fmp-tdgamesterm3

[fmp-tdgamesterm3]

Changing UV skin and eye colour

Notes: Eye hand done, used:

UV Skin uses:

https://www.diyphotography.net/quickly-easily-change-skin-color-photoshop/

and then:

[fmp-tdgamesterm3]

Developing audio for my beginning.

Premise

I new for my mysterious slightly creepy begining I wanted to go in one of two days, either make it mysterious but uplifting or dark and mysterious. After having developed the map, having had looked at lots of reference pictures and played a couple games to get ideas.

I realised I was going to have to go with the later as it’d allow me to create a sense of unease that would lead the player to be more curious instead of taking it at face value.

This isn’t to say I didn’t experiment with different audio recordings with different effects. I even played David Attenbrough under the game to see how it felt.

Making it

I tried out many different ways to achieve the affect after recording the voice myself. I first tried https://voicechanger.io/ which is a great site and allowed me to try lots of different affects such as: CustomVoice-Creepy [<-- Link]

In the end I downloaded Audition and filtered out the background static,

Tried adding some effects here but it didn’t sound quite right, I then followed the following tutorial I found:  how-to-create-a-mysterious-voice-filter-in-audacity. This was extremely insightful and after downloading Audacity as well, I had the affect I was looking for, more or less.

Voice Track below:

This could take a while...

converting buffer to a .wav file...

This could take a while... converting buffer to a .wav file...

If this fails as i’m routing it through a Voice Maker buffer. It can be downloaded off their servers here:

blob:https://voicechanger.io/583fd621-6304-4cff-9a74-ca2be773d63b - (You’ll need to reload the page when it loads)

[fmp-tdgamesterm3]

My modelling process

For the last few months it’s been a pain importing a piece of work from Maya/Blender into whatever development engine i’m using and seeing it looks completely different from what I had, this is down to Rendering, the real time rendering systems in Unreal/Roblox/Unity all have different looks. Unreal looks great, Roblox looks bad and unity looks unity. This is an issue as it makes it much harder to easily reach your desired look.

Due to these issues I have started using my own way of handling the design process, which allows for more incremental changes: (I’ll use an example to explain.)

1. Research, I knew I wanted a low poly look as Robloxs graphics is lagging behind other market leaders until they release their massive overhaul [scroll] in May/June + it’s much quicker than using something such as Substance Painter + Maya + Photoshop. This is key for such a short timeline and large world.

Aside praticality and design style I wanted someting quite magical while not sticking out too much and looking natural, in my opinion Vesteria balance this really well.

For this I decided to have a look at Vesteria as I wanted some shrubbery:

They make great use of different interchangable designs and parts to make lots of slightly different trees etc, very quickly.

2. Design it, I find it difficult to visualise how I want my design to feel and look into a formation of triangles on Maya as im something of a perfectionist. I’ve found that simply making a number of different designs quickly and keeping them all in the viewport eases pressure as it allows me to design freely and make fine improvements without worrying it may mess up the design, as I have backups. This really helped.

3. Testing, Before I import anything I create clones of every object seperatly and apply a harden and soften to each respectively this allows me to import lots of different alterations of the same design into maya without the worry of re-importing or not being able to reverse the smoother. (An issue I had designing the character.)

4.Importing, This is a very simple process and only takes around 30s. First I create a folder on my external SSD with the given name of the object as a whole:

I then save and export, all as OBJ as roblox weirdly handles OBJ better. (Discovered through trial and error)

I then, click Import Meshes, wait for it all be approved or not (and r-import with fix).

5. Building, Once imported I take however long to build the object up. I find it best to import as seperate objects so they can all have custom properties and I don’t have to rely on Textures to the same extent.

Finished builds:

[fmp-tdgamesterm3]

Weekly Reflection

What Went Well

I managed to get my intro island done (aside from importing a few more assets from Maya) as well as Rigging my character, Importing it into Roblox and successfully getting it working on roblox. I also coded in a firefly system an got a fair amount of game design and research.

I’m most proud of getting a proper process for model design, building, importing and refining which is much faster and in my opinion yealds better result as it allows for lots of small incrimental differences as how something looks in Maya and Roblox can look quite different.

What Could Have Gone Better

While I got a lot done I didn’t manage to finish NPC interaction OR Leaderboards although there was some leaderboard progress. This was mainly down to focus shifting towards aesthetics this week and it generally being an odd week in which i’ve finally recovered from my illness.

My main issue this week has been of focus as being stuck at home for three weeks has started to really sink in ( quarantined before mandotory for every school). This lead to less blog posts being written which I intend to pick up on monday.

What you plan to do next week

Initial cutscene done including: Asset placement, Musical score, Animation + What I left from last week now im starting to refocus on the scripting side + continued game design posts.

[fmp-tdgamesterm3]

Rigging my character

Issue

Previously I had imported all the model parts into Roblox Studio and rigged them there but due to infuriating Humanoids this was very difficult as you touch one wrong thing and the body positions reset. With this in mind i’ve decieded to use their rig imported and instead rig it in maya. Following this:

https://developer.roblox.com/en-us/articles/using-avatar-importer/index.html

Possible Solution

I have followed Robloxs instructions and have laid out the Bones in the body like so:

I had an issue where as I hadn’t set up the Positions/orientation etc to 0 before exporting the “true” 0,0,0 isn’t the actual one. If I were to do this agian that’s something I’d correct. I followed the example build roblox offers for Rthro for this, the hierarchy.

Including their naming structure. I haven’t added the Attachments yet as I want to export it and make sure it works as expected.

[fmp-tdgamesterm3]

Skill Trees [Research Post]

Skill have long since become and essential part of many games, used in may different ways for many different reasons to offer a sense of progression and to stagger the game and it’s difficulty curve, personalise it.

That said a lot of games use it for the wrong reason, just because it’s what games have. I’d hate to make a habit fo picking of Skyrim as it’s done so much right but their Skill Tree system lends to allowing players to farm points by using repetitive actions. This isn’t exactly their fault although they could use wheights and balances so the higher you go, the less these more simplistic tasks help or do what Red Dead Redemption 2 does and use a hidden system that lets them know they’ve got more health etc but hide the process behind it, driving the player to play their way to unlock what they want instead of playing the system.

Personaly I am a fan of skill trees but they can be overdone like in path of exile. They CAN offer a more natural progression if done right while also allowing the player to be analytical instead of just mashing some heads in but it can be costly if done wrong or simply for the sake of it. For instance they shouldn’t remove key skills the player should have to begin with or as Game Maker Toolkit says, remove expected basic skills existing in similar games and award them later on.

[fmp-tdgamesterm3]

Weekly Reflection:

What Went Well

I got a lot of productive work done where possible including:

- Lots of work on main character.

- Performant Firefly System (hundreds of moving parts all with light elements).

- Initial Cutscene place semi-built and planned out.

What Could Have Gone Better

While I got a lot done, the coronavirus hindered my progress with my mother falling ill pushing us into a 14 quarantine and later me falling ill. Leaving me with few hours of the day (3/4) to get work done or any activities as I was on bedrest.

What you plan to do next week

I intend to get a working first cutscene, finish npc interaction + complete leaderboards.

[fmp-tdgamesterm3]

The Tomb Raider Problem [Review/Research]

I really wanted to enjoy this game and appreciate it being free for a limited time. With that said I found the game to have little thought inducing tasks and felt more like a film with the number of cutscenes, even just to go under a log. Move five feet = 2 minute cutscene.

I remember back near the beginning of term, Sam showed me a research blog post about how Cutscenes break gameplay and pull the player out of the experience by restricting how they can interact with the game.

By cutting out how the player interacts with the content, you are essentially starving the player which is why games like portal have figured out ways to communicate with the player without having to stop by the old man with the van out back for some wisdom. For instance in Portal 2 your companion travells with you and can move around you as you move, this is essential as it allows you to 1. Do other things while listening in.

2. NO CUTSCENES*

num 1. is another issue many cutscenes have, even ones where they don’t cut you off completly from the controls, for instance in Skyrims intro their is dialog as the player is being transported place to place but they either have to focus on what they’re saying OR take in the visuals; This gives the effect that the player feels they’re missing out if they can’t hear what the characters are saying as they have to turn to each other character to listen in.

[Post: https://hitboxteam.com/designing-game-narrative ]

I feel as if Tomb Raider while holds a lot of promise and to many seems to be a very good game and series but the 2013 version felt like it optimised what you shouldn’t do with a video game.

It was a very linear structure with little room for player expression with moments where you could sit back for 5-10 minutes through cutscenes and do nothing. This lead to some disinterest from a player standpoint as you don’t feel as if these are your accomplishments, more like one of those interactible movies on netflix; With a linear journey and a few button presses.

*Cutscenes can work a dream, especially when there is some form of interactability. Without which you are removing that extra level of depth which snaps the player out of the experience; Which may be intended, for instance at the end of a level but generally is a bad idea.

[fmp-tdgamesterm3]

Developing my Rthro character

Creating my Rthro design was not very easy and I ran into a number of issues, mostly while in Maya.

In roblox they have have a number of base meshes and builds for Characters but they don’t include the newer Rthro. So I took an existing Rthro model from their huge collection in the virtual store and inported it into Maya by:

1. Opening up Studio, Entering a play test

2. In play test Copy my Model then exit.

3. Paste into workspace then export as an OBJ.

I then redesigned the Upper Torso and Clothes over that, using both of the following tool:

The following was used for the Clothes over.

As it’s quite hard to make something symmetrical, had tried and failed but decieded to delete half of it and mirror instead. Much quicker.

My biggest issue aside one i’ll touch on in a minute was with the upper torso, when imported into roblox it appeared very low poly and for some reason I could not correct this in Maya, I had tried the following to fix it:

-- Importing it into Mudbox + Mayas sculpt tool to soften.

- Using the smooth modifier when exoprting.

- Using “Smooth Edges” under Mesh display.

But in typical Maya fashion none of these fixed the issue as Maya had disconnected all the Verticies along the way. In the end I opened up Mudbox, selected the default human rig and exported into maya where I cut it up and kept only the Torso.

This worked the dream so now I had to export this into Roblox and doing so only took a minute or so but when reassembled (Roblox doesn’t allow full meshes for character rigs, need to be seperated into components until Character Evolution is made default).

Once within Maya and cut up, I needed to bend the torso somewhat as it was the wrong shape for the clothes I had designed:

Seperating the png’s for editing from the original OBJ so I could create new UV’s. This was simple enough although I did run into a couple issues, I did it by:

Setting up my Rthro to function in game:

Roblox uses old Humanoids for charaters (Another thing being replaced) and there’s a system for importing Rthro characters (using: https://developer.roblox.com/en-us/articles/using-avatar-importer) but I decided to do it by hand for more control by taking the neccesary constraints etc from my original Rthro base and psting them into my Mesh objects:

And then configure them for my build, this was a long process and in the future i’d probably work with the official way but worked well!

[fmp-tdgamesterm3]

How the original Half Life influenced my game intro

While doing some game design research for how to introduce the player to the world I stumbled across a “Extra Credits” video about Skyrims opening:

The point which hits home most from this is when Daniel Floyd (or Matt Krol) explain the importance of the first five minutes of any game, How they are both the hook that draws the player in:

And the context that sets the player in the correct emotional state to get the most out of the experience they’re about to have.

This really drove home for me as in pretty much any MMO, the first thing you do when you join the game, after the 10hr download time, is to set up your character. This is something I and, I assume, most love about RPG’s. Getting to carefully do research on each characters, looking over multiple sources and too many subreddits in which you learn little aside everyones got a strong opinion.

Now while I LOVE this customization, it’s also extremely unimmersive and you don’t feel like this is really your own path, more ticking off the parts you’d like to order on that maxed out credit card on amazon.

It’s issues like this which drove my enthusiasm to the genre, trying to find my own solutions to the issues that haunt it as well as how much potential the genre has.

Half Life:

While I understood the issue I couldn’t think of a viable way to sink it in with gameplay, until I played the original Half Life.

I’ve been a huge fan of Valve for a number of years with my favourite series of their obviously being Portal, especially the second. My interest in Half Life spiked after having watched a “Half-Life 2 Developers React to 50 Minute Speedrun” for research and found the game was free for a limited time, I took the opportunity and played the original (with those crisp graphics, not good, crisp).

While the game was fun, if a little dated (having been released back in 1998) the noticable stand out to me aside the..interesting AI, was how the player was introduced to the world while teaching them the mechanics.

Now I know in Half Life 2 they have a much improved and less straight “here’s a tutorial”, opting for a much better, teach through play and interaction using and building off basic logic.But the original gave me somewhat of an epiphany.

By opening the player up to the world in the environment and let the player assemble their suit and such while learning about the world in a 3d space to be very endearing and could be utilised in the players character customization so as they PROGRESS, they get to customize in the 3D space so the character customization is no longer a check list but a further immersive area of the world building up a deeper level of realism , and, well depth.

[fmp-tdgamesterm3]

R15 vs Rthro

In roblox you have three options for character models, this entire system is soon to be overhauld by Avatar Evolution [Link:** ]**, entering, out of beta, mid this year.

Types:

R6 - Made up of Six Joints, the first iteration, ugly and old.

(Ripped from a Gamejam I did, featuring a custom R6 rig.)

R15 - Made up of 15 joints, nice but very roblox.

(Ripped from Vesteria, an openworld MMO.)

Rthro - Custom morphs, more realistic and customizable, limited to 15 joints.

(Ripped from Pacifico 2: Playground Town)

Custom - Custom, Custom Morphs. Limitless anything but difficult to implement, and finicky.

My Pick:

I was intending to use R15 for this but felt that I, for one, didn’t want to mimic Vesteria and second that with Avatar evolution being released so soon Rthro was just the right way to go as it allows for more scalability down the road.

This mattered to me was I intended and am comitted to treating this project as if it was a real full game, and intend to make decisions based off this. Aside scability for the future, Rthro looks a lot more realistic and allows for greater customization, player expression and immersion while challenging my skill.

Avatar Evolution:

_https://bit.ly/399WSsW | https://roblox.github.io/avatar-evolution/

For a long time roblox has supported meshs and textures coupled with a limited amount of material and cusomization options but this new update looks to bring high end tech to low end devices through roblox and is extremely exciting as a developer.

The reason I chose Roblox for my FMP aside my fondness of the engine and programming language were two things: Scalability/Support and the engine itself, how it functions dynamically for multiplayer games with free hosting.

And while this is a huge selling point for it, it’s always lacked in the visual aspect and while this isn’t too big an issue as it’s still possible to make GTA level looking graphics but you do have to make sacrifices.

New additions now out in beta, no live server:

[fmp-tdgamesterm3]

Developing a Firefly System

This was actually a very interesting challenge when compared to how mundane it sounds. I have around 75 parts all flying around the map, all being rendered with light effects. 

My Initial Attempt:

When I initially approached this I ended up with this script:

It was logical, took all the parts, assigned them a thread each to run  their own code. The issue was overhead coupled with a ill coupling of Raw Threads and Roblox Threads as a developer explained to me. 

The Issues:

1. Overhead

In Roblox we are constantly trying to figure out how to keep thread counts down to boost performance on lower end devices and the server, this played against that logic. Instead creating around 75~ threads and looping within them so they never die. 

Overhead with coroutines is a real problem as they are more exhausting than loops. Thanks again to the developer who talked me through performance enhancement. 

2. The ill advised combo of Roblox and Raw Threads.

A coroutine is a Raw thread while wait() now runs as a roblox thread and is put through robloxs task schedular. Now imagine 75 of these running at the same time. 

This had the jolting result of:

As you can see its working but has Lag and Task Schedulers getting full. This isn’t an issue when you lower the count but that simply wasn’t the answer, long term.

How I fixed it:

Run off heartbeat, not coroutine loops. The great thing about heartbeat is they run *after physics simulation, every frame. So you it doesn’t effect your framerate or physics simulation as it runs after everything else is complete. 

Still using coroutines, I’m still using coroutines, and for performance in the future I may limit the number of coroutines to 10~ then assign randomly. The reason for this is so it can calculate a new Vector3 value locally and in it’s own time. The reason this system isn’t simple is because every move has to have a relation to the last move so this will move in a certain position for one second before changing direction. Without this it just umps around the scene.

Why CFrame? CFrame doesn’t utilise any physics, simply sets the position so no extra overhead but we simulate this physics with the while loop. 

This has the smooth result of:

[fmp-tdgamesterm3]

https://clickitupanotch.com/skin-color-rgb/

[fmp-tdgamesterm3]

Why oh why, Roblox?

What Is Roblox

Roblox is a huge social focussed platform, targeting a predomintly younger audience between 10-18 years old ( #1 gaming site for kids and teens*), Despite this Roblox is not accepted as a viable market for a variety of reasons although non-tech brands are starting to see their relevance for marketing and developers and game studios are reportedly making millions of the platform.

^ This was surpassed.

Why I chose to do it for my Final Major Project

I got a lot of concerned faces when I raised the prospect of making a game on the platform, and while this is somewhat due to missunderstanding of the platform from a professional development standpoint, it is understandable when it’s a “closed” development engine, bad publicity and devalued depite the company having a net-worth of 3$ billion.

I chose roblox for the home of the MMORPG for several reasons.

1.) It’s a great development engine with state of the art services for hosting and developing a large server hosted game. The top game on the platform at 2:21pm (Friday) has 214,100 people playing, right now.

2.) It’s a community centred platform, while I may not be making a fully fledged game, simply making and designing gameplay and mechanics for one. I intended when I started to treat this as if I were making and designing a full game.

3.) Free server hosting, Roblox hosts and maintains all servers (with a size up to 200 in each) for a cut of revenue.

4.) Free Datastore Hosting. https://developer.roblox.com/en-us/articles/Data-store

5.) I know the engine, I have spent years in the developer community and making open source systems, my top from the ground up game having over 200,000 visits and a constant concurrent player count. I also used to maintain a game with over 5.8M+ visits and a large active player base.

I personally have grown up with the platform and while I rarely play on the platform now (faviouring steam), I have grown to learn lua, game and server archetecture, some game design and met many other faces. I’ve also been able to make some cash on the side and teach others.

6.) Roblox is so lightwheight it can run on almost any device including mobile, xbox and VR, and all on the same server with large worlds and proper social tools.  

* comScore

[fmp-tdgamesterm3]

Character Choice/Customization experimentaion

Before I started the project I knew if I ever did an MMO it had to have some player customization. Having played GW2 they have this really cool class choice system with smoke enveloping the differenct classes:

I’m a big fan of this and wanted to include it into my own although I don’t intend to have classes as i’m following Starting Point: Ordinary World intend every player to start from the same place, following the learning, developing and advancing path I found while looking into this point until each player grows and blossoms into their own character with their own strengths.

There were two ways I could go about this fog/smokey effect, images which allow me more control and I can switch between them to create a moving effect or Robloxs in built system which you can see a test of below:

I then have two ways of displaying this, either use my camera system to switch the players camera to this point, or use ViewportFrames to display it within a 2D UI. Example of ViewPorts below:

Each of these do have their limitations,

viewports have the following limitations:

This means it may look somewhat plain and I wouldn’t be able to use the Smoke Effects.

While using the Camera has the following:

- All held by the server

- Less succinct with UI elements

- May need to create duplicates of the same area so more than one person can edit at a time, lag/waste.

[fmp-tdgamesterm3]

My Mock Up

Background is from a map made previous to the project as some concepts for Alterdrift. 

Leaderboard/Stats made from hand today (12/03/2020) using symbols from Black Ops 3 and colouring’s from Black Ops 3 and Vesteria. 

Hotbar ripped from an early version of Vesteria which I just loved but was sadly removed. Very difficult to find. 

Improvements to the Leaderstats were suggested by Sam Gibbs and Sam Read (Tutor) as the Leaderboard stood out too much so after using alpha channels and masking I reduced the opacity (wasn’t a simple process as it had more than one level with the gold stretching box behind).

Another issue I discussed and crafted a fix for was long usernames, i’ve decided after some debate, to limit it to eight char and add ‘...’ to any shortened versions. 

[fmp-tdgamesterm3]

Gantt Chart

https://docs.google.com/spreadsheets/d/1gS7qxGpPnXR0pw44GlH96VLvNoNbC0sXCte0DdLtPM8/edit?usp=sharing

Laying out a rough approximation of my focus each week.