Recently we were asked to test a game called WaterCooler, which is currently in the alpha phase of development. The premise of the game is that you are a boss at a game studio, with the responsibility of choosing what game to develop and allocate to a selected team member. You also must manage the teams moral, which can fluctuate depending on the activates they are doing. So, if they have not worked in a while they may get bored and aggravated, which is where the watercooler comes in. You can go up to them and suggest going over to the watercooler to talk. Here you are prompted to pick one of a few dialogue options in a response to what they said. Depending on how well you talked to them, there moral can boost or lower which gives some consequence to your actions. However, the more I played the game the more I realised that each time I would talk to somebody they would often repeat the same things they said last time. Vice versa, I also normally only had the same things to say back, so repetition was a problem very early in the game.
The game also had a cartoony, colourful and bright look which complimented its quirky concept. It had a very “mobile game” look, which is one of its possible target platforms. Progression also came in the form of each game being developed and completed. This lead to more people being hired and more management involved in talking to them, and going over to the watercooler. As more recourses became prevalent, the more you could improve the look of the place with the player being able to place plants within the office. My play time was limited however, playing only half an hour at max. I did see that there were more offices to unlock as the game progressed, giving the player a sense of progression and progress. It also breaks up slightly, the relativeness of seeing the same environment.There are also points where it will reward you with an achievement like moment, when you achieve certain criteria.This one was for boosting the mood of a co-worker by talking at the WaterCooler. You also get an experience reward for this, along with a bonus if you meet the extra criteria which I missed. Moments like this will keep the player interested and motivated, and make them feel like the extra effort they put in is worth it.I also filled in a questionnaire before and after playing with many questions about enjoying the game, to bugs and what types of games I play. It was clear to see there were many deep analytics so the developers can carefully test their games, whilst also seeing what new features need adding. Whilst playing the game for half an hour it did become apparent that I was not the target demographic for this game, as it did not have the depth needed to keep me interest for a long period of time. I can see this being aimed at a younger mobile orientated audience who are interested in management like games. I did come across one bug whilst playing, which is when using the mouse scroll wheel I was able to zoom in indefinitely with no limit. It is also important to remember that the game is in alpha state, and could improve and iterate on its current features.
I recently came across a speed level design by a user named Maverick on Youtube, who has time lapsed himself creating a street in Unreal called “Overgrown Ruins”.
He starts of with a simple street but it quickly progresses into a lush environment.
He also vertex paints the textures onto the road. Buildings that are added also have rubble attached to them, instead of being added separately. The terrain tool was also used to block of the sections of the streets. The texture also changes realistically from a ground/mud texture to a rocky texture on the raised sides. By using unreals terrain methods you can create an area such as a street, and segment it off making the area manageable and believable. Something I need to consider doing when creating my project for my final year.
Rubble patches being added in modular sections. Moss being added to tram line increasing the detail within the scene.
Different variations of mud was also added to the environment. There is also wildlife in the city street drinking from a puddle, along with a radiation warning sign in the background which adds a little context from the scene. There is also a narrative being told through the vegetation within the scene along with the wildlife, giving the viewer/player some idea of the time frame since the destruction of the environment.
Birds on the tram.Lighting was added making a huge different to the look of the scene. This is something I need to work on, and consider a lot more often when working in Unreal. As it really can have a drastic change on the scene. Here is the incredible final look.
Seeing something like this is certainly inspirational, and has made me starting thinking more about what I want to do for my final project, along with what techniques I can use to acheive this. Vertex Painting is one technique I have just recently learned, but even then I still have a lot to learn in that regard. Such as in this video he vertex paints grass which I do not know how to do yet. Everything in this video was already made, so it is important to remember that it would take a lot longer than this to make with assets, textures, wildlife+animations etc.
One skill I needed learn in order to make my vault look a bit more realistic, was a way to make the textures look worn, dirty and seamless. Part of this is to include dirt and grunge within the textures, so when you apply it to the object/environment it already has a specific look you wanted. This I have already started doing and improving by learning more of substance designer 5, where I can blend gradients with grunge maps. However, sometimes it is needed to add a little more detail onto the texture, whether that be a puddle on a floor, or dirt and dust on the lower part of the wall. Vertex painting allows this by plugging your maps into a lerp and multiply, along with a vertex colour node.
Ceiling material that I have edited to be used for painting.
I can also make some textures look a bit more seamless, from two different ceiling textures connecting I can add a little dirt in between covering the line that breaks the two. One issue I have encountered is the way I have modeled the environment. Due to me not considering that I was going to use vertex painting, my models do not have enough verts for me to efficiently paint, or even paint onto the bits that I want to. This means that I will need to go back into max, add more verts to the model and re import which is of course time consuming.
This is something to think about for future projects when I know that I am going to use vertex painting, which means I can save time and make sure there are enough verts in the model for it to be effective. This would of course be the most efficient way of working. So now I am in the process of adding more verts to walls and models in order to take advantage of painting. This will be worth the lost time however, especially for major pieces like the floor which is something that will be seen many times, so adding a bit more detail onto it won’t go unnoticed. Here I added some dirt to the ceiling.
I have also started to vertex paint for our heritage project onto one of my buildings, adding moss in a couple of areas adding a bit more realism. This will now also allow me to add details onto the paths and tram lines, with materials I could create such as dust and dirt.
Since the last update I have started to experiment with the use of darkness. Due to it of course being underground, lighting is artificial with some light sources down there being more adequate than others. So in order to translate this across in Unreal I have been using the post process, mainly the sub setting called colour grading which allows me to tweak the contrast, brightness etc.
Here you can see the post process settings in the bottom right hand corner.
Here are two examples of how different the enviorment can look with a couple of changes to the post process. You can see with this shot that it looks a little more atmospheric and darker, possibly adding a little more than what the other one did.
I also created a dust material which I then put into a particle system. I can then use this is make the environment look a little more dynamic and believable.
I also removed the middle roof pattern as it looked a little out of place.
I have also been filling empty spaces and with papers and boxes, but I need more assets creating as these quickly become repetitive. Though I have used the darkness to mask a couple of these details which make it less noticeable the viewer.
I have asked a couple of people for options and many have mentioned the lack of dirt on the floor, roof along with others assets so this is something that needs addressing, and I do completely agree. To acheive this I am going to create a couple variations of dirt textures, then apply this in Unreal through vertex painting. I also need to model a few beams, pipes and lighting for the environment.
For my Vault environment, I have decided to import it into the Unreal engine in order to test how it will look in there. I was originally going to render out a video on 3ds max with a camera sequence moving through the vault, but but due to the amount of polys and my computers performance, it makes it difficult to use dynamic lighting on the high quality setting to viably continue. However, on Unreal this is not an issue so I should be able to place lighting within the environment, with instant, dynamic lighting results. This does not rule out rendering in 3ds however, as this is just to test whether Unreal would be suitable. Also when I transfer the max project to a separate computer the textures reapply themselves to the wrong objects, which does not happen on Unreal. I have started to import some of the scene and object, along with textures which is somewhat time consuming. I can also clearly see what textures need improving and changing. I have found that my roughness maps need changing and my brick textures needs improving dramatically. One thing I could make more efficient is how I transfer textures. After I have created a texture on Sub designer 5 I could export as a sbsar file, which can be immediately imported as a material in Unreal, instead of individually connecting each map up. However, this was not planned for due to me not anticipating using unreal.
I am going to continue adding textures and objects and update with another blog on the progress.
In order to continue filling out the space for our heritage project, I decided to model the Monument building which was distinctly missing from our square. I started by getting a wealth of reference photos, some of which I’ll link here –
I broke the building down into sections in order to understand how I could utilizing the modularity of the building, whilst also making the scale seem less daunting. I have also been trying to push for creating my own textures using substance designer, which I have been trying to use some more advance techniques with.
I created a good few variations of the brick, using two separate gradients going into a blend – one for background colour, and the other for a foreground grunge colour. Then masked with a grunge map allowing for a variation in the way it appears on the brick.
I had a good few variations and toned it down from the top, until I reached one that looked a little subtle in the worn down look. One thing missing is moss and other bits of dirt that it desperately needs to make it look more realistic. I have been looking into vertex painting on 3ds, but does not seem to intuitive. For example, I cannot see what I have painted until it is rendered. This could be a miss step on my part however. I am going to import the building into unreal now and learn how to vertex paint on there, which should hopefully turn out better.
Due to the building having a lot of intricate details, I decided to use mudbox in order to sculpt these out. But has turned out rather difficult to get the shapes I want. Still, this is one of the few times I have used it, with only really using it for creative futures for practice. I do aim to continue using this software, as it is one of the industry standards for sculpting and is widely used among 3d modellers.
Today I developed a pause for my game whilst I was refining everything. Whilst not absolutely necessary for a game with such a short play time, I figured it would still be good to have and wouldn’t hurt to implement it.
First I created a simple pause widget with the text simply saying “Pause”.
I then added a simple pause button which still needs to be refined in terms of aesthetics, but serves its purpose for now.
I then put if button clicked – go to branch, which then triggers a set visibility of the pause widget; thus pausing or un-pausing the game depending on the state.
For the highscore screen I tried to watch a couple of tutorials in order to find out how to code this into the game. I found one tutorial on how to overlay a scoreboard over the top of a game, which would not work in my game but I gave it a shot anyway in hoping to find that some of the code could be used for the highscores.
I used a load game from slot, which would allow me to retrieve the save game slot which had the high score on.
The tutorial also suggested using a separate user widget, which would control the text and numbers for the scoreboard widget. Essentially controlling the formatting of the scoreboard. In the end however, I could not get the highscore screen to work, but I did start to see how the save and load state worked, so it wasn’t all for nothing.
Here is my old highscore screen which I changed to this – This looks slightly cleaner to how it previously looked, but there is also a fair amount of wasted space, which could either look minimalist or to empty. I think putting a background image in there would fill out the rest of the space, to add a little extra to the look.
Shane also helped me try to set up the highscore code but we could not get the game to load the save game state, with it overwriting the previous score rather than keeping the highest score saved. Shane did help in relaying the points system to appear on the highscore screen by creating a binding of the number.
This gets the information from the points system within the ship blueprint, which connects up to cast to ship instance. My next step is to carry on working towards getting the highscore screen functional, with the player names to be made editable, along with trying to figure out how to get the highest score to save, without being overwritten. I would preferably want to have the top ten highest scores saved but I still don’t understand code enough to acheive this. Still, this project has made me become more comfortable with unreal, and I am picking up simple code here and there, but I’m still not progressing as fast as I would like and road blocks still prop up regularly with what I want to acheive with this game. So I may start a separate project not related to this game in order to learn more about code, without being confined to specific bounds such a ship selector screen, with very specific camera movements and player interactions.
I decided to start another project to further advance my 3d knowledge and skill set. After looking at other peoples portfolios online and on poly count, it motivated me to create something different to what I would be usually be making. One of my recent blogs I looked into how some artist create a small scene which look great for portfolios, so I new I wanted to create a scene of some kind with that in mind. In the end I decided I wanted to create a floating island in the sky which featured, well, I didn’t know to be exact but it just came to be as I continued modelling. I was originally going to have it as some kind of oasis, with a waterfall and a river flowing through, along with trees and a cabin. But I somehow ended up with a wild west setting, with a saloon and a train line cutting through the middle. A slight deviation from the original idea to say the least. I accumulated a bunch of images on Pinterest which can be found here – https://uk.pinterest.com/sam137069/island/
The project is not finished yet and I’m unsure whether I will have time to texture it, due to managing time with the other modules. I also don’t know what style to go for, I’m torn on whether to go a more cartoony style like a Blizzard game, or whether to go for a more realistic style.
Here are a couple of early renders of the progress of the project so far –
After modelling my Biplane the other day, I was going to leave it un-textured and just leave it as it. However, I decided against this as one of my weak points is texturing, so I decided to go ahead and texture it for more experience.
I did not particularly come into any difficulty unwrapping the plane, as I tried to build most of it modular making it easier to unwrap. I did however target wield seams within the unwrap, which I was unaware you could do until now. I also used Marmoset Toolbag for a couple of renders, but I had forgotten that you had to invert the roughness map due to how the software works. This explains why the plane did not look as glossy as I would have liked, during the first couple of screenshots.
Here are a couple of screenshots from Toolbag –
Number one I had a very low gloss value, but after increasing the value it allowed me to achieve a more realistic look on the biplane, as you can see by the second image.
Here are a few more renders from 3ds max –
I am fairly happy with how the project turned out, but I do think I could improve on the textures by quite a bit. For example, the lines on the wings don’t fully align due to way I painted them on, with Substance Painter. This is because the uv maps for the top and bottom where not fully aligned which caused this issue. There is also a lack of detail on the wear and tear side, with a lack of scratches and dirt. I added a little but are hardly noticeable to the finished renders. I do tend to rush the texture process a bit in order to get to rendering, which is something I need to slow down on and work more carefully over this process.