Week 1 – Introduction to Maya & the bouncing ball

In this first week we were reintroduced to Maya. In class we were shown the animation workspace in Maya, became familiar with the playback controls, project settings, how to set up Maya scenes and projects, as well as how to create playblasts.

Although I had done some basic key framing before it was nice to have a refresher and learn more about achieving higher control of animation by using the graph editor and motion trails.

The first assignment we have been given is to create a bouncing ball animation. Although simple this tasks asks us to think about important principles of animation such as timing, weight and motion.

I first made a ball fall normally with no weight or squash and stretch added

After this I added the squash and stretch, this is most likely the wrong way to go about the task and was different to the workflow we were shown in class, the reason I did it this way was simply because adding the squash and stretch was the aspect I was most intimidated  by and I therefore wanted to experiment on it first. In future now I know how to achieve the squash and stretch. I will aim to make the basic vertical and horizontal movement first before adding finer details to the animation.

I then moved to the horizontal movement adding a motion trail to help me control the arches. I also had to work on adding in the squash and stretch keyframes aiming for the idea that every time the ball touches the ground it squashes slightly and when it gain momentum in the air it stretches out

This is the animation without the motion trial to enable clearer visibility.

I am happy with this as a first attempt however I know I could do better with more refining and care taken. I aim to improve this over the next week or so.

Finding reference for the ball bounce

I really like this video for reference of different kind of ball bounces, it uses a variety of different sizes and weights of ball to showcase the different results of the bounce. The ones that inspire me the most are the tennis ball and the ping pong ball. I think the light nature of the ping pong ball is really interesting as it rises quite sharply and make sweeping high curves. The tennis ball is a pretty uniform bounce but I think it is a good size with a good way of averaging out the arches and the height of the bounces. When I come back to this bouncing ball exercise I will try to reference this.

Week 2

This week we continued to work on the ball bounce. We recapped the Maya animation workspace and I worked a new version of the ball bounce using a smaller ball, keeping the reference of the ping pong ball in last week’s video in my head. I needed the refresher in controls so starting a new project was a good way to achieve this, I also learnt more about keyframes and using the time slider to add more or less time between movements. The troubles I had this week were mainly rotation and timing, rotation is something I struggle to figure out in 3D space but I think I’ll get better with practice, and timing is something I tinker with once the bulk of the movement is in place. The main thing I am learning is how much observation comes into play with animation.

 

After looking at this result myself and gaining feedback I know what I have to do to improve this animation which is primarily giving the ball more time to be on the ground when it hits, the allows for the anticipated rise to have even more impact.

We were then shown a rig of a bouncing ball that had squash and stretch built into it, this was really interesting for me as I had been struggling to get an accurate squash on stretch on my own before seeing this. This is because the ball would squash and stretch from the middle whereas in actuality the bottom of the ball would not be impacted to the same degree as the top of the ball when it hits the ground. This is what I had been trying to figure out in week 1 however the rig was the first time I had seen a solution

Seeing the rig really inspired me to try and create one on my own so I did that using a relatively simple YouTube video, I’m very pleased with this result however I would like to learn more about how to manipulate it to perhaps make a smaller ball as now it feels like the plan of my animation needs to change in order to accommodate for the size of the rig, which is okay just not a change that I was expecting. My favourite thing about rigs so far is that they organise the control you have on the object each handle or controller, only controls one type of movement, for me this makes keying and animating simpler and it feels easier to understand.

The final thing we were shown was an example of an obstacle course that was built for our balls to interact with, this is really interesting and I’m excited to get on to animating something like that as it seems like a fun challenge to animate object interaction. Unfortunately I didn’t get to that in the lesson today but I will aim to get to it soon.

Making a ball bounce using the new rig

My first task was to create the basic movement of the ball

After this I added squash and stretch to the ball as well as figured out timing. This took a lot of trial and error but I think I have it in a good place now. One thing that really helped me was using the graph editor and this experience has made me a lot more comfortable with that tool.  I really like the amount of information and control the graph editor gives me as well as being able to see all my keyframes for all of the different elements. Timing was a lot easier to figure out once I got used to using the graph editor.

Along with timing the things I struggled with were figuring out where to put the stretch (initially I had included it on the ball ascent as well as the descent), how to gradually speed up the bounces as the ball started losing height and finally adding a bit of time when the ball hits the ground to give it space to pause slightly. To give the ball more time on the ground I ended up looking at that particular movement (translate x) in the graph editor and levelling out the anchor points.

 

Adding rotation

My final task for this piece was to add the rotation to the ball. Something I learnt during this exercise was the importance of hierarchy. Initially the squash and stretch wasn’t adhering to the rotation however I fixed that my moving elements about in the outliner.

I then thought about the rotation more and wondered whether the ball needed a spin rather than just a sideways back and forth coinciding with the bounces. I liked this result however I noticed that it ruined my squash and stretch effect due to the bottom of the shape changing when it span. Therefore the orientation of the effect was incorrect, this is particularly noticeable at the end of the animation. I also created a floor plane as I realised that also due to the spin, it doesn’t reach the floor on every bounce.

I figured out a way to improve this, realising that I could also add rotation keyframes to the squash and stretch controller meaning that and the rotation control would sync up. Adding this change meant the ball now hits the floor and spins accurately

Constructing an obstacle course

My next task for this assignment was creating an animation of the ball interacting with an obstacle course. My first move was to come up with what I wanted the obstacle course to look like I came up with the idea of the ball bouncing from platform to platform in a horizontal motion eventually squishing against a flat wall and dropping into a bucket

Here are some screenshots in Maya of the obstacle course empty

I liked the idea of the ball bouncing along the different platforms and at the end coming up against a straight wall and falling into a bucket. I thought this would be interesting to watch whilst also not being too complex to animate

This was me initially just blocking out the movement and the placement of the ball

This is a render of my first attempt at the ball bouncing along the obstacle course. I used the same rig that I built before so it was the same principles. I’m glad I chose to block out the movement first as that really helped. My biggest issue with this piece right now is that I feel like on some of the platforms it doesn’t look like it is truly making contact and also some of the squash and stretch is not completely accurate in terms of hitting the platforms however the squash and stretch does match up with the rotation and I think the stylised nature and speed of the animation helps cover some of these minor faults. I will continue to work on this, but I’m very pleased with this start

Ball bounce animation, interacting with obstacle course

Ball Bounce Revisited November

Now in week 6 of the course I watched by animation back and decided that there was room for improvement so I have made some adjustments. I first lessened the overall rotation on the ball as I thought this was taking the focus away from other aspects of the motion such as the squash and stretch. I also made the squash and stretch a bit more visible by adding in some in-between frames in order for the ball to go back to its original spherical form when it’s in mid air. The last change I made was the positioning of the ball when it hits the back left platform from observing it I didn’t think the contact looked complete enough so I moved the position of the ball slightly to make that more believeable

Below are videos of both the final outcome and a version with the motion trial active

Week 3 – Exploring and making our own rigs

This week we took a closer look at animating rigs. It was really interesting so see the power of rigs and see examples of the kind of animations you can make with them.

The first rig we were given was this flour sack rig, this rig has so many possibilities and it was nice to be able to see a rig that is so effective while being relatively simple. As this was  an in class activity we didn’t have much time to work on this but I am pleased with my short animation and I think it was a good experiment.

The next aspect we were introduced to was how to make our own rigs, for this we blocked out a hand and learnt how to parent object to each other therefore creating a hierarchy of control. This was a really satisfying project as it was really nice to have a working rig that I had built from scratch at the end. Parenting was relatively simple and I was pleasantly surprised by the ease of the workflow. Of course this is a very basic model so I can imagine the complexity of more detailed rigs but this was a really nice achievable starting point for me

This is a screenshot of the rig hierarchy I really like that Maya allows you to visualise the rig in this way as I find it easy to understand and spot errors.

This was a little animation I made with the hand, although simple and not very long this felt like a milestone for me as it was the first time I had built and animated something from scratch, this definitely makes me excited to work on bigger projects.

Building my own skeleton rig and making a small animation

To take my learning further outside of class, I decided to build my own rig that was a basic skeleton.

Th next screenshot show different  depictions of the hierarchy

I then needed to think of something that I could animate the skeleton doing. Again I wanted this to be relatively simple and something I was familiar with. After bouncing around some ideas I ended up deciding to have them dance to ‘Macerena’ by Los Del Rio.

I found the audio on internet archive.

archive.org. (n.d.). 04MacarenaLosDelRio directory listing. [online] Available at: https://archive.org/download/04MacarenaLosDelRio [Accessed 19 Oct. 2023].

Although I knew the dance I did use this reference video to make sure I remembered it accurately

Raphi, D. (2022). The Macarena Dance 2022. [online] www.youtube.com. Available at: https://www.youtube.com/watch?v=rVBHH5DwYFA.

‌I then worked on the timing of the animations. To do this I added a basic ball to the scene so that I could have it hit the ground whenever I wanted the next movement to happen. This was really key to making sure my workflow ran smoothly and ended up saving me from a lot of problems later down the line

I also made this table in which I wrote out the frame number and then what movement would happen at that time.

The final animation

This is the final result of my animation, I’m happy with it as a first one and I think I did well to limit the scale and scope as I managed to complete this in a relatively short space of time, it definitely gives me confidence in my ability to work on bigger more complex projects in the future

Week 4 – More advanced rigging, forward kinematics and inverse kinematics

This week we learnt about more sophistacted rigging using FKs, Forward Kinematics and Iks, Inverse Kinematics. Rigs give artist more control over their models and also make animating easier as models can be controlled from a minimum number of controllers. This means that all animation is placed on one or two components of a scene, rather than having animation key frames on individual element, which saves keyframes from being hidden and allows for clearer readability. This workflow is also an important part of working in within a team of animations as different members will be able to use and understand your rig.

To create an IK rig we placed joints on the model, followed by IK handles and then parenting the objects.

The difference between FK and IK rigs is that FK is mainly used for mechanical movement and IK is more used for organic movement as seen in humans or animals. This is because in FK rigs, each point has it’s own rotation and works relatively independently from its neighbours. This means that they are great for precise movements and straight forward motions however capturing a movement where a lot of the muscles work together is not as simple with an FK rig. This is because with an FK rig you would have to rotate every point manually they don’t influence each other. So to create something like a hand in fist, although technically achievable is a pain staking process with an FK rig. In contrast an IK rig is a rig where the joints work less independently. For example on an arm the movement of the shoulder will automatically influence the movement of the elbow and the hand to a much higher degree. Making a fist with an IK is much easier as the fingers can bend more smoothly and if the IK handle is applied at the wrist the fingers can be bent together from that origin.

Cup Grasp with IK rig

Both have distinct uses cases so it is important to analyse what sort of movement the end result will be when choosing which kind of rig to use. More natural movements use IK rigs, for mechanical movements, use FK rigs

A tool that we were introduced today was using ghosting tools in Maya, this toggles the visibility of the frames surrounding the current one in a translucent form, this means we can see where our animation is going before and after this. This can be useful as it provides more information to the artist to fine tune their approach to animating the current keyframe.

Animation without ghosting

Animation with ghosting

Another tool we were introduced to today was baking our animations, this is used when we want to move our animations to another software once we have a final result. This is used because our projects built in Maya are software dependent meaning they can only really be used and manipulated in Maya, unless using a converter which isn’t recommended. Baking gives a final output of our animation that is compatible with other programs such as video game engines.

Animation timeline before baking

Animation timeline after backing

Rendering in Viewport 2.0 and Arnold

This week we were introduced to rendering and the differences between rendering in the Viewport and with Arnold. Due to my previous experience with Maya in my first year, I am already comfortable with rendering in Arnold however it was interesting to learn more about the differences and pros and cons of each. Arnold is more powerful and gives better results however it is an offline render meaning the animation won’t render in real time, depending on the complexity of the scene it can take a very long time to render however rendering in the viewport gives you result that are closer to real time.

For my renders I use Arnold and render out PNGs or TIFFs of each frame. For a previous project I used TIFF because I needed transparency in the backgrounds but typically I use PNG files. For backgrounds I use HDRIs downloaded from Polyhaven.com that are free to use, I then turn off the visibility in camera and make a background using flat planes.

This is a small video of the first IK rig I made from scratch, the animation isn’t good or natural but it proves that the rig works correctly/as expected.

Creating my own IK rig

I followed the system that we learnt in class just using the base mesh that I made the week before.

Making the finger joints was the most precise rigging I had done so far and I had to research to how to resize them so that they would be accurate in their positioning

I then learnt through a tutorial that I could make controllers that are called pole vectors, these tell the IK handle where to direct the movement that is solved by the IK handle. In this instance it is bending the knees of the character forward

 

This is to show that I also put IK handles on the finger joints as well so that they would move in conjunction with the wrists and shoulders.

Result shown in a small animation

This is an animation of the rig. It is imperfect and I know that things could be improved but I think it proves that the rig works and everything is set up correctly. The biggest problem I see is the rotation of the leg IK handles, however I think this is a problem with the base mesh, such as accidentally forgetting to freeze my transformations at an earlier stage. Because this was a week of focusing on rigging I chose not to go back and fix the mesh due to time constraints but through this I did learn the importance of a making sure the base mesh is correctly prepped.

Week 5 – Using a human rig

This week we started using pre-made human rigs in Maya. We were taught how to use the rig what the different controls were and how to use pole vectors to correctly orientate the knees and elbows

This was my first animation using the rig and I’m happy with how this turned out. I found figuring out the mouth controls quite tricky and looking back I could have made this animation more expressive particularly with the mouth but I think it was a good intoduction

Posing the character

After this we were tasked with posing the character, we were asked to find reference pictures to use that we could try to replicate, I decided to use dancers as I knew I would be able to see the pose of the character clearly and have well defined shapes

I really enjoyed this exercise, I liked solving the puzzle of how to replicate something and seeing it come together. I definitely this helped me get more acquainted with using rigs and it made me think about the possibilities they supply animators with. These exercises are something I would like to do more of.

Week 6 – The Walk Cycle

This week we learnt how to do a walk cycle. This is a fundamental exercise that I was looking forward to learning as it felt like a good milestone in my journey of learning animation.

We started by looking at a lot of references and learning what makes an effective walk cycle. I was inspired looking at how simply an effectively walk cycle could be created by just using a small number of repeated frames.

Examples from Richard William’s and his books “The Animator’s Survival Kit”

Examples from online resource rustyanimator.com

I really liked how these simplified the walk cycle and made it into a much more manageable task even though the results are very effective. These examples provided a road map that we could copy for the basic structure of our walk.

I copied the timing and frame selection precisely and tried to copy the poses accurately too.

I started by posing the contact pose on frames 1, 13 and 25 and then adding the in-between passing poses on frames 7 and 9, with the last step being to add in the up and down movements in between those. I really liked using this frame work as it gave me a really nice base to work with and made the task feel a lot more achievable. It was also a great reminder of how valuable looking up references and referring to other peoples work can be.

I was comforted to find that a lot of the work could be done precisely by copying keyframes and mirroring the action depending on which leg is leading. This made the process simple and most importantly precise meaning that everything matched and the movement looked cohesive. I really enjoyed seeing this come to life while I was making it.

One I got the legs stepping it was time for me to focus on the hip and shoulder rotations. From discussing it in class we recognised that the hip rotate in and in line with the leg (ie, the left hip rotates inwards when the left leg is stepping forward) yet with the shoulders the opposite shoulder is affected. This is because we move our arm in line with the opposing leg (so if the left leg is stepping forward the right arm swings).

I found the hip rotation difficult to grasp at first and after gaining feedback I learnt that this was because I was moving the incorrect controller. Once I did find the right controller I ended up making a very exaggerated walk which can be seen below

I felt that this needed some fine tuning to look more natural so I refined in futher and this is my final walk cycle. I’m happy with this and it enjoyed the feeling of having completed it. In the future I would like to try and make some more walk cycles potentially with different rigs as I think there is a ton of personality that could be applied here to make some really fun results and I think I would enjoy that

A final thing to mention is that this week we were introduced to animation layers, this is essentially a tool that means you can have different layers on animation in one project. This means you can set up animations separately that won’t be affected by each other. In my project I created one for the eye blinking movement as this is an action that can be easily isolated. Another example of a good use of animation layers would be if you wanted a character to hoover or bob in the air consistently throughout your animation. You could put that one movement on its own layer meaning that you wouldn’t have to work over the top of it when doing the bulk of your animation. It makes your timelines and graphs easier to read and use by having repetitive actions or actions that you want to section off being put in their own layers. I think this is a great thing to be aware of and can think of a lot of instances where this tool would be very useful and important for streamlining workflow.

Animation layers are also great because they’re non destructive meaning you can test things out on a layer without disrupting your main animation.

Week 7 – Refining the walk cycle and learning about parenting constraints

This week we started with a recap of last week where we delved in to walk cycles. We were given additional information and examples to help refine out walk cycle which I really appreciated and I’m looking forward to spending more time making my walk cycle better for future assignments.

We looked at things like stride length (typically beginner animators tend to give characters shorter strides than average) as well as leaning the feet out when the contact pose happens

The first practical exercise we completed was extending our walk so that the character takes some steps forward. This is something we hadn’t tackled yet as initially I only animated a couple of steps forward.

If we were to simply duplicate the animation then the character would reset its translate position between steps (as seen in a video below) which wouldn’t create a natural walk. To negate this issue we used a technique called ‘relocating’. First we launched the time editor, selected the whole animation (all NURBS curves) and added a track, we then duplicated that track and placed it after the first one. To adjust the starting position of the second step we navigated to the last frame of the first cycle, selected the IK handle for the left foot, right clicked on the second cycle and selected the option to ‘match relocators’. This signalled to the second animation where it should take its location data from, resulting in a smooth transition between steps. I then copied this technique to make the third step

This is the result when I followed this relocation method for 3 steps

This is the walk duplicated 3 times without any relocation. It is noticeable how the character returns to the start position at the start of each cycle

Starting to learn more advanced parenting constraints to prepare for assignment 2

The second half of the class was dedicated to learning parenting constraints as for our second assignment we are tasked with making an animation of a character lifting something up. The easiest and most intuitive way to do this is to use constraints to parent the box to the character so that when the box moves, the character does too.

The first way that we set this up was by simply creating a box for our rigged character to lift. We then used locators (empty objects that capture data of a point in 3D space) to assign the position of the hands and the box before parenting everything to the box itself. For this example we have 4 locators, one for each of the hands, one for the box and a fourth to group those together as a final parent. Moving this fourth locator allows the box to move and the characters hands accordingly

 

These screenshots show where I placed the character and the locators on the box. It is important that the character is in the position where it is going to interact with the object when we make our locators.

This is a small video of the working box constraint

 

Our last exercise for this class was to use parenting constraints which creates a channel in the channel editor that can be toggled on and off. This is a more advanced technique yet in my opinion easier to use as it is more integrated in to the software.

To carry out this technique you place a parent constraint on the object that needs to be parented (in the box case this would be the box itself). This generates a blend parent channel in the channel box that can be turned on or off (toggling between 1 and 0, 0 meaning off and 1 meaning on). This means that if the character is going to be putting something down or intermittently interacting, such as throwing and catching a ball I can specify whether or not the ball is constrained during sections of the animation.

This is a video of my first experience using parent constraints from the rigging menu. The anvil weight, was constrained to the wrists.

Following a tutorial to try and understand constraints more

I struggled with grasping constraints when I tried on my own project files so I followed a tutorial step to step that constrained a ball in a see-saw type animation. It was very rewarding to make this to myself and has improved my confidence in figuring out this same concept with a character rig

Original video

Sir Wade Neistadt (2018) How to Make Characters Grab Objects – Parenting Constraints. 14th June 2018. Available at: https://www.youtube.com/watch?v=U8JXO39aAqY (Accessed: 14th November 2023).

Having watched that video I tried again with the Eleven rig trying to add in a ball throw/catch. I’m really happy that I managed to figure this out as I wasn’t sure how long I would be stuck for. Something did go wrong with the rendering here though so I had to lower the opacity to bring the white level down. I do think the animation and the use of constraints is still clear and can be seen here.

Starting my own weight lift project

The first thing I did was to make a new walk cycle. I struggled with this in a way that I didn’t predict as I thought I would be able to use the walk cycle I had made previously in class.

However I fell upon an issue as I made my walk cycle as an “on the spot” walk cycle meaning that I couldn’t figure out a way to easily translate it forward so that I could relocate it to make more steps.

I tried to find a solution but ended up making things far more complicated for myself so I decided to make an entirely new walk cycle instead.

This was a learning curve though as I find it easier to make them on the spot considering keyframes can just be copied, but with walk cycles that move forward this is a bit more tricky. I managed to figure it out through trial and error and I’m pleased with the result.

I checked that I could duplicate and relocate the steps well and I can so that was good progress

The next element I wanted to tackle was making my character stop. This is because in the weight lift animation I will need my character to stop asses the weight of the box and how they are going to lift it, before proceeding to lift the box. I was a little apprehensive of this as I wasn’t sure how to duplicate an animation and the edit it. Through research I managed to learn that I could mute the time editor, meaning that I could manipulate the keyframes. Therefore on my duplicate animation which I named “stop” I could change the animation keyframes so that the character stops in place.

This was a milestone for me in this project and I was very pleased to have figured this out.

Revisiting on day 2

I made a lot of progress at the end of day 1 that I cannot show documentation for as I deleted the files once I realised I couldn’t continue working with them. I made the character pick up the box but unfortunately I made an error when I tried to use constraints. The constraints were working but I think the impact of the constraints meant that something earlier in the animation got controlled in a weird way. Essentially the arm was rotating unnaturally in an earlier part of the scene. As previously stated unfortunately I don’t have any documentation of this as I just ended up deleting the files and starting again from scratch. I am disappointed in myself for not documenting it here on my sketchbook but I will learn for next time!

This meant that as described I restarted the animation. I wanted to try and resurrect some of the files I had made before but I felt it seemed something was still wrong with them, like I had added constraints to a larger area of the rig than just the hands or something.

Therefore I started from scratch and this is my progress so far. Not interacting with a box yet but at least the character is walking, stopping and pondering how to proceed to lift the box.

One thing I did struggle with today even though I only did my first bit of experimentation with it yesterday was remembering how to use the time editor. I managed to relocate the footage fine but I couldn’t figure out how to add or manipulate keyframe from it for a long time, this made me nervous as I already felt I was behind having to start over however I remember through the process that I needed to almost save over the keyframes from the time editor onto my animation workspace and then create the new section of my animation that follows on from those key, then select the new animation and append it to my time editor.

One worry I do have with the workflow of using the time editor is how easy it will be to amend my animation if I want to change anything. Right now it feels as though once it’s in the time editor, it’s locked in as the final result. Which is okay but not the nicest workflow.

Week 8 – Reference and Rotoscoping

This week we discussed using and importing reference into our scenes and also how rotoscoping was used and has evolved from the beginnings of it to modern day

The first advancements in rotoscoping correlate to the time of Eadweard Muybridge. Muybridge was a photographer who made advancements in the medium by taking successive stills of movement which allowed analysis of how humans and animals move. This was effectively the first version of rotoscoping.

The picture series of his above is particularly significant because before this sequence was created it wasn’t believed that horses ever had both feet off the ground. Even in motion humans were convinced that horses always kept at least one foot on the ground. This image disproves this as seen in frame 2.

It can be seen here how pivotal this technique was for getting data around movement and this technique began to be used more widely in animation since that time. Walt Disney, especially was an early adopter and used this method a lot

This is a comparison of the reference footage for Peter Pan and the animated counterpart

Making Alice for Alice in Wonderland also used this technique. I like this picture as it shows the early ways Disney were working too where they would draw over the frames of movement to replicate it, of course adding in their stylisation on top

Using reference in animation and art more broadly can be a contentious topic. However generally reference is highly encouraged as it brings a sense of realism and believability to the finished product that is very hard to achieve without it. How artists approach using reference is a topic in itself. Copying directly from a reference is usually not advised as it may evoke copyright and as well as this the goal should always be to make something new. An approach that is widely praised in the “Steal like an artist” by Austin Kleon is to take a variety of references from different places and collate them to make something new, all while also inherently adding your own style. This means that in the case of animation the artist isn’t alone with creating the movement and it will be grounded in reality however the result will be different enough to stand out and be more unique.

Understand how to add reference videos to a Maya scene

We then looked at the technique in terms of how to add a reference video to Maya.

The first step is to find a video that you would like to use and convert it into an image sequence. This can be done in many different programs such as After Effects, Premiere Pro or Quicktime. The image sequence should be exported as individual image files number sequencially, these can be PNGs, JPGs or TGAs. For this example we have TGA files

This is a screenshot showing the TGA files that have been exported.

Once you have the image sequence there are multiple ways to import in into Maya. The three we looked at are PolyPlane, Camera Image Plane and using QuickTime.

I have used image planes before for modelling reference so I’m familiar with that setup however the other two were new to me.

The one I chose to experiment with was PolyPlane as this sounded like a better workflow than Image Plane, this is because image planes are tied to the camera meaning it’s not as easy to move them. Using a polyplane means that is just an object in the scene, another element that can be placed and controlled wherever needed because it is another object in the scene the way to get the image sequence on to it is use texturing.

To do this you assign a new material to the polyplane, make this a new lambert. Then in the colour attribute of the lambert you assign a file which is the first image in the image sequence. Making sure the ‘use image sequence’ box is ticked. This will mean that the video get transposed on to the plane, frame by frame.

This was me trying to copy the reference, I found this quite difficult primarily because I struggled to get the viewport and the character lined up with the reference, particularly as the character is very stylised the proportions were different but I did like learning the technique and I can definitely appreciate it’s importance when it comes to having grounded animation results.

Week 9 – Starting on assignment 3

This week we started learning elements that will help us with completing our third and last assignment for the module which is to animate to an 11 second sound clip. For this we need to import audio into our scene and create animations that sync up to it

First we learnt how to import audio which is done simply but going to File > Import or right clicking on the speaker icon to the right of the time slider and pulling up this menu which then leads to an import audio button. Audio is imported as a .wav file.

Once the audio is imported a waveform is added to the timeline like this. Giving a visual representation of how loud the audio is at certain points and any quiet moments, this audio can then be scrubbed with the time slider allowing you to fine tune where the different movements take place.

Camera setup

As we were just learning techniques to help us in class today we were also made aware of camera techniques in Maya that can enhance a shot, particularly focal length and depth of field. I think these are really interesting things to think about when it comes to designing our shots as a lot of meaning can be placed on how the camera is used. To get used to the tools here are a couple of experiments I did

Camera Placement

One thing I wanted to experiment with in the lesson is what it would be like to set up camera as if two people were talking to each other, I wondered how I would compose that as a shot

This was my first time having two characters in one scene in Maya, I angled them to face each other and placed a camera to focus on each

Here you can see the different viewports, the top right shows the perspective view and the bottom two show each of the individual characters.

I then used the camera sequencer in Maya to be able to cut between the two cameras. There is no animation to this yet and I don’t think I will use this setup for my submsission for this assignment, I was just curious about what this would entail and how easy it would be to accomplish.

Tangent styles – changing to stepped tangents for blocking

One final thing we learnt in class this week was how to change to step tangents when animating. By default in Maya animation is done on spline curves, this means that the animation is smooth and automatically curved in the graph editor, making a smooth transition between the two points. This is beneficial for the final result as unless you’re working on a project that specifically needs direct cuts between moves such as stop motion, this is typically the desired and most appealing look however sometimes when animating a lot of motion it is easier to animate when you cut out that transition and just look at the affected frames. This is where changing tangent styles comes in. By changing your default tangents to stepped tangents you cut out the transition and are able to see your animation without any in-betweens. This was a really interesting thing to learn as it does strip back the amount of information you are being shown and will prove useful on high action scenes

Here is that in motion. Both characters have the same animation just the left hand one is using stepped tangents instead of the default. It can clearly be seen here that the right hand example is a lot more fluid.

Continuing with the weight lift assignment

Since learning to relocate the walk. I have carried on using the time editor to add progress my animation however I have encountered a lot of problems with this workflow. Most notably that once I add something to the time editor I cant seem to change the keyframes. I thought I had figured out a way to do this (by disabling/muting the time editor) however this hasn’t provided a true fix for the problem. As a work around I would go to the last frame of my work in the time editor and commit that pose to the scene, therefore being able to progress further in the normal workspace.

However, with my most recent additions to the piece, even after working on new elements I could add those to the time editor. I have tried to research this and read the documentation however I have yet to find a clear fix.

My way around this has been to render out what I have in the time editor and then add the new elements to it in a separate render.

This works well however I did notice that the keyframes weren’t aligned perfectly and at one point I was experiencing some trouble where it seemed like one arm was being constrained when it shouldn’t have been. I couldn’t figure this out through key framing as my blend shape was keyed to activate at the right time. So I again resorted to rendering out an older version and stitching the pieces together.

To make sure my different renders matched I created a camera and lock its position in each of the respective files so that all renders would look the same.

In terms of matching the keyframes I unfortunately had to result to doing this by eye as every time I wrote down the values and copied them over I still found there were things that didn’t match. I am still unsure as to why

Unfortunately this has resulted in an animation that isn’t created in a very seamless or transferrable way and I am disappointed in this. However I am proud of my actual key framing and animation work. It is worth nothing though that if I had time I would remake this and not use the time editor so that my animation would all be make in one scene and be completely editable throughout the whole of the timeline. This workflow would have been a lot easier and more manageable. It is also why this character doesn’t blink. As I could figure out how to add animation layers to clips that were already committed to the timeline.

I am excited for the second task where I can have a more adaptable workflow but for this task and the challenges I’ve faced I think I have found a good enough result. And I’ve had lots of time to practice key framing, parenting constraints, setting up camera and rendering.

Something I would like to document is a mistake I made when exporting from Premiere. My frames didn’t match and I couldn’t figure out why. I realised it was because I was exporting at 25fps which is the default instead of 24fps so that is something to look out for

Moving forward with this piece I am not using the time editor so it should be easier now. I would just like to make a small lean/stumble back when the character returns to upright position and then a walk out of the scene.

Weight lift animation – First try

This is my result for my first try at this assessment. I would have liked to have developed this more however I felt very hindered by the workflow. It’s not the result I was hoping for but I think the techniques are still displayed

Starting assignment 3 – the lip sync

I was pleased to be starting with a blank slate for my next assignment, the lip sync. The first step was to choose my audio and for this I decided to go through the 11 second club website as I knew they would have a good catalogue to chose from

This was the clip ended up choosing as I could very quickly see the type of character I scene I wanted to make for it. The dialogue felt very expressive to me and I thought it would make for a fun project

This is how I’ve framed my scene. My current plan is to have the character speaking at some kind of event such as an awards ceremony. That’s why I have created a podium and hope to create a background and some lighting to evoke this setting too.

Having a small part of the character available to me to animate means that I can be really expressive with the face which is my main goal with this project.

Week 10 – Learning the camera sequencer and jump animation

This week we learnt how to use the camera sequencer to edit out animation together using different camera. This was a feature I familiarised myself with last week without realising we would be taught it in more detail this week. It was great to get more experience using this and I have a lot more knowledge now. One thing I did struggle with was generating an ubercam. An ubercam is one camera that collates all the data from all the cameras used in your sequence and makes it into one camera. This is really useful if the animation is being exported to a game engine as game engine often can’t read Maya cameras natively. Therefore being able to export all your data with one camera is very useful.

The issue I came across was that I had animated my cameras that I had made shots from, factoring in this motion meant the scale and frame numbers were not correct and for an ubercam to be able to be generated the scale and leangth of the shots has to match the source input. To fix I deleted and remade all of my shots, looking at the attribute editor of each and making sure to set my start and end frames for both the shot start frame and the frame I want it to be in the sequence. Figuring this out has made me a lot less intimidated by the camera sequencer and I can see the potential of using it for a lot of projects.

For the jump animation. I quickly recreated it using the slides. I didn’t animate the hands here which would make the animation a lot better however I was just focusing on the weight shift and getting the majority of the keyfarmes correct. This was a quick fun exercise that helped me refresh my memory of key framing and weight mechanics.

These are some of the key poses that are in my jump animation. Is interesting to look at it dissected. Using this silhouette mode is useful for focusing your mind on shapes and weight distribution in the body

Restarting Assignment 2 – the weight lift animation

After getting feedback and clarification from my lecturer on assignment 2 I have decided to restart my weight lift assignment from scratch and avoid the time editor. This will make for a much cleaner result and hopefully a better outcome

So far I have created a rough version of the beginning walk cycle which can be seen below. I put the hand animation on a separate layer and aim to add a blink in the same way too.

Final submission for assignment 2

I have now completed the weightlifting animation for which I chose to do a box pick up. I am happy with this result and very grateful that I was able to redo to a relatively high standard without using the time editor. I tried my best to add in some stumbles and leans to the animation although I found this quite tricky when it came to returning to upright after putting the box down. This was because I absentmindedly created my keyframes very close together which meant when I played my animation through the movements were very quicker. Quicker than the earlier portion, therefore I had to put in some time to resolve this. Interestingly parenting constraints I didn’t find too difficult to deal with it was more the movement around them. In the final outcome the box is probably a little too bouncy but I struggled to find an alternative that would still make it look like the box was moving with the character, I experimented with the graph editor but felt this was the best/most realistic result I could get.

Completing Assignment 03 – Lip Sync

I then moved on to working on my lip-sync. I have already shared the set up of my scene, the audio and what my initial idea is. In order to start animating I first made a variety of mouth shapes for my character on a separate layer. Now initially I really struggled with the layer system. I was fine in class when we learnt the blink but created a layer for the mouth proved tricky. It was extra practice to get used to the workflow and the system of creating the layer, adding elements to that layer and then editing them. Additionally there were a learning curve for layer modes and how they influenced each other and the final result.

My understanding from the previous weeks teaching was that a good workflow would be to create the mouth shapes on one layer, then create a new layer and copy those mouth shapes onto it (with the original muted). I however struggled to copy the keyframes. I’m not sure whether I made a mistake somewhere or maybe not all of the same objects were in both layers, I’m not sure but I couldn’t figure it out. Therefore as a workaround I decided to make all of my mouth shapes on one layer and then just move them in the timeline to after the audio ends, meaning that those keyframes will never be publicised but I can still use them as reference. This meant that all of my mouth movement keyframes were in one layer.

I then went through the clip and wrote down all the frames and there corresponding sounds/lip positions as well as noting the keyframes my created mouth shapes were on. IT was then just a case of copying and pasting the reference keyframes on to the correct frames of the animation. This was an enjoyable exercise for me as it was the first time I could see the piece coming together.

The biggest learning curve for me using this technique was that I couldn’t accurately portray every sound as the animation ended up being really choppy due to the audio being quite fast. Changing mouth positions that quickly looked jarring and unnatural so there were a few places were I had to trim down the amount of animation and choose what the most dominant sound or shape was that would help me convey the dialogue convincingly.

This is render of my work in progress this was when I just had the lip syncing completed, no other supporting animation or imagery was done yet

After this was done it was time for me to look at other aspects of the scene such as camera movement, head tilts, blinks, and hand movement. Each of these I put on a separate layer. These were fun elements to figure out as it felt like it was vastly improving the appeal of the scene, it was fun to play around with the camera moves and the hands following the dialogue. Again it was a learning curve working with layers, this made me discover how much the order of the layer impacts the output and whether they are overwrite layers or additive layers. Something else I had to be careful of was making sure that I locked my camera when I was done with creating the camera movement for the piece, there were quite a few mistakes made when I accidentally had it active. After I locked my camera I did have it on the screen while I was animating the rest of the elements, I split Maya into a two pane window so that I could select and animate the element I wanted while also being able to see the final output which worked really well.

Prior to adding all the new elements I did experiment with importing a render of the animation and the audio into Premiere to mix them together. Thankfully they aligned well without me having to do any alterations which was a relief.

The last stage for me was to add the background and the floor. As I knew I wanted it to be an award ceremony I wanted to create some kind of static image that would be on a screen behind the character indicating that he was giving an acceptance speech after winning an award. So I created this very basic screen with a trophy and some information about him/the award. I like how this turned out as I think it helps sell the scene and makes it a lot more interesting. Additionally I think it is reasonably realistic in terms of how this type of environment would be set up in real life. More cold be done with the scene setting here however I know that that it’s really what we are big graded for and I didn’t want to take too much away from the character animation.

Overall I’m pleased with this project and I definitely learnt a lot. It was great to see the difference when I added in all the supporting elements, like the hands and eye animation as well as the backdrop.

Final submission for Assignment 3 – LipSync

Week 11

This week we had a recap on the module to look at everything we have learnt over the weeks and make sure we had all the information we needed to update our portfolio sites and get our assignments done. We recaped the lip syncing and I learnt more about the theory behind visemes and phonemes. In addition to this I also learnt that the jaw on the Eleven rig can be rotated rather than simply moved/translated down which is something I wasnt aware of when I made my submission for assignment 3. Although using rotation would’ve made a better more natural result it’s not a change I’m going to implement now that I have finished the assignment