

 

 

I then duplicated these and made some changes, just having fun. Our art style is wonky and non photo realistic, so I played into this adding large buckles and imperfections, sloping to one direction etc. I’m super happy with how the book with buckles came out, even though its so simple I think once textured it looked really nice.

 

 

 

 

 

Here I tested some ideas with the candle, I really like the aesthetic of candle wax dripping over it and makes these two assets seem connected. In order to add a split to the book mark I did some simple junctions, I replicated this for the ripped out page too.

 

 

 

 

 

 

 

I used the same method to UV them all. I wanted the pages uv to be one long strip as this would make the pages flow around the corner well, so if I drew a straight line it would all travel the whole way around.

 

Just going to throw this in here since around this time Dayna was asking about wood textures and how we wanted to approach this. I previously had stylised wood in one of my past objects which I provided, as well as a bunch of stylised woods from online. This was me testing one of the online woods out on the substance jade frog to send to the group incase they liked it.

 

 

Some more research into AO artefacts, I ended up just fixing this in photoshop as always – every online method I use seems to be less effective than this.

 

 

 

 

 

 

I love how the leather came out, using the same methods as always (mentioned earlier) I got loads of great variation. I used the metal edges generator and edited this to add wear and tear to the spine and edges of the book, adding roughness and darker colours as well as subtle height changes. I used a fill layer with a black mask of the anisotropy texture in substance to create a bunch of straight lines for pages, and made this be a decrease in height and slightly darker. You can see above how the UVs helped with the pages flow around the corners.

 

 

 

 

 

 

 

 

 

 

 

 

I added some metal to the accessories that needed them. Once I textured one book fully I just created smart materials to add to each book. I think I made a mistake here to be honest, I should’ve had maybe 4 books per UV tile since they didn’t need to be this detailed, and applying the textures was long and laborious both in substance and afterwards in unreal. I love using ambient occlusion as a mask on a fill layer adding further shadows, it makes these props have a lot more depth  and I always thought heavy ambient occlusion looks better on stylised assets. This blue book is my favourite, reminds me of Fable for some reason I really love how it came out with the buckles and the brown leather top and bottom, as well as the metal detail on the buckles, super happy with this.

 

 

 

 

 

Some books with a lot of wonkiness and chips missing etc. couldn’t have the pages be one long straight UV. The same issue came up with the open books pages, meaning the method I had been using for the pages didn’t work on the more detailed books. My solution to this was to upgrade substance painter as I read online about their new warp projection tool, exactly what I was looking for. This didn’t pan out though, I hated the new workspace layout and the software kept crashing for me so I downgraded back to what I’m used to and did these pages manually.

 

 

 

 

 

I loved this final touch, during my research of old haunted books and sketchbooks, I came across sketches from Guillermo Del Toro’s sketchbook – mostly for Pan’s Labyrinth. I made sure to check with Henry if it was cool for me to use these in my books as I really loved them and I thought it was a cool little homage. I brought them in and added a blur filter, as well as going in with a paint layer to add some/ remove some from them. I didn’t want them to be that readable or detailed as I thought it would take away from the books and the style but still it had to resemble the original, I think what I came up with worked really well.

 

 

 

 

 

I created four  colours, these are the final textures before bringing them into Unreal. I included a close up of my favourite book. I made so many books as well as variations in texture since we all planned on decorating generously with them.

I’m super pleased with how these came out and think they fit into our style really well, I maybe could’ve pushed the wonkiness and imperfections a bit more? but I think they came out good and fit well with the candles too. I was a lot happier with them following one of the later presentations as both Henry and Mike liked them and used them as an example on what to do for another person in the class working on books.

 

Candelabra

 

 

 

 

I ended up with the candelabra and chandelier assets since  no one put their names down to model them – I chose the candles, books and armour and purposely waited to take on whatever assets we needed that no one specifically wanted to model. I was a little lost for direction with these props though so I did some concepts, I wasn’t sure what the group would like in their scenes so I did a bunch of stuff and let them choose the chandelier, then based the candelabra on this.

I attempted to use the curves in Maya to create a cool shape then extrude the cylinder along this, a similar method was shown to me last year for trees. I gave up on this pretty quick though since it was too organic? The shape didn’t seem right.

 

 

 

 

 

I was a little confused on how 3 cylinders would intercept and combine in terms of edgeflow, so I looked about on sketchfab for an example of this. I managed to find this pipe model and I used the same methods they used.

 

 

 

 

 

 

 

 

 

 

 

 

I wanted to use a similar method I used for the books, modelling it symmetrically then afterwards I could use this as a base to build my wonky version. You can see me begin to experiment with the wonky versions above.

 

 

 

 

 

 

 

 

I tested out trying to do a claw/talons at the bottom instead of a conventional stand – seen in my concept art, we wanted some props to have animalistic qualities to them so I attempted this for the candelabra, but didn’t like it at all. I extruded some faces to add spikes along the ‘arms’ of the candle holder then placed a candle in to see how this looked in terms of proportions and was happy with the outcome.

 

 

 

 

 

 

 

 

Super happy with the texture job on this asset, I took a bunch off smart materials such as medieval steel stylised and steel painted rough damaged. I didn’t like any of these materials by themselves, so I took the height information from one, roughness from another, did my own colours etc. It ended up a combination of about 5 smart materials and a bunch of my own masked fill layers. I had the idea to the my candle wax smart material and apply this to the top/ rim of the candelabra with some height information. I used a mask for dirt/splatter then a black max on this to paint on the areas I wanted to have this effect.

Overall I was happy with how this came out, but Mike mentioned he wasn’t a fan of the design of this/the chandelier during our latest presentation. I was going for a gothic dark spiky sort of thing but I guess it missed the mark a little and going into this asset I was definitely lacking some direction so I guess the design itself might be a little weak. He mentioned something with antlers which would’ve been a very nice idea for the chandelier, especially since the taxidermy deer wall mount we wanted didn’t end up in the final project. Despite this though I’m happy with the asset itself, proud of both the topology and the textures.

 

Chandelier

 

 

 

 

I started out with a different shape from the concept I ended up using, I wanted to experiment with a more conventional chandelier shape inspired by spider legs. I was trying to see how this would work topology wise and if it was a viable option.

 

 

 

 

 

 

 

 

This was more experimentation with modelling techniques and different silhouettes. During my organic modelling assignment I had problems before with topology with many cylinders coming together and meeting at the same source, so I wanted to make sure this was all super clean.

 

 

 

 

 

This was the concept closest to the candelabra’s design, if we wanted some consistency among the candle holding assets. I sent the three block outs of these to the group and they all preferred the 3rd one , closest in style to the candelabra. They thought it was cool to have an unconventional design for a chandelier like this.

 

 

 

 

I tested out my candles in these too to make sure they fit together and the chandelier had the right proportions. I was happy with how this turned out, I could’ve went more fun with the design but I thought it created a nice sharp silhouettes and could cast cool shadows, which would tie in with the German expressionism influence we were looking at earlier on.

The texturing was just the same as the candelabra, created a smart material from that then I had to go in with a paint layer to fix up some parts, like in this picture where the uv seams were a bit visible. I used the same candle wax dripping method on this too.

 

 

Knight Armour (+weapon)

 

 

 

 

 

 

 

 

Over the summer I was working on a Spider-man character model so I used this as a base for the knight. I didn’t want to sculpt a full character, I was using this to kind of block out the shapes and I would build the armour pieces ontop/from this. Once I was generally happy with the shapes I imported this into Maya.

 

 

 

 

 

 

 

 

 

 

 

I modelled the leg armour pieces separately then connected the topology up afterwards. I built them using the quad draw tool in Maya over the sculpts legs, then adjusted these into the shapes I wanted. I really like how the leg armour came out and I think the topology was very clean. I used the bevel tool on an edge at the top and extruded this to add a rim to the armour, and some vertical edge loops to add chips in the armour.  The top part had 12 subdivisions, the lower had 10 and the shoe had 8 around the circumference so there were so I had to connect these up by using some junctions and adding subdivisions on the shoe.

 

 

 

 

 

I built the shoulder pads from the sculpt and modelled these armour plates around the torso separately. I added this skirt thing at the crotch because it looked weird where the legs met the armour at the torso. I was a little unsure with what to actually do here but I think it came out alright and looks good by the end.

 

 

 

 

 

 

 

 

 

 

 

I took the knight back into blender and made some adjustments using the sculpt tools, without changing the level of detail in the topology. I then brought my retopologised spider-man in and simplified this down to place the armour over top and connect up all the pieces. I extracted the chest and used this to build up the chestplate. You can see I had a bunch of various knights in the scene at this point – I was still figuring out what looked best and would make a duplicate when I was making a big change.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I like how the chestplate ended up, I think its pretty solid. Overall the pieces came together really well and I think the topology was all looking well. I made the helmet from the sculpt, retopologising then making some adjustments to this. I wasn’t using the sculpt to bake from since it wasn’t a finished sculpt just the basic shapes I was going for and something I could work from, so making changes like this was fine. I’m not sure if it would’ve been better to do the ‘jaw’, upper part and the back of the helmet separately and just combining the objects after but I was trying to keep most of the knight as one object except the shoulder pads. The last picture shows how the topology ended up looking, trying to keep it as efficient as possible.

 

 

 

 

 

I used the smart material steel painted rough damaged as a base to work from and added various aspects from other materials that I liked. I imported the candelabra metal and used everything from that except colour at 50% opacity, to have some consistency among metal assets. I added some dust that was slightly discoloured, raised in height and rougher than the metal. For some variation I made the belt leather holding up the armour at the crotch, quite similar to the book’s leather. You can see I tried experimenting with a version covered in dirt but decided I preferred the cleaner version, the dirt was a bit too much. Following feedback from the presentations, I made the highlights of the armour, like at the shoulder pads and knee pads, a slightly lighter and less rough metal. I made the faces under the mask pure black to simulate very dark shadows and added some extra overall ambient occlusion.

 

 

 

 

 

 

 

 

 

 

 

I also posed the knight since the lecturer’s didn’t like the A-pose he was in. so as they suggested I posed him to be holding a weapon, and modelled a simple halberd for him to hold. I posed him roughly using masks in blender sculpt mode, then cleaned this up manually – selecting edge loops in Maya and adjusting them. To make the blade I modelled a plane into the shape I wanted in orthographic view then I extruded this to make it 3D. The blades and the part where the metal and wood meet are separate objects just combined. I think the halberd turned out well, I started by making it way too big trying to make it fit in and be pretty stylised but eventually I figured out proportions I was happy with.

 

 

 

 

 

I like how the textures ended up, the wood had nice colour variations and some cracks and scratches for added detail. The metal is a combination of the smart material iron forged old and a smart material of the knights armour, as well as the usual masks and procedurals I use for adding variation, baking from the lowpoly mesh actually provided some cool detail I liked at the centre of the metal part.

 

Extra bits 

 

 

 

 

 

 

 

Alisa said she was having some issues with her table’s UVs so I checked out her model and it had a bunch of triangles and a few Ngons, so the topology was pretty messy to work with for doing UVs – this might also cause some issues in UE4 with lighting. I basically just did some retopology for her and show her how she could set up the UVs. It was simple enough just took a little figuring out with the paws of the table legs.

Here you can see the differences in topology a bit better, I managed to reduce the polycount and keep the silhouettes the same so its a little more efficient and I got rid of any triangles and ngons it had, as well as send some Uvs she could use if she wanted. I kept the legs and top separate in case she had adjustments to make.

 

 

 

 

 

 

 

 

 

 

 

I helped Caithlin out with her gramophone since she was new to 3D and didn’t get the Uvs and topology part too much. Again I just cleaned up some loose vertices, Ngons and triangles. I also reduced the detail on some of the smaller parts and deleted all the unseen faces to make it a bit more efficient and give some more resolution to work with. I set up these UV’s and did a little texture test showing Dayna how I set up the masks to separate the materials. since she normally likes to separate materials with different Uv tiles. I sent it over to Dayna to texture while Caithlin textured her other props, and I got to work on the Piano.

 

Piano

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I added this into a separate section because it took a lot more work and changes, almost completely remodelling it. Caithlin worked on the piano early on and sent us some images of it rendered and it looked good, but then we realised at the end it wasn’t uploaded anywhere, and when we looked at it – it hadn’t been textured and had a lot of topology issues.

It was modelled as a lot of objects split up in many different pieces so I started by removing any unnecessary pieces (that could be replaced by extruding a face etc). Then I went piece by piece and remodelled these so the topology was cleaner. It ended up as a full remodel and if you look through the images we went from 20000 to 3000 faces while keeping the exact same silhouette. It’s not documented here but I attempted to reduce the number of individual keys and replace this with texture information but it didn’t work too well so I kept them as individual objects.

I kept the textures very similar to what Caithlin had created originally with lambert materials so the piano still matched her original concept, I just added some dust and scratches as well as some colour variation and metal edge wear – so it better fit the decay of the haunted mansion. I think it came out really well and matched Caithlins original piano pretty accurately. The UVs were definitely the messiest of all my assets just because of all the different pieces, I considered splitting it up into multiple UV tiles since it might be a little low resolution, but this wasn’t an issue except for our lightmaps in Unreal, which we could manually override anyways.

Unreal Engine

 

Flames

 

Before going into Unreal I attempted making the flames in Blender, I wasn’t sure what would work best so I tried out a bunch of different techniques. This video shows me testing the lighting technique in Blender and the animation in Maya, but I realised this wouldn’t transfer easily to UE4 with the same results and I wasn’t that happy with how the flames looked.

 

 

 

 

 

 

 

 

I tried making the flames by animating a mesh fire using a displacement modifier in Blender, but I read about people having issues with allembic files which is how I was planning on importing this to UE4 so I looked into Unreal’s Niagara systems. It was pretty easy to get the hang off, the node system was surprisingly intuitive. I followed this tutorial to learn about the different techniques used to make fire, then made my own flame from this by modifying the velocity, spawn rate and colours, as well as adding a jitter effect.

 

You can see I had 5 or 6 versions of the flames before the final one, I wanted to test out a bunch of variations and send this to the group so they had as much input as possible on what they liked. I was super pleased with how the flames looked and tested them out with the candles. They don’t emit light so I added a point light with an orange colour to have the flame actually give off light, then tested out how this interacted with an early version of the candles subsurface scattering. I realised you cant export this as an fbx which was annoying since my group all needed flames in their scenes, but Henry showed us how to share this easily using UE4’s migrate feature.

 

Setting up Scene

 

 

 

 

These are some of the videos I watched/listened to in the background while working, covering what 3D artists think is important in environments, all focused on unreal engine. Its useful to see their workflow and techniques used, giving me more insight in how to create and structure my environment.

 

I’ll just cover any issues I had with assets or anything I had to learn about in order to set up my scene, since most of the setting up is just setting up material/ instances and connecting up nodes.

 

 

 

 

 

 

 

 

 

 

I had some trouble with my candles for a while. First I had to see where the Sub Surface Scattering details were stored when exporting from Substance for Unreal. Then I had to figure out how UE4 deals with SSS and I noticed once I finally figured all that out that there were visible artefacts/seams when I baked my lighting and wasn’t sure why since they weren’t visible in substance or the material preview in ue4.

 

I sent this over to Henry for some help and he figured out that it was an issue with the lightmap’s texel density, due to how I set up the UVs for the candle. He sent back the candle with new UV maps for me, but since I set the UVs up the same way for each candle most of them had this issue, so I had to go back and redo them all which was a little annoying, since it all seemed fine in Maya back when I did them. I based each UV on how Henry did it, hiding the seams in similar ways and keeping it as one shell.

 

 

 

 

 

 

I had a little issue with how the metal looked in UE4. but eventually figured out it was an easy fix, UE4 was importing my OclussionRoughness map as a normal map automatically, Alisa had this same issue too. Still not sure why this was happening but it was an easy fix.

 

 

 

 

 

It was a little harder to find interior lighting examples un UE4, especially stylised ones, but I found these 3 videos super helpful. I would also refer back to the 80.lv articles i linked on our Miro board a while ago, breaking down moody/horror lighting in interior environments, having a blog going into detail behind the techniques and thoughts behind these environments was useful.

 

 

 

 

 

 

 

I started by importing the assets needed to build two contrasting scenes I know I wanted to showcase, the cosy study lit by candles. and the cold and spooky door lit by moonlight. I knew since very early on I wanted to have this theme throughout, with reds and blues contrasting each other, creating some purple hues were they meet. The red areas would be conveying comfort and warmth, while the blues would be the haunted aspect of the mansion, cold and scary. I wondered if I was going too moody and dark with my lighting for showcasing an environment, but Henry and Michael both liked this level of lighting with the candles that I included in the presentations, and used it as an example of the amount of light candles would give off since a lot of other groups wanted candles in their scenes.

 

 

 

 

 

I wanted god rays to be cast in from the window to match my original thumbnail composition studies, It was useful in leading the audiences eye and loved the effect it gave. This was easy enough to set up using volumetric fog, but I had to swap from using directional lights to spotlights, as the directional lights to pop in and out of existence depending on the angle, while the the spotlight was constant – this might not be as realistic but it matched the style I was going for.

 

 

 

 

 

Having a fully interior scene with moody lighting and areas consumed by shadows made assets hard to see, so instead of setting up 100s of lights I used reflection captures to showcase the assets better, especially those that would reflect light stronger like the metallic assets.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This was just me saving angles I liked so I could recreate them when it came to rendering as well as keeping a record of me developing my lighting, as you can see it starts out much darker than it ended up. I experimented with a more obvious ambient orange fog about halfway through but preferred the more contrasting and saturated colours of the deep blues and red/oranges.

 

Pretty chaotic when its all selected like this but this is my final lighting set up, I set up a bunch of fake bounce lights using spot lights, and had spot lights with low lighting highlighting the different areas that my cinematic would focus on, adding rim lights or just ambient light. I had a general reflection capture covering the whole scene, with smaller reflection spheres for the things that needed to reflect a little more like the piano which was very dark, the knight and the gramophone. I added a lightmasss for my whole scene since there were messages appearing telling me to, as well as a lighting portal for the spotlights casting god rays since they were looking a little pixelated. I added a bloom to all my lights using a post process volume and the god rays come from volumetric scattering within the spotlights, as well as an exponential height fog. I’m really happy with the lighting as it’s what I had in my head from the start with the foggy and moody lighting and saturated blues and reds. Despite it being quite dark, every asset is visible, but the further away they are the more silhouetted they become.

 

I think my final scene came out really well, unlit you can see the actual set up of the scene more clearly. I played into that basement comment Michael made about the stairs and set the scene up as if it was a basement/back exit of a mansion, so its a lot cosier than I originally planned with the thumbnails. I wanted to convey the feeling you get when playing Dark Souls, when you are in a dangerous and scary place but you actually feel safe in the bonfire room sitting next to the bonfire, the only resting place in that dangerous world. I really like the contrast of that feeling of stress and danger vs comfort and relaxation, so I tried to recreate something similar here. The sort of storyline I tried to convey in my cinematic hopefully reflects this.

Forgot to mention I had overridden a lot of props to make sure most of the lightmaps were green (just left the red ones red). The roof is blue since its never in shot and isn’t lit up. This just made sure the shadows and lighting was detailed on these props. I wasn’t sure if going too high would make UE4 more resource intensive, so I left some blue parts like on the clock or the floor – they didn’t cause any lighting issues.

 

 

Cobwebs, Dust and Embers

 

 

 

 

Dust wasn’t something we ever mentioned when listing our assets or anything when concepting this environment, but something I came across during my own research into environments. It’s common in video games like Horizon Zero Dawn and the Last of Us, as well as a lot of other artists environments, especially common in god rays which is something my environment had. I thought ambient dust was something so simple that added a lot to a scene and since I had some experience now in the Niagara system I knew dust would be something I could make using it.

 

 

 

 

 

I created a Niagara emitter from an empty template then added a spawn rate and a box location for them to randomly spawn in. Then I added random movement paths and lifespans, random colours and opacity within a range and random size within a range. I added a fade in fade out effect to make it seem more natural and a random velocity within a very low range, since dust doesn’t travel very fast. A video I found online helped me out with this but his dust had a different effect than what I was going for so I drifted from this pretty quickly. The last picture shows what my final system looked like, and from it I made 6 variations, allowing me to have more concentrated areas with dust like the godrays.

 

I wanted to keep it quite subtle because the render made it show up a little more obvious, and I didn’t want the dust to take away from the environment itself, since in games and other environments it is a very subtle effect. I love how it came out, not only does it add to the god rays, it adds movement to the scenes which makes it a bit more interesting when the camera is on a stationary object – a great example of this is when I focus on the chandelier in my cinematic.

 

 

The whole scene ended up full of dust, selecting it all makes it more obvious. Since it has the possibility to spawn so small and translucent though it didn’t take from the scene and came out pretty tasteful and subtle I think. I added some more concentrated pockets – areas in which the camera would move through, since the further away you are from the dust the harder it is to see.

 

 

I modified my dust system with new colours, more opaque and a higher velocity and spawn rate, this was to add embers from the fire. It was missing a little something with just random movement around the fire, so I also added a new system based on the direction burst template, to have sparks that would fly out randomly in a set direction at a higher velocity – this happened at random rate between a set range. I set up my fireplace with three flames since one large flame looked a little silly, and really loved how it came out at the end with both the flames and the embers. (Megan also did a good job making such a cool fireplace)

 

 

 

 

 

I also (very) last minute made a cobweb, I found this tutorial online but it was more realistic so I made some changes. Instead of using a texture online I drew my own in photoshop. He uses masks in his tutorial, I changed this in mine to use transparency, I thought using masks made the webs too visible and light couldn’t pass through them so if a light hit them they were very visible. I also changed my material setup to be slightly emissive since my web was transparent. I thought the grass wind part was cool though, letting it slightly wave as if there’s wind.

 

I’m happy with how it turned I feel like it fits in quite well and I like how it moves, it’s pretty simple but adds something extra to the scene. Since I made this the day of submission, while waiting for some assets, I wish I had more time to do variations as well as clean up the textures in photoshop though, this would also let my group members add them into their scenes, since mine was drawn to be in this corner it might not fit into their scenes too well.

Rendering

No idea what happened but when I opened up my project again to show how I set this up, my sequences are all missing and I can’t seem to get them back (somehow deleted them?). I’ll just type up how I did this part. I found a video of me testing out camera movement but that’s all the documentation I have of the sequences.

 

 

 

 

I looked at some videos linked during to make sure my render would come out as good as possible. Just before I started rendering I added another set of lights to my flames, with a slight flicker light function following this tutorial, this is more obvious at the fireplace. I ended up rendering to PNG image sequence as it was much higher quality than everything else for me and allowed me to review each frame, I compiled this in premier.

 

 

 

 

I wanted my first shot to start by going down the stairs, as it would be going into the dust and establish that the rest of the mansion is upstairs, so this is some sort of basement/lower back door – since Mike said the stairs looked more basement like than mansion grand staircase. I used the techniques Henry showed, switching my camera to film dslr and animating using the debug focus plane to animate the focus and get some cinematic depth of field effects.

 

 

My basic methodology was going to my screenshotted camera angles from before and recreating the camera angle – I’d use that as the start or ending of that shot. You can see at the bottom in the video above I have a sequencer open, I keyframed the transform and the focus distance (manual). I animated the transforms to have some imperfections that made it feel a little handheld, I like that effect in the final cinematic. If I was trying to convey some stress I would use camera rotation, like seen in the clock shot. I tried to match up the end of one shot and the start of the second one, so the transition wouldn’t be jarring. For example if I was moving left to right at the end of one shot, I would open the next by moving left to right. I played a lot more into this with the clock, chandelier and door shots, transitioning between them with a heavy rotation (rotate left to a wall at end of shot1, rotate left from a wall at start of next shot) –  I really like the transitions between these shots, I also lined them up with the music so I think it’s very satisfying.

 

 

 

 

I had saved some previous cinematic to reference when doing this part, just to see how long to hold certain shots or the camera angles used in these horror environments, one was even a haunted mansion which was very relevant. This Youtuber William Faucher has so many useful tutorials I used him a few times throughout this project.

 

I edited it all together using premier since I exported as image sequences and I am used to editing in Premier. I had a cross dissolve between almost every scene since it worked quite well with the slow nature of my cinematic. I added an audio track I found online called Alone In The Darkness (Sad Music Box Melody – Original) DIMD, I put this in .9 percent speed to better match the length of my cinematic. I generally cut on the beat since it adds good timing to the edit and feels satisfying when watching with audio.

Super happy with how it came out think it works really well and feels very cinematic. I thought interior environment would be a much harder than an exterior but I think it worked out good. It conveys what I was going for and feel like it will have a very different vibe to the rest of my groups cinematics. I love the slow pan shot revealing the knight, with the dust in the god rays. I really like the transitions between the clock, chandelier and door, as well as the camera movement on the clock. I definitely got more experimental and confident when working on the later shots. I feel like the piano shot could be better, and the transition to the next shot. I think the opening shot was very strong other than that though.

Something I realise when watching it not in full screen is that it’s hard to see with the white background, since its such a dark scene, so it might be harder to see on certain monitors or with the screen glare. I wasn’t able to go back in and render the shots brighter for this since I somehow lost all 8 sequences, but I adjusted the brightness in premier afterwards and didn’t like how it looked so I left it how it was. (I wonder if a better approach next time I do a lowlight/night scene is to render brighter than needed, and lower it in premier afterwards) I feel like the level of light is good if you are watching in fullscreen with the lights turned off, I watched it on my phone in case the colour and brightness is different from my monitor, and it was just as clear when I turned up my phones brightness so hopefully its all good. I don’t think this is a big issue because any low lighting cinematic or trailer is hard to see with screen glare/lights on.

Really proud of all my assets and the final cinematic though, I definitely get a similar feeling to being at a bonfire in Dark souls when I’m next to that fire showcasing the books before immediately cutting to the uncomfortable angles of the clock. I think the particle effects and god rays make it feel very stylised and atmospheric and all the props have ended up with a consistent style.

Reflection

This above paragraph was kind of me reflecting on the final cinematic so hopefully I don’t repeat myself too much here. I really liked this assignment and module. Group work is always fun because you get to work on something with  a much larger scope than if you were doing it individually and I had a good group to work with. I feel like we produced a high quality assets in an cool style, and individually I produced a pretty solid cinematic. This environment assignment has definitely got me interested in Unreal Engine, after seeing the capabilities, I already want to start another project in it now that I kind of know what’s going on. The lecturer’s have been really helpful too in both assignments, Henry helping with problem’s I had with assets and Alec providing a video of recorded feedback when I asked. I can’t speak a whole lot on the other assignment because most of my focus has been on this group environment but I’m enjoying it so far too. There are definitely things I could’ve done better like not spending weeks on just candles, and leaving more time for the cinematic since it was made on the last day. I feel like I’ve improved my knowledge of topology and edge flow during this project and have developed a pretty decent workflow for texturing in substance – producing a bunch of assets I’m really happy with. Overall I think my final environment has come out super well, it feels very atmospheric and the contrast of the reds and blues gives a nice effect.