DOWNLOADABLE FILE HERE (images included, not attached)
What would you do differently?
If I could go back and do this assignment again, I would dedicate more time to it and study more. I didn’t do as much as I’d wanted due to other projects involving modelling and animation that took up more than 50+ hours of my time a week. I believe I would’ve gotten far more done and to a higher quality if I was able to spend more time on the work.
Another change I would make is to take 3D animation 2 before this class, it felt like I was missing a fair bit of information going into this class that other students had over me.
As for tasks I went to high poly modelling too early, instead of working the proxy. I had issues with too many vertices and maintaining loops when working from a high poly mesh, outwards.
Highlights
The thing I’m most proud of in this work is the face and how it animates, but the face was already showcased in the video below so it’s not seen as clearly in the final render video:
CURRENT PROGRESS VIDEO ABOVE
Click here for the process post
Schedule for the remaining weeks:
Week 11: Continue work on the production model and reattach rig
Week 12: Clothing and other props modelled on
Week 13: Unreal export
Obstacles:
This model is my first attempt at creating a human character and a facial rig, and since it is I ran into a few problems in the procedure of how to make the head. Following a tutorial I’d been previously using only led to poor results (as seen in the process post) that needed to be scrapped completely, which wasted a fair amount of time since it took me so long to realise the model wasn’t acceptable.
I’m also still running into a workload issue with other classes and this one, but with enough time management I should be able to find an easy solution to juggling my work.
Critical Tasks:
If I run out of time, ideally I’ll have the rig properly attached and all pieces of the body connected. To achieve that currently I’d only need to reattach the rig to the head and then the head to the rest of the body. From there it isn’t too hard to showcase the rig and what I currently have by forcing myself to figure out unreal exports instead of progressing further on the model.
Started with the proxy model and simply connected up the joints in the simplest way possible, following the provided tutorial video for the unit. I planned to come back to the mesh for the joints after to make sure they bend correctly. This took an hour.
Fixing the hands took me an hour and twenty minutes, since I had trouble navigating the higher poly count of the hand versus the poly count of the wrist. I currently have several triangles on the hands and aim to do some more research to make sure any triangles are in the correct position or removed altogether.
Modelling the base of my first attempt was a flawed process, as I attempted to use a basic tutorial that featured a cartoony character being modelled rather than a realistic one. The first attempt to model the base took me an hour and a half. I used my own sketches of Zeal as a loose reference but made the mistake of not including real women to reference.
After I’d created the base it took me a further two hours fiddling around with the flawed mesh I had to make details like teeth, ears, and eyes. It all ended terribly and can be seen below.
After showing the face to several people and being unsatisfied with how it looked, I scrapped it completely and went searching for help with my work. I found a handy tutorial you can see here that went step by step through how to model a human head. I also made sure to use real human references along with my sketches to keep it anatomically correct or at least near correct. It took me two hours to do the process seen below.
After doing the base, it took me a further three hours to do more details such as the brows, teeth, inner mouth, ears, and so on.
After completing the head, I went about working on building the face rig as I believed it was going to take the longest. I filled in the shape keys following a process seen here quite easily and went about setting up the facial rig to control the face keys. The shape keys took me two hours, and the rig + custom shapes took me two hours also.
Compiling the video to show took me half an hour to quickly animate and get into premiere.
Schedule for the remaining weeks:
Week 10: Begin Production model
Week 11: Experiment with facial rigs, implement facial rig
Week 12: Clean up rig and implement a production rig
Week 13: Texturing and UE4 export
Current Progress and Problems:
Compared to the last progress update, that being the milestone, I’ve made no modelling or rigging progress over the past holiday week for this subject. During the week I did spend time practising my modelling on hard surface models for a game studio 2 class. The table pictured below took 3 days to fully complete due to problems I had with moving to high poly too fast and learning how to use rigs in programs outside of blender.
Practising with this helped re-establish the importance of following the workflow and planning ahead. With this knowledge I started doing some character sketches of Zeal to better learn how her face looked and the important blocks of her body shape.
Beyond these sketches and practise with other hardsurface models, I was unable to complete further work on toolkit due to the demands of other classes. One class had a plan scheduled to take over 48 hours during the week, which restricted my time for other subjects.
With this week, work should pick up once more for this project and follow the allocated schedule properly.
Time Spent:
12 hours 40 minutes - Creating the model, rig, and animation
40 minutes - Writing the reflection and progress post
Process:
The process is available here
Reflection:
This semester is the first time I’ve worked with rigging and modelling a full character without using assets, and it’s been a real learning experience. I’ve taken a while to complete the basics and feel I still could’ve done more with this milestone, but ran out of time and was unsure where to go from the proxy that wasn’t just to the production model and rig.
I’m happy with how the rig and model came out, but worried about my time management. While I am on time and schedule with the last progress check’s timeline, the work I have done was mostly completed in the last two weeks rather than in the assigned weeks and I haven’t set up enough of the rig and model to complete the assigned tasks I’d written for week 10. The idea now is to work through the holidays with more research and pre-production to ensure the project will stay on task for the next university week.
I’d also like to create some more fluid animations to better show the rig, instead of the blocky ones made for this milestone. This will need to be done at the end, so time management is crucial.
Detailed below the cut is the process that went into making the proxy geometry and rig for milestone 2A.
Using my references for Zeal’s concept and human anatomy, I started working on the model. I spent an hour working on my first proxy attempt, before realising I’d used too many loop cuts and started with too high a poly count. I scrapped this attempt and started again.
The lowest poly version of the proxy took 2 hours to complete, which I did using the provided proxy timelapse from blackboard for reference.
I then went through editing the proxy’s geometry and adding some loop cuts to make the form more clear and less rectangular. I also cleaned up some of the topology to better suit the anatomy. This took 30 minutes.
I continued doing this, adding more cuts and fixing the positioning of forms to better suit the rigging phase. This took an hour.
After this I spent an hour setting up the basic rig, using some of the time to watch reference videos and make sure it was set up correctly.
A further 3 hours was spent setting up the IK rig modifiers and setting up limited rotations and locations. I also spent this time fixing up some of the rig’s locations so the IK controls functioned correctly.
Custom shapes were added along with further limit fixes, this took 2 hours along with tutorial and reference checking.
2 hours were spent doing basic animations for the proxy rig, including a walk cycle, an animated pose, and a spin cycle. The animations are slightly off and not as smooth as I’d like, but with it being a test of the rig it’s a fine proxy animation. 10 minutes were spent compiling the animation videos into the main one shown in the milestone post.
Details: “Series 01: Blender Character Creation” by Darrin Lile, 2017. URL: https://www.youtube.com/playlist?list=PLyelx0TsmSpf-8xZfjZveokSsIwj0F3lo
Summary: This series details how to create a blender character to work into unity and video games. Following this tutorial has helped me create several characters now for practise and helped give me some knowledge on rigging and modelling to further look into those streams.
This experiment was also for my IGB200 class, in which I’m the animator, and I wanted to test myself to help practice for KNB221. I created a mannequin for my game and started practising habits that I’d need to use in my final assignment for this class.
The mannequin in this game needed to be poseable, but wasn’t going to move in the game, so I focused on making the mannequin look real enough but not making it flexible.
I opened up the blender file and got a turnaround as a background image, from there I started modelling the mannequin.
After doing some basic forms, I worked on the ball joints, which I made by having each joint have an indent with a space for a T extension to roll through it.
After completing the joints, I put them in the correct locations (pictured below is just shoulders and hips, in the final version there’s joints at the elbows and knees).
I then continued with modelling the mannequin, trying to practice removing ngons and triangles as I went. It didn’t go as well as planned due to not thinking through a few of the fixes I tried to use. I then realised that it wasn’t important to remove them because the mannequin was a hard surface model.
After 4 hours of modelling, with my attempt at removing ngons and triangles, I had the rough model I wanted. I took another hour to fix some parts of the model I wasn’t happy with and produced this end result:
I did this project just after choosing the current stream project, in an attempt to follow a tutorial and create a character my friend owns, named Judas. Unfortunately because of when I did it and because of why I was doing it, I did not record any wip images or how long it took me to complete the model and rig.
During the process I ended up dropping the attempted IK rig for the arms, as it broke the shoulders and lower arms and despite trying, I couldn’t fix the problem. After doing more research and looking back after a week or two, I’m fairly sure I hadn’t set the bones up correctly and had done no retopology to correct any mistakes.
Below is the final model and rig, along with a progress shot of when the IK rig broke.
(above: Rig)
(above: Model)
(above: here is where the IK rig kept breaking, twisting the arms and shoulders uncomfortably. I could not find a fix for this, and ended up turning off the IK rig and animating each bone individually)
(above: pose after turning off the IK rig)
(above: pose after turning off the IK rig and adding an eye tracker)
My chosen stream for assignment 2 is character animation, the modelling and rigging brief, using Zeal from the polycount forums:
To achieve this I’ll be using Blender, a program I’m familiar with, to model and rig Zeal. I’ll also take the model to different sculpting programs to work on it, so far I’ve not decided on the program I want to use for this but may choose Mudbox or zBrush.
Before I do any work towards modelling and rigging, I’ll need to do some studies based on Zeal’s concept art. I’ve already started these studies with a quick skeleton sketch and shape layout.
I still plan on creating turnarounds for Zeal, or finding an anatomical reference similar to her build, that I can model from.
To start the process I’ll study and practice creating accurate topology when modelling, and what models are usually used and how (Proxy Geo, low poly, high poly). I’ve already started saving tutorials and references on how to properly do geometry.
https://au.pinterest.com/pin/414331234451252720/
After researching and practising, I’ll start work on Zeal’s model by first doing her anatomy, hair, then her basic clothes (tighter, underclothes, etc), over lain clothes (tassels, sash, etc), and accessories (swords, bracelets, etc).
After creating these models I’ll start more research on rigging complex organic surfaces, since my rigging knowledge is limited to hard surfaces. I’ll also need more practice in this and IK rigging. After practising these I’ll start setting up the rig for Zeal. I’ve done some basic rig work before and after studying and practising I hope to be more confident and efficient with it.
https://au.pinterest.com/pin/564075922056906924/
To my knowledge, after this I sculpt the model and work on higher res/poly models before then texturing it. When the model and rig is done, I plan to animate it in blender and take it into unreal where I’ll build a basic scene and place the animated character inside it.
For this assignment I don’t have any idea how long doing all the research and learning will take, since I have some base knowledge but don’t know many of the details. I’d prefer to dedicate a week to each learning period to allow myself time to work on other projects that require this information (essentially practising for this) and take time to step back from my project and judge it objectively.
Week 7: Research and concepts for Zeal
Week 8: Modelling low poly/proxy
Week 9: Research for rigging, basic rigging
Week 10: IK rigging, facial rigging
Week 11: High poly modelling
Week 12: Texturing, touch ups and fixes, Animation
Week 13: Unreal import, scene render and editing
Ideally I’ll try to do work earlier than planned to allow more time for research and problem solving.
Throughout the three milestones that were completed over the last three weeks, I’ve learnt a lot about Unreal and Maya and how to work with them. Before the milestones started, I already had, at least, a basic understanding of animation and had developed a few animated scenes in Blender.
(Video above: “Test Animation Walk sequence” by Madison Muir, 2017. URL: https://youtu.be/0wytRX6-82s)
I didn’t have any knowledge of Maya or Unreal however, and had even thought Unreal was just a game engine like Unity. I still have a way to go with both programs, I’m hoping to learn a lot more about Maya and how to integrate the knowledge I have of blender over to it. I’m also aiming to learn more about using After Effects to I can make my final product look as polished as possible.
FOCUS
For the next assignment, I’ve decided on going with Character Animation as my specialist stream with a focus on modelling and rigging one of the concept characters shown below:
I’m an animator before a modeller, and rigging is still something new to me. I chose this task instead of the animation one so I can learn more about modelling and rigging, something that interests me and could open more opportunities down the road.
Areas of Priority:
Using Maya for character creation
Using Proxy Geometry
Modelling hair and clothes
Facial Rigging
Fixing broken rigs to prevent broken meshes
Personal Objectives:
Model the human concept with at least a basic facial rig
Create the clothes and hair with physics simulation
Animate a small action scene to accompany a rig demonstration
INDUSTRY RELEVANCE
The chosen task and stream are relevant to quite a few jobs, but one of them in particular is that of a Character Artist for video games. A Character Artist is someone who models characters based on concept art, they can also be called 3D modellers but that role can also include props and environments. Character Artists are specifically focused on creating the people in the product they’re working on.
(Image property of Victor Petersson: http://polycount.com/discussion/160600/characters-for-character-artist-portfolio)
A character artist may not work with rigging, but in some cases 3D modellers will when a studio is smaller.
Required Skills and Responsibilities
Within all jobs in both the film and game industry, employees are required to have communication skills and the ability to maintain their own workflow. This is a basic requirement, as the jobs require a team of people to work together and share work effectively and efficiently. They also expect employees to be able to better themselves and learn, therefore increasing their quality. A video covering that is here:
(Video above: “ 5 tips to boost your career ” by Bobby Chiu, 2017. URL: https://youtu.be/FRR9BFYaL_Q)
Within this stream of animation, a lot of importance is placed on having knowledge about the foundations of art and being able to draw from life. This information is essential to making realistic and anatomically correct models and rigs, which need at least a basic understanding of how the human musculature system works. Other requirements can be summed up as knowing how to use the required programs to create and optimise the character and rig for use in animation. A job requirements listing can be found here: https://creativeskillset.org/creative_industries/animation/job_roles/366_cg_modeller
The job also comes with responsibilities:
Collaboration with other members of your department (concept artists, animators, directors...)
An understanding of meshes and rigs and how to create them
An understanding of UV wraps and how to optimise the layout
An understanding of animation requirements
The ability to work to a budget and follow directions from directors and leads
A working understanding of the primary modelling program
Current and Future Skills
My own set of skills in this field are partially developed, and I mostly need to work on the technical aspects of my skill set. I’ve already been studying art fundamentals and been familiar with them for approximately 7 years. This includes attending life drawing sessions.
(Image property of Madison Muir, 2016. URL: http://feveranimation.tumblr.com/post/152245012995/life-drawing-sketches-from-week-12-model-alana)
I also have some knowledge of modelling, rigging, and animating 3D characters. A video of previous animation work I’ve done is earlier in this post. Below is a render of a character I created in blender.
(Image property of Madison Muir, 2017)
Most importantly I understand how to work in a team with other people being reliant on my own work. I can communicate effectively what I require from other team members and meet my deadlines with what they require of me.
There are many things I can still improve upon. Studying art foundations never really ends, and further study improves the quality of work that is produced. I could also work more on my composition and level balancing, among other things related to final polish and unifying a piece. My technical knowledge of modelling and rigging is also limited, as I’m only familiar with Blender which isn’t a program used in professional workplaces. My rigging also needs to be worked on as the one I have made in the image above is incredibly stiff and prone to deforming the mesh incorrectly. There’s a lot I need to learn to work through this assignment and ready myself for the industry.
Starting Points
(Video above: “Maya Character Modeling Tutorial - Cartoon Mouse HD #1“ by Edge-CGI 3D Tutorials and more!, 2014. URL: https://youtu.be/2dLMctUGKXw)
This video series and the channel contain information on how to model characters in Maya, which I’ll need to learn so I can begin to understand it. As I’m unfamiliar with Maya, I’ll need these beginner’s tutorials.
(Video is a part of the playlist: “Introduction to Maya Rigging” by Jamie Telford, 2011. URL: https://youtu.be/FHcVNBweIo8?list=PL9EF913A8175DD2EC)
This is a beginner’s video series on how to rig in Maya, it includes videos about primer rigging, bipedal rigging, and some misc rigging. Using this will start my understanding of how to transfer my knowledge of rigging from Blender to Maya, and then develop that knowledge so I can make better functioning rigs.
(Video above: “ Creating Proxies Tutorial - Autodesk Maya 2016 “ by AC Animations, 2015. URL: https://youtu.be/kr9Krpq8sGc)
This is a video about proxy geometry and how to employ it in Maya, which I’ll find useful to see how to create it within a 3D program and how exactly it functions.
Time Spent
1 hour animating in maya
1 hour and 30 minutes animating and building the scene in Unreal
1 hour for editing in After Effects
30 minutes writing the reflection
Process:
I spent the first hour animating in Maya, doing a scene where the character, Muscle Mann (the ‘mann’ is not a typo), did a spinning kick and then posed heroically.
The next thirty minutes were used setting up the scene and importing the animation into Unreal. For this last milestone I wanted to go further with my environment, so I made a small enclosed courtyard with some shapes inside to add some interest in the background.
I also wanted to try something new, so I added some lights and then changed the materials of the objects I wanted to be spotlights or fluorescent strips so I could mask them out in post. I also changed the materials of the character so he was easily distinguishable from the background. This change and the following animation took an hour, partially due to a crash that wiped the camera animation halfway through.
I animated the camera to spin around the animation, making the animation also pause at certain poses to do small introductory screens to Muscle Mann.
The next hour was spent in After Effects editing the animation to add all the effects I’d learned. I started by adding the light effects so that the spotlight and strips glowed.
I then added a composition that had titles that went across the screen with text, to introduce the character.
I did some colour correction and animation on the final scene of the composition, to add some punch to the name reveal.
The last thing I did was add some fog so the skybox and corners would be darker than the rest of the scene, bringing out Muscle Mann.
Reflection:
This week I employed everything I’d learned from the last two weeks of milestone work and experimentation to make the final product. I did have to google a few things such as shortcuts but most of it was information I’d gathered from previous works. Doing this revealed a few things, one notably being that this milestone’s timing is a bit off and could’ve been snappier which probably could’ve been fixed with a storyboard or time chart. I’d also noticed that some methods I’d used to make the light and title effects felt like they took longer than they should, so finding faster ways to do masks and effects would be ideal for my future works.
This work did turn out much better than my previous milestones however, due to more time spent on it and more knowledge on how to make it. There are some things I’d have liked to fix like that hard lines on the fog and making the ending hold for a bit longer, some of the animation could have also been smoother but that can be blamed on the lack of any storyboards or timing charts.
The thing that pushed this milestone further was the colour correction, the two lit colours that contrasted was a nice addition and using the effects in After Effects added a level of polish that wasn’t in the other milestones.
With this experiment I wanted to test myself to see what kind of movement I could get with the camera, using an overwatch highlight reel as reference. I also wanted to do some simple scene building within unity, since it’s currently something I dislike doing and would like to have slightly more practice in.
Reference Video:
Setting up the initial animation of the camera and the cylinder (replacing the character) took 25 minutes.
Building the scene around the animation took 10 minutes.
The effects and post processing took 45 minutes, since I employed more effects than I had in previous experiments with a dark fog and an outside light. It required a lot of editing of render passes to get the luma trackmat to work.
What this experiment taught me is that I need to work on my scene building more. I had trouble making the room look interesting enough while still producing useful base image render passes, I had to return to unreal and render the images again after screening out the door so I could mask it properly. I did learn that I have at least a basic understanding of how to animate a camera and object to match a reference video, and with more time I could’ve matched it more exactly with timing.
Details: “Maya Animation Tutorial for Beginners” by OnlineMediaTutor | Maya modeling and animation tutorials!, 2016. URL: https://youtu.be/rNkJ-OEO7qg?list=PLsPHRLf6UN4kAt46X0_Bw1IuOQ7Q1tvCS
Summary: This video and its playlist are a collection of Maya animation tutorials. I’m required to watch the series or at least take some notes for relevant information so I can use Maya right. Currently I have no experience in Maya at all, and have trouble copying key frames. After watching this video series I’m hoping I’ll have enough knowledge to do my project animation at a higher quality.
With this experiment I wanted to test how the lighting looked and animated in Unreal, since I’d already had experience with lighting in other 3D software previously. It was also a test for myself to see if I still knew how to do three point lighting.
First I set up the scene in the Unreal engine, placing a cylinder and removing all the other components so I could see how lighting affects a simple object. I followed the image below for lighting reference, but replaced the back light with a high intensity rim light instead.
It took me 15 minutes to set the scene up and place the lights how I wanted them.
It took me another 10 minutes to set up the render passes, and then change the colours to produce another set of passes.
I then wanted to try some more animated lighting, so I keyed the lights to rotate around the cylinder. I rendered that animation, and then added keys so the colour of the lights would change over time. This took about 20 minutes, due to me also doing some basic google searches for shortcuts I’d forgotten. The scene setup overall took 45 minutes.
Finally I opened After Effects and edited the render passes together with a black background to make the lighting more obvious. This took 15 minutes.
I did notice in my final renders that a white outline appeared in After Effects, which may become a problem if it wasn’t due to some incorrect trackmats or choppy edges. What I did learn though was how the lighting appears in Unreal, which is a lot more neat and clear than other editing programs I’ve used in the past. Another experiment to do would be to add glow to edges of objects to aid the rim lighting.
RENDER PASSES:
Each pass for the different scenes were near identical, as only the final image was changed.
Summary: This video details how to open and use the orthographic views in the Unreal Engine. I’m using this so I understand how to navigate the 3D view more effectively, which is important in all fields to learn your tools.
