Pre-production process

My pre-production process worked very well. I started out with one artist whose work I really enjoyed, sparking interest to research further into this area and continuously generating new ideas to be explored and evaluated. This, combined with the research paper that we did for our first assignment, where I explored different versions of platform games, gave me a good idea of what I wanted to do for my second assignment, giving me enough time to really think my ideas through and develop them to their best potential. I take this as a good approach for research and development for next year; to start early and thereby creating the extra time to really research and develop an idea instead of just following the first initial thought. Improvement wise, I wish I had documented the process a bit more, with little sketches, mindmaps, etc and not just gone through it in my head a lot. This is something I would like to implement into future projects in order to better structure my work for myself and also make my thought process clearer for the tutors grading my work.

 

Asset creation

Following the Pipeline of modelling in ZBrush – Retopology and modifications in Maya – texturing in Substance, proved very effective throughout my work, letting me produce assets a lot faster than originally expected. There were minor issues though that I need to pay closer attention to in the future and that was the inconsistency in detail. While the tree and the tree stump appeared quite high poly due to the fact that they were created in Zbrush without retopology in Maya, they have a lot more detail than, for example, the mushrooms do, causing an inconsistency throughout the scene. This is something that I will have to pay more attention to in the future as the game could have easily handled mushrooms with higher polycounts as it is a rather small game and there is no need to keep the polycount this low so it won’t interfere with performance, and if the polycount is kept low, it should be consistent throughout all assets, including the trees.

 

Time Management

In general, the time management plan was good and in itself, the plan was not the issue, as it left room for constant evaluation of my own work and even room for improvements. Had I been able to stick to my plan, and I would have been able to, I would have been able to finish the whole game as initially planned without a problem.

However, what the plan didn’t leave room for was getting as ill as I got, being unable to attend classes and needing treatment back home, basically overthrowing the whole plan, leaving barely enough time to finish a less complicated version of the initial concept to a satisfactory level.

This was something that I couldn’t have factored in and that was absolutely out of my control and it was an issue that I faced throughout all of my classes this year. Still, I think I can take a few lessons from this into my final year and that is that all the planning in the world won’t help you if you don’t look out for yourself and your health. I know now what to look out for and which steps to take to not let my health get out of control and I know that in the future I need to plan in even more time than others may need or that time management schemes suggest, just to accommodate the possibility that I may not be able to do work for a longer period of time as well as time for breaks. As I still have an unfinished assignment that I would like to complete this summer, I can take my realisations and directly practice them on the assignment before coming back to university in September, to make sure that this approach really works for me.

 

Switching from Unreal to Unity

This is the one thing that went absolutely wrong during my project and I still don’t know why. Even though I did extensive research on this subject and watched countless tutorials on youtube and pluralsight, I still haven’t found an answer, making the switch inevitable. The only way that this could have been avoided would have been going to classes and regularly asking for feedback and help within sessions which wasn’t an option in this particular case. I am sure that had I been able to ask for advice within lessons, this problem wouldn’t have happened. This only shows me to really take the opportunity to collect as much feedback as possible if circumstances permit it.

The switch from Unreal to Unity itself went surprisingly smooth, following multiple tutorials on Pluralsight to set up my scene and always having someone on hand to ask questions and provide feedback or assistance, really helped to finish the game quickly and without major issues, showing that asking people for help teaches you more and more efficiently than trying to figure things out yourself with online tutorials.

 

 

 

Originally I wanted to create cut scenes that look similar to illustrations in children’s books. However I really wasn’t happy with the outcome of my first attempt:

It wasn’t terrible but also not great and I felt like it would take away from the quality of the game and therefore decided to go with a text cutscene using the Parchment font in Photoshop.

Now that I am using a different engine I can’t use the generic character from the Unreal engine anymore which was the one I was planning to use. The choices for characters in the unity asset store were very limited so it became clear that I would have to create an alternative for my character. luckily one of the asset packs came with a generic box character which any other asset could be attached to. I decided to use one of the particles from the other asset pack that I downloaded as my character and my flatmate helped me attach it to the box so that the particle would be able to act as my character.

The inspiration behind this was that the character would be a willow wisp rather than a pixie, giving the game a more finished look.

Considering that the game would be only one level now, the original storyline doesn’t make a lot of sense anymore. therefore I decided to adapt the storyline to better fit the current level. The wisp would be looking for his little friend who got lost on top of the tree and once found would follow the playable wisp back home. I created this code using this tutorial:

as well as the expertise of my flatmate to adjust the code to my specific needs.

Knowing that this works I started importing more assets like the mushrooms, windows, doors and platforms into the game engine and placing them in the correct spots. In order give the character the ability to jump up onto the platforms I placed invisible collision boxes over the mushrooms on the same line as the platform, creating the illusion that the character is jumping onto the actual mushrooms, even though they are placed way behind the axis that the character is moving on.

I decided to go with a 2D Template to set up my game, importing all my assets into the engine as well as adding the assets from the asset store.

https://assetstore.unity.com/packages/templates/2d-platformer-controller-69772

https://assetstore.unity.com/packages/essentials/asset-packs/unity-particle-pack-5-x-73777

I used the following tutorials to set up a rough playable preview of my game, consisting of the background, the tree and the character moving back and forth on a platform, testing collision with the tree.

https://app.pluralsight.com/library/courses/unity-2018-fundamentals/

https://app.pluralsight.com/library/courses/unity-game-prototyping/

Luckily this proved to be no issue and the tree remained collision-free, allowing the character to pass under the tree roots without any problems, solving the problem I had with Unreal.

I had to admit that I would not be able to set up my game in Unreal engine because of unsolvable problems of unknown origin so I had two options left:

1: give up – Not a good option…

2: Try and set up the game in Unity

So obviously I went with option 2. This had some major disadvantages over Unreal, mainly, that I had never worked with it before and we hadn’t been taught Unity in class. However, at the same time, it really didn’t make much difference considering I had missed most lessons due to illness anyways so I would have to rely heavily on tutorials to set up my game in unreal, therefore it made no difference whether I was watching unreal or unity tutorials. And some clear advantages: There were more game specific tutorials about Unity on Pluralsight than about Unreal, to be precise, almost double the amount, and my best friend and flatmate is studying a Games design course, meaning that if I get stuck at any point, he might be able to help me resolve my problem. So what choice did I really have?

 

I started out with importing the tree into unreal as it was the most important item, the item that the whole game would evolve around and encountered my first mayor issue.

As far as I was aware, every model that is imported into unreal comes without a collision map, that’s how all the tutorials showed it and I watched a lot of them, unticked and ticked every box that I could, deleted non-existent collision maps within edit mode, trying to fix it somehow, but no matter what I did, the problem stayed: As soon as I placed the tree, the character would stop moving.

I tried to place the tree further away from the scene and the character started moving again. approaching the tree showed, that there still seemed to be some kind of collision box around it, prohibiting the character from even approaching the tree. I tried finding a solution for this for probably 5 days straight with very little success and all the advice from the internet about removing collision maps did absolutely nothing and I couldn’t find a single article or tutorial to the problem I was facing with this so eventually I had to admit defeat and that I would not be able to set up my game like this in Unreal engine…

I used substance painter to texture all my assets, using only materials provided on substance share.

for the mushrooms, I used apple, banana and potato textures in various overlays and for the tree, I used the rough wood texture.

I also modelled the tree  and the treestump in Z brush, however this time I kept the polycount fairly low so I wouldn’t have to retopologize the models. As both models would only exist once in the whole scene, they didn’t need to be as low poly as other items.

In order to retopologize the mushrooms I followed Manos’ tutorial on retopology that he provided during one of our Animation Practice classes