Character Design

DSC_3497

Better images will be taken tomorrow, when the sun is back out. Apologies for the dark images for now! This is my little character. She is a happy go lucky kind of squirrel, who is dainty and soft, but you know she might beat you up if you try to litter on her patch. Little details like the necklace and the coat indicate that she might be a collector or scavenger, and her rounded face gives her an approachable, goody kind of look. I gave her very prominent eyebrows so that exaggerating emotion would be easier.  DSC_3498The character turnaround-The side view could do with some work, but otherwise I am happy! Unfortunately I do prefer her design in the early sketches, but hey ho.

DSC_3495.JPGIn the sketch above, I am trying to show a bit of overlapping motion. As her hips move to the right, her coat and ears swoosh to the left. I am also using this drawing as a way to capture exaggeration in the face. I chose to use overlapping motion as I think it is one of the most dynamic of the 12 principles. I’m actually really happy with how this sketch turned out!

DSC_3496I tried to be a lot more free with my drawings for this exercise, instead of keeping them really refined. So if you think the drawing above looks extremely rough and ready…it was meant to be. Here I am attempting to show the use of arcs in motion. I used the nose as the point that follows the arc, with everything else following after it, and I actually think it works really well! I chose to show this principle as I find myself using it most when animating in maya, and thought it would be interesting to try in 2D!

DSC_3499Here I am attempting to show ‘Anticipation’. The idea was that she is preparing to creep along…but need up looking more like she had been pushed backwards..OOPS. But I suppose you could use that as showing ‘Follow Through’. I think the facial expression was probably what caused that problem!

Advertisements

Life Drawing-Reflection

(Photos of my work will follow, they were removed from the classroom the other day and I have yet to find them.)

Life drawing this semester has been up and down for me. Due to a lot of life commitments outside uni, and the pressure of keeping the animation project going, I was finding it harder and harder to find the time to practice and grow. However, one thing I really enjoyed with this semester was being given more freedom to use my imagination, especially in class. I particularly enjoyed the exercise where we had to transform Robert and a cardboard box into a new character. To me, this has always been an easier way of drawing. I actually found that class extremely helpful for the last assignment as well, as it meant I was able to take human poses and apply them to my character.

I can slowly see my use of form growing over time, and definitely think that I have begun to really understand exaggeration and gesture drawing (maybe a little too much at times). Using exaggeration makes drawing so much more fun to me, and also gives your drawings a real sense of movement and life, as if they are about to spring from the page. Another thing I found to be very interesting were the homeworks on drawing the hands and feet. I used Burne Hogarth as my reference when drawing the legs, and ever since have seen an improvement even in my quick 2 minute sketches of the calfs and feet.

DSC_3484DSC_3485DSC_3486DSC_3489Personally I can see my drawings becoming a lot more confident the more I go to life drawing. I used to use very timid strokes, whereas now I feel more free to experiment.

My character design is of a little squirrel. I chose this just because I thought the bright eyes and bushy tail would be appealing to an audience. I decided to go for an oval kind of shape for the head, to make her look very innocent and shy; with big ears and eyes to make her look a goofy and loveable. The ears also function very well to show exaggeration of poses, as well as the tail. Having softer elements to the character allow for different weights which can travel at different speeds with motion, creating a more interesting look than one solid block. It is for this reason that I also chose to give my character a coat.

The character is roughly three heads tall, and is rounded and weighty towards the hips. I chose this because I always find it easier to draw models that carry their weight lower to the ground. I also thought that the skinny legs and big feet would be a great way to exaggerate motion.

DSC_3497DSC_3498

 

The Retopology

It is a shame that I couldn’t get more of the head retopology finished in time-I had a lot of problems with Maya crashing on my laptop, and with the room in uni closed for shows over the weekend, it was impossible to work anywhere else. Screen Shot 2017-05-15 at 17.28.31Screen Shot 2017-05-15 at 17.28.54

Screen Shot 2017-05-15 at 17.35.43

I really enjoyed completing this section of my retopology. It’s actually great fun to work out all the angles, making sure the edges look smooth etc. Sadly I had to leave the project here…For some reason my computer started taking almost 2 minutes to load each polygon once it was selected, and the quad tool became almost unusable. I am happy with how much I have learnt with the quad draw tool, and think if I had had access to a mac in uni with mudbox that I would have been able to continue with my work and get it finished.

 

Imaging and Data Visualisation-Reflection

This module has been a massive learning curve, and has been extremely enjoyable. The Floating City project was a daunting task at first (and throughout) but I would happily say that if given the same task again now, I would feel quite comfortable with it. Mostly this is through learning by doing, and making many, many mistakes. But it’s better to make these mistakes now and learn!

Starting the project, I was new to modelling, didn’t have the foggiest how to render and only knew a little about editing. Now, following the tutorials given by Alec, I would feel comfortable to tackle all of these areas myself. I think the New Narratives module works really well with this one in creating a solid knowledge of workflow, which to me is one of the most important things to master. I particularly appreciated the class on how to keep a tidy Maya workspace and how to send the files in the best way to team members-if not for this class, I am sure the floating city project would have ended up even further from finished!

In reflection, I wish I had had more time to play about with the various rigs that are provided each week by Alec. With all the other projects on such a tight time schedule, I only got to use them for the short class each week before getting on with the rest of the work. In saying that, each class provided me with so much knowledge that I was able to use in our 15 Second Animation (creating lattice rigs, blend shapes, the 12 Principles). I would love to go back and look over my notes on how to create and animate arm rigs using the method of overlapping motion and work on that over summer. My teams animation didn’t require arms, so it would be fun to work with something completely new and expand my knowledge further for next year!

The head topology task is definitely a tough one, and one I honestly felt very unprepared for. Apart from being show some sculpting tools in Maya, I had never seen any of the techniques, tools or menus being used before. I am still struggling to get my head around the retopology part of the task, but hopefully with time, I’ll get there!

Even if I don’t manage to complete the retopology in time, I hope to figure it all out over summer so I can come back knowing what to do!

Starting to Sculpt

So, possibly naively, I thought that Mudbox might have the same menus and set up as Maya. Of course not.

The process began by just getting used to the navigation, learning how to add image planes as a reference, and how to use your tablet or mouse in the best ways possible. Thankfully, Mudbox does come with some handy videos to help you get started.

I began with the default head model and worked from there…soon finding that the androgynous shape of the face actually make things quite difficult to sculpt. Given more time I would love to play with the settings in Mudbox more, and create a far more realistic representation of Yazz (my model is far from looking like her). If you block out the eyes and forehead though, I don’t think it’s that far wrong!

The part I found most tricky (and due to time ended up slightly giving up on), was the eyes. Creating a rounded surface to add the eyelids to was harder than expected, and getting the angles right between the brow bone and eyelid was an even harder task. In particular, I kept finding my mesh ‘creasing’, which causes a lot of problems.

An Early Attempt- I found that by the end, the head had been too overworked…the general structure here is a lot better.

Screen Shot 2017-05-14 at 00.29.09Screen Shot 2017-05-14 at 17.15.14

Having fun^ Dent head…oops.

Screen Shot 2017-05-14 at 22.21.35

The ‘final’ (but accepted as ‘not the best I could have done’) head.

Screen Shot 2017-05-14 at 22.23.53Note-No time to finish the ear!

Screen Shot 2017-05-14 at 22.29.36

^This bit actually looks almost real! Whaaaaat.

 

The Head Model

Some research into what this should look like. Not expecting mine to look anything like these ones!

For this task I will be attempting to sculpt Yazz’s lovely face in Mudbox, before transferring it to Maya in order to ‘retopologise’. We have to alter our model until it has under 10,000 polygons.

So what is Retopology?

Retopology is the act of recreating an existing surface with more optimal geometry. A common use-case is creating a clean, quad-based mesh for animation, but it’s also used for most any final object that needs textured, animated, or otherwise manipulated in a way that sculpted meshes are not conducive to.’ – DR.Google.

What does ‘Autodesk Knowledge Network’ say?

‘Retopology lets you create new topology based on the features of a reference surface. Maya’s retopology toolset offers a streamlined, one tool workflow that lets modelers focus on the creative process instead of edge flow and polygon count.

The Quad Draw tool gives you complete control over the retopology process, providing you with tools to quickly create new topology and refine it on-the-fly.’

To retopologize a mesh using the Quad Draw Tool

  1. Set a reference mesh: snap the Quad Draw tool to the mesh you want to retopologize.
  2. Open the Quad Draw Tool by doing one of the following:
    • From the Tools section of the Modeling Toolkit window, click .
    • From the main menu bar, select Mesh Tools > Quad Draw.
    • From the marking menu, select Quad Draw Tool. (To open the marking menu, Shift + right-click in the scene.)
  3. Drop dots: define vertices.
  4. Create polygons: make quads and fill triangular shaped holes.
  5. Refine polygons: clean up your new topology.
  6. When finished, click  again to exit the tool.
    Tip: You can press Y to reactivate the Quad Draw Tool after switching to another tool.

The Final Two Animations!

The First Draft

Things we knew about-

  1. Lid bounce had managed to go out of sync/stopped completely in places.
  2. Quality had to be reduced to 720 to get it rendered in time-This will be changed
  3. Lights and camera settings were grainy and noisy
  4. Sounds were a little off, had to be done very quickly before the deadline-We want richer sounds with a LOT less bass.
  5. Some of our camera angles don’t make you feel like part of the action.

Feedback-

  1. Could change the shape of the bouncer (big shoulders)-Tried before…looks a bit off.18387233_10211579704624808_257523284_n
  2. Lighting needs altered-White Light between colours- Altered by adding Fog and a ‘Moonlight’ spotlight. Box lights now only 2 area lights.
  3. Lid bounce out of sync-Was already being changed
  4. Rope Barriers outside club?- No space for the action to take place…also for realities sake, where would they have found a perfectly proportioned rope barrier? Barbie set?18405634_1535857119779002_800999767_o
  5. Include Fog/Glass Box?- DONE
  6. Timing and camera Issues-RESOLVED…we hope.

The Final Edit 

 

 

Lighting

Originally, our team was a little in the dark with lighting (ha).

The Original Set Up-Edit One

We had our whole scene lit by an Arnold Skydome light, which cast an even but unrealistic light over everything; taking away any opportunity for shadows and interest. We didn’t make a wrong choice in doing this…it just wasn’t dynamic in the way we wanted.

Our original studies and sketches show a beam of moonlight flooding into the room and lighting up the box, with lovely dusty light, but at last minute, this just wouldn’t have worked with our render time.

In saying that…neither did the EIGHT LIGHTS that we had put into our Box nightclub. In 1080p renders, our eight lights did look fabulous though…but only when they weren’t interfering with the animation and creating noise.

The Box Set Up-Sarah created our box lights-which would then be imported into the final scene. They consisted of eight different coloured spotlights, each with a cone to the back to prevent light from spreading past the desired area. The area over the ‘Windows’ was then covered with an ‘AiMesh’, which created the ‘disco ball’, ‘clubby’ look. It really does look beautiful when it shines over reflective surfaces and over the faces of our characters…but was just too complex and confusing in the final close up scenes.

Another thing that we found out later on is that you can save your renders in one of two ways…RAW, or Colour-Managed. 

For the first set, I think we were working with Colour-Managed, while RAW would have been better (darker). (We didn’t know about this setting at the time, as well as camera settings).

For the second set of renders, we chose to go with Colour-Managed saves, as it was a little brighter in the shadows, which we could darken down in post. We couldn’t undo the crushed black look we got with RAW files in after effects.

This slideshow requires JavaScript.

To resolve the lighting issues we had in the first edit, Clare and I spent the day researching lighting-in particular, how to create a dusty moonlight look. We also looked into the best way to create the box lights, without over complicating the scene and creating white light.

Researching the moonlight was easy…applying the moonlight was tricky. Clare and I wanted it to be obvious that it was moonlight, and not sunrise. We needed it to be blue and cold, but not too dark or ominous. The room had to seem foggy/dusty, but only the right amount…too dusty and you couldn’t see all the lovely models that everyone had made…It began to look more like the club was on fire and the room was filled with smoke. Not a good look. Over and over we kept finding that we couldn’t get the scene to brighten up…only to realise about an HOUR later that our screen brightness was way down. FACE PALM.

But in the end, with much fiddling with fog densities and spread, we got to a happy medium of cold and cosy.

 

 

Animation

With the models made, room arranged and characters rigged and textured, it was time to start the animation!

This part of the process was simultaneously great fun and unbelievably tedious. Having the most knowledge of how the rig worked, Clare started the ball rolling by animating the first few scenes (checking for any random glitches etc), while I worked on the simple animation of the trap door closing, the box lid bouncing and the mobile slowly spinning. I love that the two of us are always on hand to help each other out and give critique/advice on the best way to do something.

With the first camera pan already rendered, I also started to design the ‘Fitting In’ title on after effects, while Sarah and Holly thought about different ways we could make the end credits interesting (adding little drawings from each of us as a logo). I would have loved to put more time into the title, its’s my kind of thing. With the time we had, I just ended going for one of my favourite fonts-a very simple font called ‘Peach Milk’-to give a paper cut look. I think it also reads really nicely as a description of the story, the ‘In’ is different to the ‘Fitting’ but it can still fit in…the difference is part of the harmony of life. Yes, that much meaning in a typeface. Graphic design, boom.

Screen Shot 2017-05-13 at 14.14.03

The plan – We would have the room completely animated so that the basic animations listed above stayed constant. Then, each set of character animations would be imported into that working scene at their specified frames and saved as new versions (frames###-###). Unfortunately we must have got an old file and an animated file of the room mixed up at some point, resulting in the box lid stopping it’s bouncing at points 😦 By the time we noticed, it was too late to restart renders.

Clare then blocked out the beginning and end positions of the characters in each section of the story, and set the OUTLINE file to each of us. From there on, we could work on our own section of animation, knowing it would at least roughly line up with the next section.

Clare-Opening Hop Up, Dance, Cylinders Jumping In

Me-Cube Hop In, Crash, Fall Back, POV shot

Sarah-Get up and Turn Around

Holly-Sad Walk and Turn

Yazz-Lid Pop, Turn, Jump, Freeze Frame

Lots of squash and stretch used in all of our animation to give it a very traditional feel. We made sure to follow the 12 Principles of Animation-Blog Post.

Screen Shot 2017-05-13 at 14.10.46Screen Shot 2017-05-13 at 14.10.54Screen Shot 2017-05-13 at 14.11.23Screen Shot 2017-05-13 at 14.11.44

Some shots from my section above^ Look at those little concerned faces!

 

Rigging and Texturing

Rigging

Clare, our resident rigger, did an amazing job rigging the characters that we had all sat and modelled together. She used a simple lattice rig, with cluster handles at the head, belly and feet, with expression controls for squash and stretch on each one. Very similar to the Jonas rig (I watched the process many times to learn from her).

We did have a lot of trouble with the rigs to start off with. Having rigged the characters, we didn’t realise that if you changed their UV map that the rig would no longer apply. While Clare had been testing out rigs, I had been designing UV mapped textures on another model…and when the two were brought together, weird things happened. We thought we were being time efficient! BUT, in the end, the two came together beautifully to create the squishy, squashy little characters we now know and love-Shout out to Clare who must have rigged each character at least 5 times-STAR!

I’m sure Clare will have loads on her blog about the rigs, so I’ll leave a link here.

Textures and the Facial Rig

It was actually great fun learning to UV map the characters, and then paint over the maps to create a custom texture. I had also worked with projection maps to add the faces onto the characters, which was fun! The only issue was getting the face to scale correctly, which could be a bit of a nightmare. In the end I stuck with this option though, as the tutorial for creating 3D eyes was a very old one and seemed to have a lot of issues (Of course I found this out 3 hours into trying it over and over again).

No matter how hard I tried, I just couldn’t get this tutorial to work. It has a lot of grouping and parenting issues which I couldn’t resolve in the time frame.

My attempt to follow this tutorial^ I was only one step from the end when things started to go wrong…Maybe it just needed more time…But in the end, I loved the 2D faces so much more!

In the end, I used this tutorial to learn how to add the faces to the texture of the characters. Unfortunately, due to time constraints, the expressions we could use in the animation were limited…but I did design and create expressions for 9 different faces on each character. These were set via an added attribute in the channel box of the top nurb of each characters rig.

^Some of the earliest tests, with quick doodles to check that the placement of the maps worked out. Note-Encountered problem with the UV map ‘Sticking’ to the background of the editor…possibly problem with laptop, but was extremely frustrating as I couldn’t move the UV around make painting in Photoshop easier. 

Cubeeyes.001eyes.001triangle

 

 

 

 

 

 

An example of each texture map-there were nine of each, with different faces on each one. This was then used as an ‘Image Sequence’ in Maya.

The ‘wood texture’ texture was based on my original character rotations. We wanted something that looked illustrated, and not true to life-just to differentiate them from the world around them, and to give the whole thing a ‘handmade’ feel.