Good Evening Jay dot Tumblr dot Com | Dev Blog

Recap

On Tuesday I continued with my procedural research. I found the papers I was reading quite difficult to parse which made the experience rather frustrating.

To Do Today

Today I will be starting my implementation. The goal today is to complete all of the preparatory programming and engine work required before implementing any procedural algorithms. For this first Sprint I will be focusing primarily on the locomotion of the legs, and I will be be adding upper-body procedural animations during the second Sprint, which will now be iterating on the first. I will be attempting an Inverse Kinematics implementation.

Recap

Last week’s radio silence was due to other commitments. I ended up overdoing it a little bit and required a couple of days to recover. I did include buffer time in my project planning for unexpected interruptions due to my health so hopefully last week’s disruption won’t result in a delay to the beginning of the next sprint, currently scheduled for March 20th.

When I last touched this project I had just completed the character swapping function and had begun my research into some procedural animation algorithms.

To Do Today

Recap

On Tuesday I was able to complete the traditional animation implementation and conduct some basic functionality tests.

My free time to work on this project has been limited this week due to running in the student elections.

To Do Today

Today I will be implementing a very quick and simple player character swapping feature and beginning my practical research into Inverse Kinematics, Gait Synthesis, and Physics-based animations.

Recap

Last Thursday I made a start on the implementation of traditional animation techniques. Today I will be completing that implementation using the assets I currently have to hand.

I found when I started my implementation last week that I may have vastly underestimated the difficulty of implementing traditional animation techniques. This doesn’t yet contradict my hypothesis for the out come of this project however.

To Do Today

Today I will be finishing the traditional implementation techniques for a walk cycle. After conducting some testing to make sure it’s met my bare minimum functionality requirements I will move on to my procedural animation implementation research.

Today I also need to create a couple of new Tasks for the next Sprint. I want to iterate on the movement system to give it some more weight. I also want to try and source some new traditional animations that I can swap out before I begin working on the Running animation. The animations I am using right now don’t cover the full range of movement I need for full implementation.

Recap

Yesterday morning I concluded my preliminary research into the fundamentals of traditional 3D animation. I have other commitments on Wednesday afternoons so this was the only task I achieved in the time available to me.

To Do Today

Today I will be beginning to implement traditional animation techniques to my base character to give it a nice walk cycle. By the end of today I expect have a functional implementation.

Recap

Yesterday I was able to conclude my research into the basics and fundamentals of 3D character model creation, and completed a fair amount of research into how traditional animations are creating in software such as Blender.

To Do Today

This morning I will be concluding my research into traditional animation fundamentals. There’s some jargon I still don’t understand, and I would like to familiarise myself with the pipeline.

It occurs to me that yesterday’s poetic title didn’t accurately convey the goal of this Sprint so I’ll outline that now.

By the end of this sprint I aim to have delivered an implementation of a character that walks using traditional animation techniques available in UE4 - this is the control character. There should also be a second implementation of a character that walks using procedural animation techniques.

These are the primary deliverable of the first Sprint, but there are some secondary deliverables as well. I also aim to include a simple sandbox to test the capabilities of these implementations, a method of switching between the two characters in-game, as well as a static “enemy” actor that the player can lock onto. The first Sprint is scheduled to be the longest (after pre-production) as it will include the majority of the learning and research required to support later Sprints. However, the secondary objectives will be pushed to the backlog if they cannot be achieved by the scheduled end of this first Sprint to prevent delaying other primary objectives in subsequent Sprints.

Day 1 Recap

Yesterday I managed to achieve all but one of the priority goals I had set for myself.

A simple sandbox space is suitable (for now I’m just using a default map provide with the UE4 start content) and the player class and movement system has been implemented from scratch. It’s very basic at this stage, but I plan on iterating on it as I move on to animation implementation.

I have moved customisation of this blog to the backlog and will probably muck around with making it look pretty closer once I find out if I have been accepted to participate in ExpoTees this year.

I  was unable to populate the “Walking Animation | Traditional” Story with Tasks as Child Nodes yesterday, but I completed this task this morning before I started today’s daily scrum session (I.E. this blog post).

To Do Today

Before anything else there is an error on Jira with some duplicated stories that I need to resolve.

Today I’ll begin progress on the “Walking Animation | Traditional” Story.

The first Task in this story that I’ll be addressing is doing some research on the fundamentals of 3D character models and 3D animation for games. I intend to spend approximately an hour on each of these Tasks to build on my existing knowledge. I feel this is important not to overlook as this project marks my first time working with 3D models and implementing animations.

Assuming no blockers occur during my research I will then move on to implementing a basic walk animation using the same Unreal Documentation guide I followed for implementing the movement prototype.

Today marks the beginning of the first production sprint.

First I’ll briefly recap what I did last week.

Recap

Last week I got the sign off from my supervisor to conclude the pre-production sprint and commence with the project in earnest. He delivered feedback on my project proposal and I answered any questions he had about it.

One concern he had was whether I had enough time to complete the goals that I had set out for this project. I reassured him that whilst implementing the project plan on Jira I had reconsidered the goals I want to achieve and trimmed the scope slightly. For example, a quadrupedal implementation is now a stretch goal instead of part of the minimum viable product.

There were a couple of outstanding tasks at the end of Thursday that I had intended to address the next day, but I have had to roll over some of these tasks to the beginning of this sprint so not to delay it further. This is because I was unwell on Friday and over the weekend. I will cover these tasks in more detail when I address today’s plan.

Finally, my supervisor authorised my use of ChatGPT as a research assistance tool. As an example, I was not sure if the correct term was to “roll over” a task or to “grandfather” a task when I was writing the previous paragraph. A quick a Google of my question did not provide me with a clear answer, but ChatGPT’s response pointed me in the right direction. My use of this software will be carefully documented and monitored by my supervisor, and will be referenced in my final report.

To Do Today

As mentioned, I have a handful of tasks rolled over from pre-production before I can conclude it.

These tasks are as follows:

Set up sandbox that will be used for testing.

Implement movement system prototype.

Customise Dev Blog

Populate stories for first sprint with child nodes

Item 3, customisation of the Dev Blog, is non-essential for project progress, but I would like to complete this prior to ExpoTees marketing material being circulated.

I will be tackling the remaining items this morning before.

After lunch I intend to proceed with the beginning of the first story of Sprint 1, implementing a walking animation utilising traditional animation methods.

This blog will be serving as the development blog for my future projects.

The first project I will be documenting is my Games Programming Project (FYP) at University, which is also my submission for ExpoTees 2023.

The name of my project is:

“Traditional Animation VS Procedural Animation: Is it worth it?”.

Project Description

I will be exhibiting a comparison between traditional animation and procedural animation techniques. I’ll be focusing on the player-controlled locomotion of a 3rd person perspective 3D character model to demonstrate the differences between these two techniques.

I am aiming to create a character that controls and feels like a character from a game like Dark Souls or Elden Ring. With that in mind the locomotion modes I will be focusing on are walking, running, crouching, and rolling.

I’ll be implementing two player-controlled characters in my deliverable. The first will be implemented with traditional animation techniques, and will be serving as a control model. The second character will be a programmatically driven procedurally animated character. I will be attempting to match the feel and performance of the control character.

The final deliverable will be created in Unreal Engine 4.27.2. I am using Git for source control.