jacobhallunit04

[jacobhallunit04]

My Evaluation Of This Project

I’ve had an incredibly good time throughout the entire duration of designing and developing my Random Character Generator, with only 2 occasions coming to mind where I didn’t feel like I was progressing, both of which were after achieving the MVP for my RCG.

I’ve learnt much more about modelling with making each and every one of the models by myself, which in turn has made me more confident with 3D modelling, as well as being able to practice skills I learnt last year such as blog writing and researching for ideas and for documenting my progress throughout the project and while I’m still somewhat unsure about my coding ability I feel more confident about being able to do it then I was at the end of last year. Alongside feeling like I’ve performed well in nearly every factor of my work I’ve also enjoyed my time thoroughly, with it feeling like I was more actively pursing opportunities to work on something I hadn’t started yet or on an idea that randomly popped into my head as well as just being more dedicated to continuously progressing on any factor of my work I was already working on without getting distracted by something.

The only times that haven’t felt like that were in the last couple weeks due to needing feedback from other students after having achieved the MVP, which I didn’t have the confidence to seek out myself, which caused downtime where I didn’t feel comfortable adding/working on features as I personally felt it would invalidate any criticisms made of the final product by last minute changes that weren’t intended to be fixed before feedback, although I was also dealing with some kind of artist’s block at the same which made it even harder for me to find work to do.

In summary I’d say this has been the best project I’ve done on this course so far, I’ve learnt how to work in a new factor of game design, I’ve gotten better at documenting the various references and work I need to for this course, I’ve gotten more experience with coding and using the Unreal Engine, I’ve felt positive of all my work throughout the entire half term and the only times where I haven’t is because of who I am as a person and not because of the course itself.

[jacobhallunit04]

The Download For My RCG

This link leads to the downloadable .zip file for my RCG.

https://thekirboi.itch.io/rcg

[jacobhallunit04]

My RCG As A Finished Product

The link below leads to a short video of my completed RCG functioning properly in a game setting with all of it’s features on display.

https://youtu.be/H_k21xc16b0

[jacobhallunit04]

Last Minute Improvements/Changes To My RCG

I finally decided to fix the lighting issues I previously had in the game world by rebuilding the lighting and removing the roof I originally had on top of the play area and I also fixed an issue where some of the models I spawned were spawning outside of the map. Below is how I fixed them:

The lighting:

Previously the play area had 2 incredibly dark shadowy areas due to 2 pre-placed rectangles, that are included by default in every “FirstPersonExampleMap” to help the developer test things such as character jumping, which I deleted in order for the characters to have more room to spawn.

The shadows-

The build tool-

To fix this I had to rebuild the lighting via “Build Lighting Only” option inside the “Build” tool seen below:

What then happened was the entire play area room becoming pitch black due to the roof blocking out all the light, although all of the models spawned in being entirely visible due to the fact they aren’t coded to work with shadows. I unfortunately don’t have any screenshots of this as I fixed the issue and saved the project after removing the ceiling and rebuilding the lighting a second time before realizing I should’ve taken some screenshots of the issue before fixing it for this blog post. However I do have a screenshot of the final product, which is right below:

The incorrectly spawning models:

A couple weeks ago I noticed an issue with my models where the spawn point for them wasn’t centralized which was causing there to be an issue where the models were pretty much only spawning in one corner of the play area, with the 2 opposing walls looking like they had some kind of invisible wall preventing the models from being spawned too close to them. After correctly relocating the central spawn point I decided to increase the distance the models could spawn away from the spawn point by 2000 and -2000 from 1000 and -1000 for the X and Y locations. However today I noticed that occasionally no models would spawn when pressing the Q button to spawn the models, so I decided to temporarily set the models to spawn every 1 second instead of via a button press that overlapped with the controls of the spectator view and then entered the spectator view to see where the models spawned and, as I had predicted, they would rarely spawn outside of the map and fall into the abyss. Luckily there was an incredibly easy fix by just lowering the values from 2000 and -2000 to 1500 and -1500 respectively.

Edit:

I was incorrect about this. I originally thought that the way the spawning worked was based on an editable, invisible box around the spawn point’s location in the world, however it turns out that once the spawn point is in the world only the point it was was dragged into is where the origin of the invisible box, with wherever it gets moved to after that not doing anything. This meant I needed to spend a good chunk of time changing each location value, with having no idea of which value was in command in each direction, so that the empty space caused by this wasn’t missed by the box and the parts it were close too didn’t have the box over reach and cause models to spawn above the abyss. Below is the final measurements I settled on, which makes the invisible box hug the 2 side walls and have a small gap at the front and back walls to allow for easier viewing of most of the models:

[jacobhallunit04]

Feedback for my RCG 3

Unlike these previous types of blog posts the actual thoughts were written down by the reviewer instead of a recount of the reviewers thoughts by me. Below is their thoughts:

As a prototype this is looking really good, its highly functional and an excellent proof of concept. All of the characters are very distinct and unique from one another while also being easy to identify what they are. It’s difficult to see where the characters segment and all of the randomly generated components fit together seamlessly.

The only issue I can see with this system is trying to animate the characters or adding another system to it will be a lot of work. Not that it needs another system to prove that it works but it could be difficult to implement into another project.

I think it could be interesting to try adding randomly generated names for each of the characters as well to give them unique identities as well.

------------------------------------------------------------------------------

I really want to try and implement the random names for the characters, but I am unsure of how to do it and am worried of accidentally breaking another part of my RCG so I’m going to ask for help to see if it can be accomplished.

[jacobhallunit04]

My Reflection On Last Week’s Progress And What I Plan To Do This Week 3

Last week was a very slow week for me as the main thing I needed was feedback from others and I wasn’t able to receive that necessary feedback from anyone but 1 other student. However I was able to work on the feedback I was given the week before which from the one other piece of feedback I was able to get cementing that it was a good design choice and I myself was able to give a couple bits of feedback to other students on their work, which I am really glad I was able to do as it felt strangely satisfying. As for the last couple days of the term I intend to hopefully get a least a couple more pieces of feedback and to capture footage of my complete RCG for this blog.

[jacobhallunit04]

An Issue I’ve Had Today

So far today I haven’t been able to do much with my project for a couple of reasons. The first one is because of a lack of ideas. I’m pretty much at the end of my project, in reality the only additions I should adding is from feedback given by other students. I’ve added everything I originally wanted to add since the beginning and it all works perfectly fine, but this has naturally led to an issue of making today very stale and stationary when it comes to working on my RCG. I’ve been waiting most of the day for feedback to then add things to make my project better, but without it I’ve spent all day with nothing to do, not being able to think of anything to add, but my second issue is that any ideas I have thought of I haven’t felt comfortable adding, as to me it feels like it makes any previous reviews feel disingenuous, due to possibly changing factors that go against praise that was originally given at the time. I’m hoping that this issue will be gone by tomorrow as I should be able to receive more feedback and, hopefully, be able to be more active with possible additions/changes to my RCG, as today has been really boring for me.

[jacobhallunit04]

Feedback For My RCG 2

Similarly to my first piece of feedback praise was given to the models and how all the code for it worked, as well as the additional praise for the physics I added from the previous feedback. However when it came to feedback for possible additions/changes the reviewer had only one possible addition that came to mind, which was allowing the models to follow the player around once spawned. I probably won’t try to add this idea as it’s simply to hard to try and get to work with the already existing code and it doesn’t fit in too well with the idea of it being just a random character generator with the idea of every model following the player causing them all to bunch up and be difficult to see the differences of each and every one, whereas just the physics allow the models to be fully viewable, just with a small amount of interaction to help the player feel more involved.

[jacobhallunit04]

Improvements/Changes To My RCG Based On Feedback 1

I’ve now messed around a bit with adding physics to my models and so far I see a few choices for how I can use it in the final product.

The first is by using the collision type “CharacterMesh”, which when used allows for the characters to be shot and moved, which when paired with the invisible cylinder spawn for the models allows them to be shot down and then rise right back up similar to a floor-based punching bag. The only issues with this take on the physics is that the models will occasionally spawn inside one another and the other option for the physics I’ve tried so far can give a more satisfying and dynamic result.

Another collision type is called “Ragdoll”, and makes the models move around a lot more, although it has more notable issues with the models then the previous option. The hitboxes for the models are actually an invisible cylinder around the models, which are needed to help with the model’s spawning and with one of the pieces of code allowing for different heights, widths and sizes for each and every model also affecting that invisible cylinder, causing some of the models to appear like their floating. Something I don’t understand about this issue however is that of the cylinders becoming enlarged around the models only happens when using the ragdoll collision type and when using other collision types the size of the cylinder is much tighter and doesn’t seem to act separately to the actual model’s. Another issue is that while the model’s physics are now more dynamic and humorous to watch the models tend to either spawn in the ceiling, due to the top of the invisible cylinder’s spawning in them, or they clip with another model and bounce around the room, which can defeat the point of being able to get a good look at each randomly generated character.

I’m most likely going to use the “CharacterMesh” collision type as it causes less issues that aren’t very noticeable.

[jacobhallunit04]

Feedback For My RCG 1

Currently my RCG has received positive feedback for the models, the code for them and the variety for their appearance of each randomized character and the received feedback for potential improvements has just come down to adding physics to the spawned characters so they can be shot around and make them able to collapse over each other like dominos, and while I will try to implement it I’m mildly concerned that doing that make cause the issue of flying, possessed models again.

[jacobhallunit04]

My Reflection On Last Week’s Progress And What I Plan To Do This Week 2

Last week I was able to successfully finish coding all the requirements for my RCG to function properly and I received 1 piece of feedback on it. When it comes to what I need to do for this week I could use some more feedback on my work as I currently only have one person’s view on my work and other than that I want to add physics to my models in the RCG, which I want to do because of that feedback, so they can be shot and kicked around once spawned in to make them more interactive, although I might need some help to figure out how to make it work correctly, as the models are connected to an invisible cylinder to help them spawn and this could cause issues with possible resizing.

[jacobhallunit04]

Getting My RCG To Function In An Actual Game Setting 2

I’ve successfully exorcised the devil that haunted my attempt at getting my RCG to function properly from last time and have made it so that not only can I make random characters spawn at random locations at the press of a button, with an accompanying sound whenever you press that button, but they also spawn with random sizes as well., with a screenshot of the game world and the code below.

My RCG:

The Code-

The Game World-

The Viewport-

I reason the models were behaving they were in the video from the previous blog is because to make them fall, which was the initial idea, they needed some kind of physics to let them fall which was granted by an invisible shape (a cylinder), although I still don’t know what caused them to fly away, but I have a theory it had to do with clipping with an invisible spawn in the sky, causing them to gain massive momentum in the effort to get out of the spawn point. Before arriving at the images seen above I decided to test the idea I had in the last post and had their spawn be random, although with a limit of how far away and always at the same height and I noticed an immediate difference in how the models behaved. While they spawned randomly they would still end up possessed and go flying away, however now they all went the same direction instead of wherever they felt like going. I have absolutely no clue why they behaved like this, but my best guess is basically the same as my previous hypothesis, with the easiest direction for the models to try and escape the invisible spawn being that specific direction. I fixed this issue by removing the physics entirely, as well as the shape that allowed for the physics. This meant that the models no long tried to clip out of an invisible spawn and all I had to modify from there was the height they spawned at, although I took it one step further with the sizes also being random, with some models spawning normally, some having the body structure of SpongeBob and some with the mild appearance of Slenderman. I feel like I’ve accomplished all of the requirements for my MVP, I’m really proud of all of the work I’ve managed to achieve and now the main thing I believe I need to do is receive feedback from others on any other possible improvements I can add to the product in the last couple weeks.

[jacobhallunit04]

Getting My RCG To Function In An Actual Game Setting 1

So far I’ve made concept art, modelled, positioned and (somewhat) coded everything for my RCG correctly, however when trying to get my RCG to function in an actual game setting there are a couple of issues, which can be seen in the video below.

https://www.youtube.com/watch?v=ETilow1oAps

I don’t fully understand what causes this issue, nor how to fix it, but I find it funny and part of me wants to keep an element of it in the game somewhere, although what I’m more likely to do is reuse the code I learnt a few weeks ago for making random cubes fall from the sky, as I feel it will suit what I want to achieve better.

[jacobhallunit04]

Coding My RCG 7

I’ve finished coding and remaking the glove sprites with images below to show the changes.

The RCG:

The Code-

The Viewport-

Gloves 1:

Original-

Remade-

Gloves 2:

Original-

Remade-

Gloves 3:

Original-

Remade-

The main change I thought I’d have to make to the hand models was the size, as they’re all pretty much as big than the body models they’re attached too, which was a major oversight by me when making them, as well as the fact of needing to making them into singles instead of pairs, as if I didn’t there would be random gloves would be either inside the body or floating off to side of the body. However this wasn’t the case, with the one thing needing to be changed about all of them was the sleeve being 1 voxel line to small, with the hook probably taking me the longest to remake due to that one voxel issue as well as the fact the model had the main issue I originally thought they all would have, with the kind of unfortunate payoff of massively preferring the originals, although I don’t hate the remade look. With the blue and fingerless gloves I had to change their placement in MV, similar to the hats and hair models with the hook taking the same role as the blonde hair, as well as making the model’s pinky finger a bit bigger, similarly to the hook’s case. Specifically for the fingerless gloves however, I needed to make them much fatter, as the original glove model was too small, with no real way to make it taller without making the glove look really weird at the time, and as such the sleeves were constantly clipping through it’s base. In hindsight I’m able to think of one way to possibly fix the height issue, being to have premade “arms” coming out of the sleeve to connect to the shirt models, but it goes against the idea I had in mind when creating the glove models to begin with and it could’ve quite easily led to more issues, which in success cause more modelling time instead of coding time, than fixes.

[jacobhallunit04]

My Aim For Today

By the end of today I plan to have finally sorted out every single model so they all slot into place in my RCG with no issues as well as hopefully have a testable product for other people to give me feedback on.

[jacobhallunit04]

Coding My RCG 6

I’ve successfully added the arms to the shirt models, which should make positioning the glove models easier in general.

Shirt 1:

Original 2.0-

Remade-

Shirt 3:

Original 2.0-

Remade-

The only thing left to do with these models is to actually import them into Unreal and then I’ll work on the final models which are the gloves.

[jacobhallunit04]

Coding My RCG 5

I’ve finally fixed the issue I had previously, now every hair and hat model are synced up properly with the head models and they all share the same spawn point.

My RCG:

The Code-

The Viewport:

Hat 1-

Hat 2-

Hat 3-

Hair 1-

Hair 2-

Hair 3-

The only thing I changed for most of these models was their location in MagicaVoxel, except for hair 2 and hair 3. For hair 3, nothing was changed at all, it was the base for the height of all the other models, as for hair 2 I needed to add a lot more “hair” to it as a lot of it clipped into the head model at the height it was supposed to be at and I think the result of needing to add more makes the model look much better and more distinct. The next thing I plan to do is like add arms to 2 of the shirt models as well as add some kind of feet to the sandal models.