The Bravery Animation is the second group project of the semester and possibly the most challenging one. Our groups were assigned for this project and we were tasked with the challenge of making an animation that links to the theme of ‘Bravery’.
We considered multiple ideas, but we settled on ours being about a knight rescuing a princess from the tower. Except the twist would be that he isn’t brave and doesn’t save her. Therefore, it is still within the theme of Bravery.
The Initial Idea
We decided on the knight and tower idea relatively quickly, but we had to develop upon this story further which was the problem. The idea originated from an old drawing that Veronika was working on which she called “The Good Knight”, I jokingly said that his nemesis could be called “Knightmare” and then we stumbled upon an idea.
I suggested inputting these two characters into a story where they act as a knights conscience. The main reference that I used was Krunk from Disney’s Emperor’s New Groove.
(However we can see these characterizations of the conscience in plenty of other examples, such as The Simpsons).
I designed a very quick and rough sketch to show the team what I was thinking roughly…
The left one was based loosely on Veronika’s sketch and the other two were just showing good and evil while maintaining roughly the same silhouette (so it wouldn’t be too much extra modelling or rigging). I had Tyrael from Blizzard Entertainment’s Diablo III in mind when quickly drawing the right side one.
At this point I was also exploring different kinds of knight shapes and working with a variety of proportions, to see which would give the best result. My favourite would probably be the second from the right with the very exaggerated proportions…
This was the rough animatic that was made (my frames are the starting section up to around 0:09 – Roughly 34/35 shots, which was supposed to be quite an intense tracking camera movement, so it does look a bit jumpy, apologies).
The rough story was that the Princess actually doesn’t want to be saved, so she sets up speakers at the window to scare knights/potential suitors away as they will think there is a large dragon protecting the tower, but really it is just a small dragon with a speaker phone. The animatic was quite poorly executed/organised but that’s was the general gist of it.
We were thinking that the Princess would be a bit of a gamer girl – We decided not to set this back in Medieval Times thankfully and this allowed me to give the Princess more of a tom-boy chilled looking personality rather than a pink flowery Princess. I created a concept that I was really rather pleased with based on this ‘Gamer Girl Princess’ persona.
I had begun to do a character sheet drawing of the character (not shaded as you can see). However, Charlotte wanted to do the Princess, so I stopped around here with my Princess designs.
At this stage we still hadn’t settled on a knight design. Therefore, I did another few of those as well…
I was very pleased with this drawing as well, I used some of the techniques that we learned back in the Life Drawing days. I then went on to design a Dragon Drawing or two.
These were very much based of Mushu (from Mulan), as I really like the shape and style of him. Even though he is so simplistic, he has so much character and is able to show a vast amount of expression through his movements.
After presenting this idea, we decided that it wasn’t good enough, so we changed it up, we created a more interesting story about the knight’s journey to the tower, rather than him just standing still at the bottom.
The More ‘Refined’ Idea…
We drew out storyboards for the new idea, and we asked Mike (Michael Bass – one of our tutors) for some advice, he told us about a few areas which could be cut down and he liked our new idea. He wasn’t very fussed on the old one… (Apologies about the poor photo, it was drawn quite lightly and it wasn’t great lighting).
The new idea generally shows all the trials that the knight faces along the way to the tower. I discussed the idea of narration with Mike and he agreed that it would be a good idea to help tell the story, he gave me some advice on narration and how might be best to do it.
After this, I drew out a couple of ideas of some of the trials he could go through, my favourite one (drawing wise) was the geyser field idea (below). We had discussed many different ideas, ranging from jumping over spike pits/geyser fields/bees chasing him/cliff face climbing and of course the snapping rope bridge.
When thinking about this idea, we were discussing the Shrek scene where Prince Charming braves all of his trials on the way to save Fiona from the tower.
We based the rope bridge idea mainly on this clip, except we didn’t want to completely copy it. Therefore, we decided to make the bridge snap due to the knight jumping on it. We thought this would be a great transition from him jumping over the spike pits. Once the bridge snapped the knight falls, after a second or two though, he shoots his arrow up and swings heroically leading him to completely face-plant the cliff-face.
I wrote the narration, the next day I brought it in to see what people thought and my group seemed to like it, they found it funny.
We then wrote it up into a script together in Uni, Charlotte said that she had a friend who studied Drama who might be able to read the script for us. However, it turned out that he couldn’t do it, he suggested a couple of slight changes to the script that the rest of the team thought were better.
I’m still not too sure though, they were only like a couple of word changes here and there, but I felt that the changes made it more complex and longer which I felt took away from it a little bit.
The Next Animatic/3D Previsualisation
In the previous animatic, I made my section first on Maya to get the camera angles and all sorted then I hardware rendered them and drew over them in Photoshop. The team felt that this was a good approach to it. So our animatic is also our rough 3D Previs.
I decided to just take on the role of the narrator, generally whenever I was showing it to people or my teammates were, they found it funnier that it was me. Anyone that I showed the animatic to laughed at a lot of the little puns and narration jokes that were made throughout. It was also easier rather than having to rely on another person who doesn’t really have a connection to the project.
I did the 3D part of this animatic, to keep it consistent we tried to take a part each, so I did the 3D sections and character posing, Veronika drew the character line art on and Charlotte or Jessica quickly drew the environment on.
I’m reasonably happy with this animatic, I feel it turned out quite well, a lot better than the first one at least. I know the end needs tweaked in particular, but I have a couple of ideas for this. For example,
- The Princess marking a board of knights scared away, as the knight is running over the hills in the distance. (I did a very quick and rough drawing of it to show my team what I meant).
- I think it was Mike that possibly suggested having the Knight joining a queue of defeated knights.
- The Knight waking up at the bottom (POV) with the Dragon approaching with a broom sweeping him away.
- The Dragon knocking on the helmet, lifting the visor and waving goodbye, (maybe winking while he does waves) to the camera (POV with a waking/shuttering eye effect – as if the knight is just waking up again).
Bye Bye Princess
Unfortunately, after presenting our Animatic, we got the general consensus that three characters was a bad idea. Most other groups only have one character, therefore, we cut our Princess out as she was the least vital to the story (as you can see in the animatic, she isn’t much of a vital part of the animation). We are thinking that we could just suggest to her presence, she could maybe be behind a door or something, or overheard in the background (if she is playing video games or something).
Anyways, we had to try and simplify it down, so the Princess had to go.
At this stage I also started to think about the environment and how it would look like, I considered an establishing shot from inside a cave. Therefore, I decided to draw out this concept, to better visualise it and be able to show my team my thoughts on the environment.
[I have been working on this for quite a while, I had all this dragon concept designing section done at around the time of the first presentation. But I forgot to put it in up there while blogging about it…]
I wanted to try and make the dragon a little less Mushu like. Therefore, I went through quite a number of quick Dragon designs to try and discover the shape that I wanted to go with… Here are a few of them. (Some of them were very quick to just test ideas).
They started to get a bit better or more interesting at least (in the images below)…
I started looking at other creatures for reference, I looked at Pascal – the Chameleon from Tangled. He fits the characteristic of “cute” while he still has scenes where he pulls more ferocious/angry looking faces. (Image sources on the photos).
One of my main inspirations still was Mushu from Mulan he is a very energetic likable character, I would quite like these qualities for my dragon as well. (Image sources on the photos).
I also considered a couple of realistic references –
Veronika was talking about Crocodile Skinks (mini crocodiles) and we thought they might be a nice addition to the dragons qualities (“cute”) if he was similar to that. (Image sources on the photos).
I found another species that looks possibly the most similar to a dragon as is possible (except size of course). The Sungazer or otherwise known as the Smaug Giganteus. (Image sources on the photos).
I also unknowingly used my character designs rough ear shape from last year in the design (which was based from Dobby the Elf from Harry Potter as well as Ratchet from Ratchet and Clank).
After discussing with my group, we were aiming for an Overwatch type style… Or just a general Blizzard Entertainment type style. The Blizzard games are some of my favourite games of all time.
I have played World of Warcraft for hundreds of hours throughout my childhood/early teenage years. I play more Overwatch/Diablo III now, but still all their games are great, even the strategy ones (Yes, I also play a bit of Hearthstone/Heroes of the Storm from time to time). I obviously have a lot less time to enjoy these games now due to University work, but if I get to work in the same style, it’s pretty much like playing the games…
(Image sources for the images below are on the images themselves).
Therefore, the style was perfect for me and it made it so much more enjoyable to work on. I just hope I will do the style justice…
The Dragon Concept
I eventually designed a dragon concept that I thought was good and thankfully my team thought similar, so that’s what I went with. – I hope you can see the similarities between the references above that I was considering…
I drew a rough turnaround to get a better idea of how to sculpt him.
I then started straight into ZBrush with the model (as that was the most ideal way to make an organic character such as this). It was a very enjoyable process sculpting the dragon, ZBrush is great when you get the hang of it. I used Dynamesh as well to create a more smooth polygon flow over the entire character.
I then brought down the topology count from a couple of million down to around 19 thousand using Z-Remesher along with the Z-Remesher Guides – to keep correct topology around the character facial and body features so that they could be easily animated.
I had Z-Remesher guides around –
- The face – eyes/mouth/cheeks/nose/ears
- The body – neck/chest/hips/tail
- The arms – armpits//elbow/hands/fingers/thumbs
- The legs – legs/ankles/feet/toes/tail
These were to keep the correct edge flows strong in these areas. I did create a high and low poly version of my dragon using the project function in ZBrush, but in the end I didn’t end up using the high poly because of the style we seemed to be going with.
Using the UV Master I UV’d the model on ZBrush itself before doing the whole re-topology and projection process.
Thankfully, I had already done all this process before in the previous environment project (Space Vikings) so I already knew how to do it. However, I guess in that way, I used the same sort of tutorials. I also helped a couple of my classmates with their UVs and linked them the tutorials so that they would know how to do it again in the future.
I used Substance Painter for the majority of the Dragons textures… I experimented quite a lot with different textures for the dragon, ranging from different colour schemes, to scale sizes, material choices.
I found this to be a really enjoyable process. I also baked the maps inside of Substance Painter instead of Maya as it gives a far nicer and cleaner result in comparison.
I must have spent a good few days to a week texturing, looking back I probably should have spent less time on it, but it was reasonably early on in the process (and my teammates still hadn’t made their models/characters so I felt it was fine to spend that time messing with textures and learning Substance Painter a little better).
I remade the eyes in Maya just using a simple sphere, I then UV’d it and created the texture on Photoshop using the UV snapshot to draw the texture on to ensure that it was correctly positioned and that the pupil remained in the middle. I wanted to get a dragon/lizard looking eye, so I looked at a few google images of Lizard eyes. The UV snapshot is a bit hard to see, but hopefully you can see it alright.
I did the tongue on Substance Painter as well to try and get to the bumpy tongue texture in the normal map.
I would post the colour map/normal map, but it is extremely plain due to the UV being so small and the tongue details being so subtle. Ah, here they are anyways… You can see more of the details in the roughness map which I’m not sure that I actually ended up using…
As I said before, I looked at different colours, such as red/orange/blue/different greens etc. Unfortunately, I can only find the bronze one as the others I was just playing around with in Substance itself mainly.
I ended up lightening the green up a bit and going with it (as you can see from the earlier image that I posted).
I baked out Ambient Occlusion maps as well and Overlayed them on top of my Diffuse maps on Photoshop to give it more of a depth/character to it.
Here are the maps that I baked out and plugged into the Ai StandardSurface texture…
I touched up the last one and lightened up the horns/nails as they were very dull…
Here are the rest of the maps, I had to make a few adjustments to the normal map in Photoshop as well, to fix some areas that we protruding too much or that weren’t following the model how I wanted them to.
They follow the order of Height Map – Normal Map – IOR – Specular – Roughness Map
Due to me already having rigged before in first year (my Astronaut) I was keen to try it again. I actually found quite a lot of satisfaction in making a character move and do semi-complex movements, it felt quite worth-while. Below is a reminder of the rig, as you can see, he doesn’t have a facial rig either, so it’ll hopefully be fun to get all that working.
I wanted to improve upon my previous rig. In making it I followed several more basic YouTube tips and tutorials for it (not that they are bad, but I wanted to try and push it to the next level passed that)…
I followed this James Taylor tutorial understand how the facial joints worked.
I tried to match up all the joints to my character that he discussed in this video. I feel that I successfully did.
However, with my dragon, he was a bit too high poly in comparison to the model that James Taylor was using. It was a bit of a challenge and it took quite a while as my character isn’t human as you can hopefully tell, so I gave him a few extra joints here and there…
After firing Joints (NURBS Curves) on I was starting to feel that it was possibly a bit too much for the character. I actually made half-eyelid shapes at this stage too (which I didn’t end up using thankfully…)
After presenting, Alec (Alec Parkin – our tutor) seemed to think it was too complicated and seemed to agree that the face didn’t need that many movements in it. Therefore I simplified it. He recommended that I used Blendshapes to make the facial movements. Which I plan on doing, but I wanted to be able to control the face with the rig as well to exaggerate certain aspects if necessary…
I tested the rig and it was working pretty much how I wanted.
However, this was starting to look very basic around the body, which as I said earlier, I was trying to avoid. Therefore, I changed it up and started following a more complex set of videos about rigging. The series of videos was from Lynda.Com I think and it was 75 clips long. I think I watched around 40 of them, and I learned quite a lot throughout the process.
However, little did I know that this was going to be the longest and most difficult part of the whole process – Since I had spent so long on it though, I wanted to finish it – I had the old version as a bit of a back up though, I hadn’t fully finished it as I knew that I wanted to do this instead of that…
The leg was set up with 3 Different IK joints and through the “common” reverse foot process.
This is the reverse foot (Red), it gets constrained to the normal foot rig (Yellow) in multiple places… I later came back and simplified this, removing all the various toe joints, as my character didn’t need the capabilities of moving individual toes separately.
The IK joints went from the hip to the ankle, the ankle to the main part of the foot and the foot to the toe joints. Once I constrained the various joints to one another, I constrained it to the foot control and then added attributes to the Foot_CTRL to allow the foot to raise from the heel, swivel from the heel as well as bank left and right.
I then added controls to allow the foot to raise from the base of the foot as as well as from the toe.
The Spine and Main Body
An IK handle was put into the spine and cluster nodes were added to it. I then set up the body CTRLs. One at the base of the pelvis to control hip movement and stretching of the spine, one in the middle to control chest compressions and exaggerate hip rotations, one at the top to control the shoulder and head and one that runs along the spine to control the movement of all three. This is where problems came into play with the Script editor. Alec (my tutor) figured out that they were impacting the global scale control, which was a problem when scaling the character (and it was exactly what the script was supposed to prevent according to the tutorial).
Next was fixing this up and creating the Hip_CTRLs. At this stage I was also messing a bit with the arms as well (hence the blue joints (IK arm)).
As I said, I had started with the IK arm there… For the IK/FK switch to work, the body needed 3 arms, one being the IK arm (Blue), one being the IK arm (Red) and finally one being the resultant arm (Green) which used the values from both to create a switch. Unlike the Jack rig I wanted the switch to be able to smoothly transition from IK to FK, which is exactly what the series of tutorials showed/taught me what to do.
First I made the controls (CTRLs) for the IK arm and the hand (above). Then I made the FK arms and the Resultant Arms.
It was at around this point I added in the shoulder and elbow constraints and controls. However, the main thing at this stage was getting the IK/FK switch working so it controls the resultant joint with the value input into the hand (wrist) control.
I felt that my previous rig from earlier was actually more complex than the tutorials rig, so I left it as it was. It already did all that I required from it, all the other movements I was intending to use blend shapes for. It could move the direction, nose flare, ear gestures, jaw movement, neck rotation. Therefore, I felt it was good enough to just keep as it was…
Here’s a wee test of the eye movement/point (below).
Filing System Problems
Unfortunately, once I constrained all the parts together and binded it to the mesh, I ran into a few problems with the grouping system and a global scaling issue. My Master_CTRL wasn’t working and my eyes/tongue weren’t functioning properly and when I scaled it, the rig and mesh would unfortunately separate and scale at different proportions. I tried for ages trying to fix these problems and then I asked Alec (my tutor) if he knew what was wrong with them. He was then able to fix the problems and he said that he would send me videos on how they were fixed (so I’m looking forward to seeing how to fix the problems with my rig).
I then made a few minor changes and skinned the rig to the joints a bit more… I also added in a few smaller things such as finger/thumb joint CTRLs (which allow for rotation and I added an attribute to give them a middle joint bend in each) and a Tail CTRL (with the same functions as the finger CTRLs)…
(I know that the tail wouldn’t be making a movement like that, but I wanted to try and show you how much versatility it actually does have…).
Although I had done this already before starting this complex rig, I thought the blog would flow better if I put this down here…
First of all, I will show you the standard face, so the rest can be based of that…
The Blendshapes order – Angry/Smiling/Shock/Blink/Inhaling (“That looked like it hurt”)
The rest of the expressions or if they need further emphasis that can be done with the facial rig controls…
Not going to lie, at this point my team are very far behind schedule. I have been working on the tower throughout the rigging process but, other than that, we don’t have much at this point… Here is my tower process.
I started with drawing a stone brick while thinking about how I wanted the tower to look. Here it is below…
I know it doesn’t look like much, but during the time I was drawing this, I got a good few of the tower ideas sorted. I then started to actually draw a few tower concepts…
I quite liked the first one, but it was a bit too small in relation to the size it needed to be. Especially since the princess wasn’t going to be shown in the shots, I had to make sure there was enough room for a interior wall or door to block of the rest of the towers interior layout. This meant that the ‘head’ of the tower had to be bigger to allow for this space that wasn’t shown and just to be suggested.
I made a few more tower designs and researched it a little more. I found this really nice sketch in sort of a Blizzard-looking style.
I found this other piece that looked great, I more used the left-hand side tower section as reference myself though…
The main reference for shape was the tower from Tangled (shown below).
I began by blocking it out roughly using basic shapes and models just for rough markers.
Unfortunately, I don’t have an awful lot of saves when it comes to my tower, as most of the time I just saved over the previous save. Therefore, the next set of shots are a good bit further along in the process after I had worked out a lot more of the problems and proportions that I wanted.
As they did in Overwatch, I used my dragon for scale, this was an extremely helpful marker and landmark for the modelling process as I was able to create things based on the size of the dragon from then on.
The next phase involved blocking out the interior (wall/roof/floor/door) as well as the rest of the exterior. I mainly made the rocks/stones for the tower, to allow it to be possible for the knight to climb it and the wooden boards and supports.
For the rocks (above) I made half a ring different larger rocks and then duplicated it around to form a full ring, I then scaled the ring rotated the ring and moved it down, I filled in the larger gaps with small rocks and repeated the process around a few more times. I then had formed a group of several rings all filled in, I then scaled this group -1 on the Y-Axis to give the impression of different rocks. I filled the gap in with more smaller rocks and repeated the process. The smaller side towers followed the same process roughly.
The modelling was pretty much finished by this stage… (I have made a few more models and adjustments throughout the next few stages while UVing and texturing).
While UVing, I found that I had mad a mistake by not UVing the rocks originally before duplicating. Therefore, since they were only using one texture, I UV’d one and transferred the UVs using the Transfer Attributes tool in Maya – as they all had the same polycount. This too was a slow process as it was 1 by 1 (I did it for a full group of around 4 or 5 rows then just duplicated them down) but it felt faster than repeating it and UVing from the beginning modelling stage.
Introduction of another style
Fortnite – Epic Games
Lately during my breaks from work, I have been playing a couple of games of Fortnite, which is a very popular game at the moment on both console and PC. However, the main reason that I play it, is the style, don’t get me wrong, the gameplay is great too (and can be very intense), but for me the style is the big hard hitter and the thing that brings me back to the game.
The texturing style within the game is incredible in my opinion, and I found out from friends in the class that Epic Games (maker of Fortnite) use 3DCoat to texture. Therefore, I downloaded 3DCoat and gave it a go, and this software is incredible for texturing hand-painted objects and textures. It felt just like an Adobe Photoshop for 3D meshes.
I found a texture artist at Ubisoft who I found who worked on some of these seamless textures at one stage… He is called Tangi Bodio.
ArtStation – https://www.artstation.com/artwork/o6z34
Here are an example of some of the stone/brick textures that he worked on while at Epic Games –
Here are a few of the other textures that he worked on, such as wooden floor boards and tiling textures, which are slightly similar to my own later in the process.
The last texture, which would have useful if I had have found it earlier in the process or had more time was this one of a ground/grass texture.
In my tower there are around 20 different textures. I painted each separately on 3D Coat, so I won’t go into detail on them all, instead I will just show you the a couple of images for each… (Some have been altered on Photoshop after being textured primarily on 3D coat).
Wooden Plank Support –
I had to do 3 different types of this texture to be used on 3 different types of Objects/UVs, I did this twice and the third UV was done on Photoshop as I made the UVs in a way in which I could later replicate them…
Wooden Arch –
Possibly the second most used texture within the scene, as I have quite a number of arched supports throughout the scene.
Main section of tower –
There is two parts here, one is the main cylinder shape and the other is an outer section of the window with the same type of plaster material on it. (I did make both separately just using the same colours.)
Wooden Ring_01 –
I tried to make good use of my textures, so I try and re-use them as often as I could. In the image below you can see this texture in use just shy of 10 times or so in 2 different shapes…
Window/Door Frame And Window Sill –
Since the camera would be quite close to these particular textures, they needed to have more detail, and you can see that a bit better in a couple of these renders below.
Underlying Tower Granite Textures –
As you can see, this is quite a rough texture, since it is underneath the rocks and in the dark, it didn’t need to be incredibly detailed, so I drew a quick cartoony looking granite texture and made a section that was lighter than the other as it is not covered by rock (but it isn’t seen that much by the camera up close)…
Using the same texture I was able to adjust it in Photoshop to suit the side towers.
Small Window –
I did this texture primarily on Photoshop actually, as it was easier to get the symmetry on the UV map since it exported with weird symmetry lines.
Tower Top (Spiked and Rounded) –
Interior Wall/Door –
As you can see, I re-used the window frame texture from earlier for the interior door frame texture. I think it works well for both. I added cracks to the walls and small sparkles or dark areas to the walls to give it a bit more character, I also drew in the door handle and bolts to test whether I could make them stand out with the 2D texture.
I tested the rig – I felt it was best to keep it to this section, but it was tested after rigging the character as well.
During this scene I modeled the deodorant can to allow me to test the rig in a funny way.
Deodorant Can –
This was a bit of a later test showing the dragon spraying his armpits (which isn’t in the scene – but I wanted to test a lot of the controls in one movement that makes sense with the scene and the props).
After this I began animating the first scene of the dragon, where the knight looks up to a friendly Dragon looking back at him.
I then animated the next section, this was it initially, it’s quite rough at this stage still, still not finished…
I put both into the same YouTube video you can watch it here (below).
Before I could finish this scene I needed to do the model for the lighter. I just textured it using Maya, so it is made using two Blinns for the Silver and Gold and a plain Lambert for the black switch. This is also made with a simple rotation so that the lid can be flipped up from the joint in the middle to close properly as well.
After this I then put the deodorant can in his left hand and the lighter in his right and started developing the movement further. (There is a shot of the knight supposed to be in between here and it’s not just a sudden facial change).
I was happy with this so I started rendering it with a simple 3 light set up to show the interior and exterior quite well.
This scene was re-rendered with the correct texture, and I went on to animate the final scene as well.
I basically animated every scene that the dragon is in, I very much enjoyed the animation process using my Dragon Rig, it was quite satisfying.
I also made the Cave and Spike Scene environment as well as a couple of planes for cliff textures…
Veronika and I put on the majority of the scenes to render and whichever ones didn’t get finished we play blasted the rest of it, I roughly assembled the clips into the video below. Unfortunately, the voice actor who we arranged hadn’t finished the voice acting by this stage, so it doesn’t have voice and the rope bridge scenes aren’t completely finished either.
Playblast Animation – No Sound Yet…
PROJECT – PORTFOLIO
DRAGON RIGGING/BLEND SHAPES