|
I've realised that although I've made a plan of how long it will take me to produce my assets and make my project I'm not sure I've ever said what my plans are in full and listed what I aim to achieve by creating it.
My plan for my project is to make a functioning, basic endless runner mobile phone game that I can create a full collection of character animations for. I hope to spend the first negotiated study creating the assets need to make it and then leave the whole of negotiated study 2 for the creation of the game and animating the character performing different actions. If I have time during this half of the project I will likely create an animated introduction to use at both the start of the game and to put in the end of year show reel as I'm aware it would be difficult to show off a game on that platform. What I hope to achieve by doing this is a better understanding of the process of making a game, how to create assets to suit a particular platform and a better understanding of how an animation should be prepared in order to easily use within a game engine, with a focus on seamless transitions. This is a similar goal to what I wanted to achieve last year, however last year I failed on my time management and was only able to achieve creating one character, albeit a complicated one, and did not manage to make an environment or game to put it into. This year my time management is going a lot better as I think I'm in a better place mentally and I will hopefully be able to create something entertaining.
0 Comments
I am consistently updating my other post to tick off which models I have finished creating and which UV unwraps I have optimised to better use their space. I am, however, going to create posts showing screenshots and renders of the assets I have made so far. Building assets:  Bench:  Bin:  Bollard:  Construction barrier:  Lamppost:  Planter:  Pipes:  I have recently opened and tested the software Substance Painter for my assets to see how quickly and effectively I felt I could get it done and have concluded that I should be able to texture most, if not all of my assets in one or two days. With correct Lambert assigning to the faces of my bench model I was able to simply use two different smart materials in substance painter to achieve the correct appearance. For the sake of added detail; with most of my assets relying on materials to complete their appearance, I used the brush tool and a slight rise on the normal map to add moss and dirt effects to the wood. I could easily use similar tools to add scratches and graffiti to the bench and disguise any seems on other models.
So far I am modelling my assets ahead of schedule which is great. My plan is to see how long it takes to model my character and then decide if I have time to make the additional assets listed or whether I need to move on to UV unwrapping and texturing to make sure I have things done within the time frame.
I'm running into the problem at the minute where at home I use Maya 2016 whereas the computers in the Uni buildings all use 2015. This means that some of my UV unwrapping has gone wrong and will need to be fixed if it doesn't work fine at home. Minor problems like this are why I have left a fair amount of contingency time. Around 13 weeks for negotiated 1 Around 17 weeks, closer to 15 for negotiated 2 Asset list:
Roughly 4 weeks to model everything listed if only including lesson time, around 2 if including lesson times and at least 4 hours of work each day for the rest of the week. Potential assets if time allows:
UV unwrapping the models will be the time consuming part. It is difficult to say how long I think each part will take but I'll make a guess to get a rough time frame.
Texture list
If I am able to use downloaded textures and the software 'Substance Painter' then the texturing of these objects should not cause much of a problem and only take 2 weeks maximum. If I am required to make all of the textures myself then I may have to sacrifice a lot of the variation and quality for the sake of not spending too much of my time on the task. Potential textures if time allows:
It could take some time to rig the character if I need to do it without any software help, however I believe I can use Maya 2016 to auto-rig a character and there are other website alternatives such as Mixamo. If I can use these then I estimate maybe 2-3 hours to rig my character. If I cannot my guess would be around 15 hours.
 I finished a concept painting of the environment for my game idea. I started with the one building (the one with the black shop-front in the middle) and then duplicated it a few times in either direction with some spacing. What it proves is that if I were to do this with my 3D buildings then I could change the location of details such as the drainpipe or the shop fronts and play around with the colours of each building to make it look believable as a different building altogether. This would cut down on production time a lot and might even allow for certain parts of the buildings to be generated randomly.
I side thought I had while putting this together is that if I were to make the game playable from this side view instead of over the shoulder there would be a much better emphasis on the character animation as well as make it easier to generate new map pieces off-screen. This would help minimise the impact and might even help fix the problem of it not running well on mobile. I'll have to look into it further and find out if it would be possible to have the world move around the character instead of the other way around. I haven't updated my blog in a few days now and wanted to put up a post of what I've been doing. I have started some progress on concept sketches for the game, however it is not my strong suit and it's difficult not to get distracted by the other parts of the project that I have more interest in as I enjoy the 3D work far more. Because of this I have also created a test animation using an old model I downloaded to see how well I could make a run cycle and then whether I could get it to look good while holding a mug of tea. I will put a link to the different stages of creating the animation below. I encountered a lot of problems while creating this animation which led to it taking a lot longer than it should have to create. In the time I was using it Maya crashed many times and then after a few saves the models right arm and left leg became black as if it were missing a texture. This mixed with the body movements not being very noticeable from behind mean that this model isn't very good for what I wanted to show and I will likely not be using it any more in the future. Thankfully I plan to create my own model for the actual game so I hopefully won't run into any of the same issues.  I've been doing some research for buildings that would work for my game. I found a few images of Victorian buildings that could be repeated with minor changes without the player noticing. This would allow me to randomise which buildings are chosen to make the setting look believable. My plan for creating these buildings is to model the individual parts, such as the windows and shop signs so that I can piece together the buildings in Unreal using the different parts and create different buildings without having to create each building from scratch. I believe this is called Modular Building. No major update today, I just wanted to store a few ideas I had about a them/story for my game up here so that I don't forget them.
Idea 1: Man wakes up, comes downstairs and starts making himself a cup of tea. While making it he notices the time and panics as he is late for work. He grabs his cup of tea and runs out the front door. This is where the game starts; with the player taking control of the man, avoiding obstacles so as not to spill his cup of tea. Pros:
Idea 2: The player controls a basketball player, collecting basketballs and avoiding both obstacles and other basketball players. Pros:
Over the past few days I've been encountering the different problems that are coming with my goal of making this game for mobile. The first problem is that the game runs very poorly. I've determined that this is likely because it is constantly destroying and creating new actors as the game runs. A possible solution to this that I found is to set up an 'Object Pool' that the engine moves actors around from without having to load them more than once, at the start of the map. The only problem I'm facing here is that I have no idea how to create one and the only help I've managed to find is an example setup that I can't pick apart easily to understand, or even get to work properly.
Problem number 2 is that the lighting looks terrible. When setting up a game for mobile the best way to get the lighting to look nice is to have a small area set up in a 'Lightmass Importance Volume'. The problem with this is that with the setup I've made for my infinite runner the character quickly leaves this zone and things start looking terrible. Stretching the zone of the Ligthmass results in worse quality light overall and the zone would need to be infinite to account for the characters available movement space, which obviously would not work. A solution I thought of for this is to attempt to have the map move instead of the character; keeping the same illusion, but allowing for the lightmass to stay stationary. Again my only problem here is that I don't know how to go about coding this and have had no luck so far finding any help doing so. What's becoming clear is that I've reached the end of Prototype number 1 and will need to start again with a better understanding, however I'm not certain where or how to start yet and I'm not making much progress. I'll update with any results in the future. I've spent the last couple of days attempting to convert the game prototype into Mobile format and set up mobile controls. Sadly I was unable to find much help with how to set up mobile controls, especially tilt controls, and spent a lot of time attempting brute force trial and error; learning what each thing did as I went along. I was able to find a few images of setups that worked for people and used these to help piece mine together: https://answers.unrealengine.com/questions/390439/how-can-i-use-tilt-sensor-in-ios-ipad-to-steering.html https://answers.unrealengine.com/questions/35512/how-to-use-tilt.html In the end I was finally able to create a working prototype that allows me to control the characters left and right movement by tilting the phone left and right, with swipe left and swipe right turning corners, and finally swipe up as jump. I requires a large chain in the blueprint editor and was not easy or fun to set up. Image of Character Blueprint:
 You'll see in the image that there are a few unlabelled event chains, from where I attempted different ways of reading the tilt input, that I wasn't happy with and deactivated. Another thing that I managed to get set up is the ability to test a map I've made on my phone. This means that each time I make a change to the control scheme I can check what effect it has on the game directly on my phone, using the tilt inputs and swipe functions. My only concern now is that when I set up the project I didn't originally select mobile, but rather switched to it later and I think this has led to my lighting being broken. I have spent a lot of time looking for some sort of fix for this, however I think my solution will have to be to start the project again and copy over the function chains from the original. This won't necessarily be difficult but it will, however, be quite time consuming. |
AuthorThis is where I will be posting all of my work updates. Archives
October 2022
Categories
All
|
