Abyssonance Postmortem

Project Abyssonance for GDTV jam 2024, postmortem

The first day I started working on the concept was a day before the theme announced

It's to establish the world building called "Sorcery Tech", which is the combination of sorcery and technology

The inspiration was from Magitech from FF XV, though their fans don't like this one but it's the first FF which I actually played long enough since it's the first one which changed into action game instead of turn-based

The problem of Magitech is the distinction between magic and technologies

They're generally apart from each others and the idea of magic within FF seems to be rather on the superstitious side than the norms

Therefore, I used the world building which I already have and combine with the tech side

Essentially it's the technologies built from the universe which based on sorcery laws instead of physics laws

----

While I had the general idea of what type of game I wanted to make in mind, which is a linear action game with simple narrative, it doesn't mean I have all solutions

Basically the narrative in the beginning and the end are there to present the idea of simple narrative

As for the linear action game, it's essentially using this formula; Start => Normal zones => Boss arena

This is the simplest formula I came up with after several attempts to extract the level design philosophy and essentially simplified them into the way I could make easily

Well any gamers would know about this formula, but I doubt it'd be on their active minds to use them as level design right?

----

Despite having so many years of experiences(10 years), I only have wave spawner and target lock systems in my toolbelt...I actually wandering what I'd been trying to do for so many years

So the narrative system I developed during the jam time was just simply work but can't be expand

It'd be so much pain to work on narrative writing if I have to use this system, because it couldn't take the next line...well maybe I'd have to add "slash n" ? I don't know, but it's not very intuitive for narrative writers right?

There is also the problem of the lack of QoL systems(settings and key binding)

While I have extensive skill in pseudo-code to make various algorithm which is good for working on gameplay systems, it's not really useful when I need specific information about keywords in order to work on various things which influence more than just basic algorithms

----

As for the theme "Last Stand!"

Well the moment I heard the theme, I knew people would went for survival stuffs or food stalls, though cowboy is a nice depiction since it's the cult classic when it comes to "Last Stand"

But I don't want to make games which have gameplay different from the one I wanted(Linear action)

So I decided to just make the narrative which fit the theme, world building and my gameplay choice

Essentially the idea of the story is "There is no other guardian, so you are the only guardian here. Deal with the enemies whom already breached our base"

----

Aesthetic choice

I knew that neon-aesthetic will always have spaces in people's heart since it's such a charming aesthetic which I also personally love

But the choice was made from the context from world building rather than my personal preferences

Though I could choose between scrap mechs style or clean hi-tech mechs, I decided to just go wild with the idea of "Sorcery Tech" and make use of neon-aesthetic to emphasize on the otherworldly energies like mana and stuffs

Even the design on stuffs are not based on physics engineering and have the "Arcanist architecture" instead

The most obvious thing is the Nexus, which is the crystal in the boss room

But even the pistol in the player character's hand made by using this design

I made everything floaty with vague and questionable assembled instead of fully connected mechanism

While it's simpler to make in Blender, it's still provide such dreamy design which is rather hard to find in games nowadays

----

AI tools integration

I have to admit that I'm using Microsoft Copilot to help me in various stuffs

But I'm not entirely relying on the AI, since I used it to help me find relevant information within the contexts I provided

Which helps me save a lot of research time, it's a lot more time saving than you might think

As for the images from the main menu, the operator(Rosarita), and the credit page, which generated by using Microsoft Copilot

I made them since before I knew about the jam, so they're purely from my personal interests than trying to make something for games

Some of my fellow artists got intense rejection for AI images

It seems the usage of AI images bothered them so much that they refused to play the game

While I respect their decision, I still think it's fine to use within scope

For the music, the base of each songs generated using Udio AI

But I remix them a little to make them more fitting and more to my liking, well basically I wanted something similar to early days Hatsune Miku

I actually made these songs long before the jam, so it's just another happy accident here

----

Now let's get into the important question that I believe everyone also have

"How would you integrate AI tools into your workflow"

Simple, just use them to enhance your workflow rather than replace

As I mentioned above, I used Copilot to help me organize relevant information by provided contexts

So if I did my own extensive researches, which I'd done for the Sorcery Tech world before I actually establish the world building

Then the AI will help me come up with ideas or relevant information that I could look into to further enrich the concept

As for images

Well I'd say it's better to use them as inspiration instead of directly use them inside the game

People could tell if you're using AI images even without mentioning it

While I don't know how would you want to work around this, I'm sure a lot of people already have their own workflow established

More importantly, it's crucial to establish your own identity in games

For my game when we look passed the AI images, the monsters are already kinda unique in a way right?

That's my attempt to pave my own path

Also, the current generative AI won't be able to generate the exact vision of the images you want to have especially when you're someone with great visions

Unless you're kinda "mid" in this case, don't bog your own potential by using AI to replace your workflow, use them to enhance it

I have to mention, I spent 5 days just for the models alone

There is also like an hour that I spent to make SFXs after I'd done with the models

By models, I meant including material slots and animations

I think the most time consuming part was the modeling itself

The shapes and topology are the most important part anyway

I'm kinda having fun when making animations, so it's not taking as long

----

Right...Animations

The clunkiness which you found while playing the game actually came from the animations and how I implement the combat system

The attack mechanics directly bound with the animations, so if you cancel them before they reached the trigger point, the attacks won't launch

The false lies in both wrong frames calculation and support systems

So if I restricted the player's input, then there won't be animation cancelling problem, though it's also provide some exploiting tricks that allowed you to rapidly fire the pistol and deal hefty dps or using mana blade in quicker succession than it should be

The other is how I didn't take into account of how much frames for the animations would influence the game pacing

That's why the game combat ended up as similar to old school RPG, which wasn't my goal but I didn't have time to think carefully about it until now

----

About target lock system

I'm not sure if there is any problem, but the tutorial mobs are not well made, that's why it's not obvious when selecting

Even the mobs and the boss which designed for target lock got the problem of not obvious enough

I could try to find a good solution for it, but I can't guaranteed if I'll be able to make it

I mean if it's too hard to make, I'd use a more cartoony solution instead

Someone said that they couldn't lock the boss, but I don't know why

There is no replicable way to allow me to see it

When talking about all the mechanics, they're rather straightforward

Target lock system is the one I extensively work on for quite a long time, so it'd be hard for the system to make any bug since I generally fixed all of them

----

I think these already covered the jam build

So let's talk about the future for "Abyssonance"

Before that I'll have to mention that there was a project I worked before the jam period, it's called "Arcane Horde"

Which as the name suggest, you're a mage and you have to fight waves of enemies

I actually was ready to move toward the next stage of development and potentially finish the vertical slice version

But I didn't have clear vision for the project along with the jam was a day away, so I decided to wrap up the project and start working on the ideas for the jam as I already mentioned above

Originally I was thinking about using generic fantasy as the theme for Arcane Horde, but as the name suggested, it's "generic"

Then thanks to the jam, I found the theme that I wanted to work on

Sorcery Tech

It's combined half of my interests into one and potentially simpler to develop

The other half of my interests are hard to explain but let's say they're Eastern stuffs + nature friendly stuffs

----

The future of Abyssonance

As you might've guessed, I decided to scrapped everything I made in Arcane Horde to make the foundation for Abyssonance

So this is the first ever project which have rather complete initial concepts

I mean the previous projects I made were all vague and kinda lack the proper direction, that's why I never make a proper game before(though there were a bunch of tech demos and stuffs)

Not to mention the jam build was considered as a success in my part

It's already shown everything in the conceptual forms

I'd normally called this as "conceptual build" which will shown everything but maybe lacks the proper pacing or how stuffs will actually work later on

So I'd have to continued to work on all the systems until they're polished in order to make the vertical slice

I could make horizontal slice, but my game don't really need it since the game would be rather simple and relying on narratives and variation of enemies instead of having a bunch of questlines like how Tim Cain would do

----

As for now, I'm still in the process of figure out how I'd want the game to be

Simply put, game pace

As another action game, I have to consider the bar of DMC and Elden ring, which I'm leaning toward DMC but should be a bit slower and not focus on combos

I tried to adjust the variables in order to simulate the feeling, but it's ended up as a failure

So I'll have to make new animations in order to work on this part

Then I'll introduce the special resource system, think of it as Mana but a bit more complex

I'd love to just use mana, but it's not making much sense in the world building context if I only have mana to use

Also there is the idea of having variety of player's abilities to choose from

This way the game won't be staled when playing more than once

My plan is to have limited levels along with rather short narrations

I wanted to make a concise game in order to learn about publishing on steam

At the same time, I think it's alright to make a short game with some fun and sell cheaply

There were many short but memorable games back in my days too

That's why, this is my goal

Finish up the combat system, adding some alternate abilities, and have a concise narrative

Well I nearly forgot that I still need to work on the narrative system lol