Prototype Project

BA2a Indie Development: Prototyping

For this unit I will be making a 5 minute playable prototype, focusing on game play and mechanics, with the theme of inference implemented within.

To begin brainstorming we randomly chose game titles and used random word generators, and then had to base a game concept on what we were given. Within 10 minutes I had a game concept, and I continued this for couple of hours, producing multiple game ideas varying in mechanics and styles. This helped me clear my minds and focus less on narrative, as my initial ideas focused heavily on story. I have narrowed my ideas down to two and began thinking up the game play and mechanics that would be involved. My preferred idea, when analysed and thought about further, did not suit the brief and relied heavily on animations as well as mechanics. The idea I have chosen to pursue has lots of opportunities for me to implement simple but effective functions and create a good feel to the game play.

10 Games Ideas

Spirited Away
Red Sheet
Island Routine
Swell Torture
Royal Thief
Dreamember

As iterated in this video: https://vimeo.com/139094998 (Everything is a remix) I am using other influential games to help create my prototype. I'm analyising the aspects I like to why I like them.

A game that has heavily influenced my mechanics is the PlayStation 1 game Pink Panther: Pinkadelic Pursuit. I enjoy this game as it requires memory whilst you explore, has fun themed levels and a tense game play when you're being chased. Similarly to this game, I shall use a health bar or some form of tracker on the health of the character, and have them collect something equivalent to coins.

Little Nightmares uses lighting well to create a gloomy mood, and blurs aspects of the background to create a dream like atmosphere. The camera sways and the levels themselves look like they're tilting, creating an uneasy environment and furthers the dream-like atmosphere. There is sound, but very little, which makes the player more attentive to the environment. You can hear the patting of the players feet and breathing of the man, making are more tense environment. The screech he makes if you get caught is disturbing. The level design and asset placement also leads the character through the level without telling you what to do, but letting you figure it out. In one area the lift is places near where you enter, and you can hear someone trying to break in. The panic causes you to want to find somewhere to hide or escape quickly, so you use the lift which has the next clue to the puzzle. Although I am not going for a horror or thriller style game, by analysing the mechanics I can understand why the game play is interesting and keep this in mind when further down the line creating my game.

Level Planning

Keeping this 2D side scrolling idea, I have began planning out level ideas and exploring the use of visual cues and directing the player to help make an interesting and functioning level for the player. Using Photoshop I made a rough level design, showing the path the player must take to complete the level. This helps me visualise where the player will go an trying to make them move around the level a lot and rely on memory of where they saw doors/passages for later. Iterating this further I boxed out random shapes and tried planning levels within them.

Brainstorming a rough narrative is the starting point before I can visualise my levels and character. The initial idea was stealing a crown, however this wouldn't be great for reoccurring levels just having the same initiative. I settled on the idea of the main playable character crash landing on a planet and his space ship rips apart as it soars through the atmosphere. The aim of each level is to gather a part of your ship, and by the end of the game once all levels are complete, you shall have all the pieces to your ship so you can rebuild and leave the planet. The collected items through the level will be nuts and bolts, which help put the big parts collected at the end of the level together. With this ideas, I have explored the route he shall take to help visualise level terrain.

I have used a map of earth as I can then base the terrain similar to the continent they are travelling through. This map could be shown to the player as they move from level to level or hidden. Hiding this means the player doesn't know what planet they're on, so they must use the clues to decide whether it's earth (future or present), or a different planet.

I read this article (http://games.ucla.edu/resource/visually-directing-the-player/) which has been useful whilst white boxing my level out in UE4. I need to consider visual cues such as lighting and using colour for certain objects and areas.

Using one of the level layouts, I planned a route and began mapping it out in UE4 to see what works. The layout soon altered as I added moving platforms, so I took screenshots frequently and re-planned platforms and routes.I used a tutorial for the moving platforms (https://www.youtube.com/watch?v=THCR9p5mIz0). Initially I put trigger points on the platforms so they only moved when approached. This method turned out hard for the maze style where I would need to be often using the platforms back and forth. I took out the trigger points and made the platforms move up and down (or left to right) on loop throughout play.

After expanding, playing through and getting critique, I have changed the starting point of my level. Originally placed in the middle and spiral left out into the maze, it was suggested they have more of a 'tutorial' section before getting stuck in without knowing the functions. Play naturally goes from left to right, so I have draw a section at the top where the player learns the functions of jump, hide and attack in that order, so they are equipped to complete the level. The two yellow stars at the bottom are where I plan to have 'keys', so both must be collected before they can open the door near the start of the level, which opens the next section. Placing the door where the player will see as they enter the level gives them incentive and they know something must be found to get through.

So far however, my level doesn't have a terrain as I have not decide which environment to use. This could cause a problem and I will likely have to alter my maze when I start planning the environment as it needs to work around the building/asset placement.

I am blocking out a town level, so window ledges and roofs can be used as majority of the platforms. I am completely redesigning the maze to suit the environment and using my prototype white box as an influence when planning the routes later. My building references are from a German town.

Blueprinting

In the workshop sessions, we created the blueprints for a character to move and the camera to move in first person mode. This will be useful once I have sculpted my character and implement them into UE4. Below is a screenshot of the basic blueprints we covered.

Below is the environment I made to test out the mechanics and the character was just a block. I successfully got the character moving around the platform and through the door, and had control of the camera angle throughout.

Mechanics I planned to implement in my game are:

• Jumping (possible jump height dependant on the length of pressed button)
• Hiding in certain areas
• Attack (one simple move to destroy any enemies encountered)
• AI for the small enemies to pace on a platform
• AI to 'chase' the player when certain trigger points are met (taking the key)

Character Designs

I have chosen 3rd person perspective so I need a basic character that fits the narrative. They will be a non-human species but need to be relate-able to the player. Using my mood board I have done some silhouettes and quick sculpting in Zbrush and then drawn over in Photoshop.

Week 3

Blueprints

To firm up my knowledge, I re created the blueprints again for making a character move and jump, and use the camera. We then learn how to put health and stamina bars on the screen, and get them to function with damage and collections. I will use a health meter of some kind, and will have nuts and bolt about for collection, in which these will be counted in the corner of the screen also.

Zbrush

I have been keeping the resolution and division low, and working on shape and form for character designs. I feel my first design isn't very approachable or relate-able to the player. Although this is an alien, the player needs to connect on some level to them.

I used an image of Ratchet (Ratchet and Clank) as reference to further my observational modelling skills and possibly give me more character ideas. I struggle with the side perspective to give the face real structure. I will analyse images of side profile and facial muscles to improve my next models.

This week we learn about tileable textures in zbrush. We started by preparing  two planes in Maya and importing into Zbrush. After fiddling with a few settings we could put bricks on the lower plane and mirror them in the opposing edges so it was tileable. Once complete, the larger plane on top can be brought back and used to fill in the gaps (concrete). This was hard to follow but I shall revisit as I may do this for the pathing on my level.

Level Design

When white boxing, I encounter a camera based issue, in that when my character went to far down in the map the play just stopped. Altering the Z axis in the world space view-able solved this, however I should've started higher up and then worked my level down.

With the terrain in mind, I've used the references to create a basic plan of how the houses are aligned. I've considered the feedback from last week of making the play go from left to right and the start has less platforms, so the player can adjust to the level slowly and be introduced to the mechanics. Through further crit when looking into mazes and how to make the game appealing and keep it interesting, it was suggested to have a twist to the maze. Suggested ideas included flipping the maze once you hit a certain point in the level, or turning platforms other ways so blocking of routes that were previously accessible. I like these ideas as it keeps the player on their toes an they don't get bored of the map I quite like the idea however of having a mirror of the map in the second section. The visuals can be different and some areas different. This can confuse the player as there will be some familiarity but something isn't right.

Earlier on in the week we looked into types of mazes, specifically unicursal and multicursal. My maze is linear in the sense that there is one main aim and only one start and finish to the level, and one path to follow to complete the game. However within my level, there isn't one specific route to follow, so it has multicursal aspects within the level design.

I have grey boxed this idea using basic blocks in Maya, combining all the basic buildings together and importing them into a 2D side scroller project.

Week 4

 Placing the windows (blocks) in may then helped me plan where ledges could fit into the environment. The initial structure is simpler and easier so the player can adjust to the controls. I then planned out the larger area by drawing paths over a screen shot in Photoshop. I imported just the platforms to UE4 to test they were all an achievable height for the character to reach. This meant I could adjust an platforms before I started adding more detail to the backdrop and merging the platforms int the buildings.

To add more depth to the level, I brought some building forward and removed the front. 

 After putting platforms on a building or two, I imported a test into UE4 and used the standard character to test the platforms where on his path and reachable.

 I modelled some low poly nuts and bolts and will begin adding blueprints so they can be collected and tallied through the level.

Tutorials used: https://www.youtube.com/watch?v=7BSc-1dk4tY&list=PL4JNxxaAxlM0-45TF9nSRZi7xBGvNCLw3&index=2

Widget Tutorial - helped with bolt and nuts binding to the UI
https://www.youtube.com/watch?v=6-NgA1zOESU

Week 5
This week I focused on completing the backdrop and test playing after each building.
The large maze area is basically complete in this image. Whilst playing myself and others noticed you were almost 'jumping blind' as the game went on, as you were unable to see where the platforms where. I altered the camera distance from the player, giving a bigger view of the environment. Using an earlier idea of having a duplicated maze area but slightly different, I have created and underground tunnel. The player must find both keys (which will be rocks) to weight down the platform, break it and drop into the second half of the level. The second half will be the maze mirrored, so the player begins at the top of the large maze. I plan to have a different atmosphere, possibly night time, with some slightly different assets and eerier music.

Following YouTube tutorials and screenshots from workshops, I added the health bar and nuts and bolts progress bar to the players screen. I struggled getting all three bars on the screen, and linking them to the appropriate blueprints to correlate the collection to the bar. I settled using trigger boxes and casting the blueprint inside the nut and bolt to the player character blueprint. Although going through an extremely similar process in workshops, I struggle to get my head around the slight alterations that need to be made to tailor to my project. I watched health bar and damage tutorials so I could make collectables with a progress bar. As this takes me longer than it should, I may not have time for all mechanics I had planned. I still want to add pacing AI drones, which was covered in workshops I just need to apply to my game. I also need the two rocks to be collected and the underground platform to give way/disappear once these are collected. I believe AI that chases the character will be too time consuming so may leave this mechanic out, which them makes my planned hiding function obsolete.


During ZBrush sessions, we experimented using the symmetry tool to make trophies, trinkets etc. My design had no reference so veered of to look more like a cake but I am now more confident using masks, transforming tools, symmetry and zspheres.

Sell Sheet drafting
This week I finally settled on a working title. 'Planet Zero' is the current name, linking to the narrative that you have landed on an unknown planet that seems to be abandoned or un-populated. I initially liked the idea of 'Character Name: Planet Zero', however I still prefer the ability for the player to name their character at the start of play. This idea followed suit of games such as Spyro and Crash Bandicoot, which are aimed at the same audience as my prototype. This rough sketch shows how I plan to layout my sell sheet, the first page flows however the second page still feels muddled and will come together with trial and error.

Going back to my first influence, Pink Panther, I like the idea of the games motives being explained in the loading screen with these basic but quirky diagrams. I will experiment with this diagram style in my sell sheet to visually show the reader how to play through a level. My game is mainly visuals, as text within game will indicate the planet type and I want this to be unknown and pondered upon.

Tutorial used for Camera Sway
https://wiki.unrealengine.com/How_to_Add_Camera_Swag_/_Lag_(Side_Scroller)


Tutorial for Checkpoints (next blueprint)
https://wiki.unrealengine.com/How_to_Setup_Respawns_and_Checkpoints

Two screenshots of the background mesh in UE4. I ran through with my character testing all the platforms. With no apparent issues, I shall chip into some of them to add a little more detail and begin unwrapping so I can texture the houses.

 Here shows the 'underground' section of the game. Using suggestions for earlier crit, I shall have a mirrored city as the second half to my game. Once you have the keys to unlock the tunnel, you travel down into the new zone and work your way back to the start effectively, where a piece of your spaceship lies.

My Blueprint within the Key

Using this screenshot tutorial from George, I was able to implement a key and door system. This took surprisingly less time than I thought, allowing more time to get the next aspect sorted which is linking the blueprints to the UI so the player can see they have a key by there being a symbol in the bottom corner of the screen.

Using these blueprints I have for the underground door key activation, I created one final door for the second part of my level.This is the final shot of the ending. Another red door was placed like before, and needs the 3rd key to activate it. The player can walk into the room and find the spaceship part and complete the level.


I have also began modelling some assets to put in the level, none for interaction but to help the game feel and make the town less empty, but still needing to look desolate.

Enemies
I began placing NPC's in the town this week. In hope to get the mechanics down before making a character', I used a sphere for the base mesh and began blueprinting. Following a workshop session on AI and guidance from a friend, I added a Nav Mesh box to the area they would pace and implemented pacing blueprints.
This is the blueprint within the NPC. On begin play, i made them start moving, using the custom event of Pacing. With pacing I references the patrol points, sending the NPC from one, to the other and so on, looped. I also added damage when the player comes in contact with the NPC, using a collision box. I had a few issues getting the collision to work so I used the collision capsule that was already there, and an additional box, and linked them both just to ensure the player hit one of them.

This didn't work first time as I had the checkpoint variables wrong. Through analyzing the blueprints in the NPC and the patrol points step by step, we eventually found the problem in the patrol point and attempted to change it back to a checkpoint reference. We couldn't do this, so we made a new checkpoint and made sure not to change it to anything else accidentally as we couldn't undo it. On simulation, the blueprint was now functional and the NPC went from patrol point 1, to 2, and then back to one and so on.
I purposefully walked into them after making the blueprint and soon noticed the health bar drain more than I had set. If the player got stuck next to them for a few seconds, as soon as you walked away your health would drain and kill the player. I knew I needed to now put a delay in, so the health takes the set amount and gives the player a chance to move away.

I originally planned to have chasing AI in my prototype, and I have looked at the blueprints to do this, however, all my other mechanics and blueprints have taken longer than planned and my estimated time fore this would get me very behind. Alongside this mechanic was the hiding aspect, which becomes obsolete without anything to hide from, so I shall be leaving that out also.  

In scene shot of the enemy. I added particle effects within the blueprint, so electric blue sparks came out of them, and a little smoke from the top. This indicates they are damaged and that they can cause damage, as without these effects they don't look like enemies. The effects are hard to see in the lighting of the top area, however are more effective on the underground area.

Checkpoints

It took over a week to get a functioning blueprint for my checkpoint. On my 4th attempt I got it working, and these are the blueprints within the checkpoint itself, and the character.

What I believe was my trouble was similar to my patrol points
Screenshot of a checkpoint at the beginning when testing whether it was working. The object has no collision and is hidden during game play.

Week 6 & 7

Zbrush

This week we revisited the Chalice and looked into more ways of modelling the same things.

To increase my familiarity with ZSpheres I attempted to model a body. I used Joy from Inside Out as reference as her pose is distinct. I am happy with the placement of the points and feel this is a good place to begin sculpting on.

 Blueprinting for Start screen and checkpoints

Diagram to give player direction. Activates when the start button is pressed, displays for a few seconds and then game play begins.

This tutorial helped my set up my widgets for the nuts and bolt count earlier on during the prototype. It also came in handy later as I was able to use the toggle visibility which I didn't need before, on a different UI element. Now on a button press the level map appears on the players screen, and disappears on release. I have also added a short animation so the map fades onto the screen
https://www.youtube.com/watch?v=6-NgA1zOESU


Extra Credits - Games for Kids
https://www.youtube.com/watch?v=NdFw8kvHAY8
This video helped me understand that a game for 'kids' can be very broad, and I need to define which age group it will be aimed at. I intended for this game to be for ages of 7 and above, and as the video states, it is important to understand the abilities of that age group and how socially aware they may be. With that in mind I need to make the game challenging enough for children aged 7 and above, but not too hard that they don't understand or give up. Similarly the narrative cannot be too complex but compelling enough to keep them interested.
Health Pickups

As there will be a health bar and enemies, a health regeneration item is required. This allows the player to make mistake but survive longer whilst getting to grips with how the enemies work and the amount of damage taken. Each health pick up regenerates the equivalent of damage an NPC would take (20% of maximum health). 

I have chosen for the heart to be gold, giving it a shine and making it enticing to the player. One i put the material that was already in UE4 for gold onto it however it didn't stand out as much as intended. I added a pink glow around it to make it more visible, and as hearts are often associated with the colour pink or red. There will be a pick up noise to indicate it has been used as the actor is destroyed. 

Sounds
I explored Purple Planet (www.purple-planet.com) for background music and shall convert a couple files into WAV files and  place in game and get feedback on whether they suit the game style and feel. I like the idea of having
I then used freeSFX for the sound effects. I chose 2 xylophone clips for the nut and bolts as a satisfying reward for collecting them.

Albedo Mapping Town and Windows
I unwrapped my town, platforms and windows, and then painted the albedo's using Photoshop. I tried substance initially, however the UV unwrap wasn't done well enough and on such a large scale it didn't paint well. As I focused more on game play and just wanted the assets there, I automatically unwrapped which had it's downfalls. In future I would spend more time manually unwrapping, softening and hardening appropriate edges and unwrap each building individually. With the whole top town being one mesh, the UV's where very small and would show very horrible textures if anything more than a solid colour was applied. I was going for a sell shades, block colour look however so this was fine. I could have also have modeled one window, and just stretched it out and shrunk it for each house to get the variety, rather than having all different windows and then having separate unwraps. With all the meshes more separated, pivot points would be more central and easier to move around the scene.

Lighting
To get the underground section dark, I have enclosed it between multiple walls, and added a dim directional light. Through my own play testing I soon noticed this was too dark to navigate and had to add additional lighting. To fit in and make the area visually pleasing I added the extra lighting to the windows.

Controls began simple and illustrated for the start screen and sell sheet. I then used a controller whilst play testing and enjoyed the feel more, so decided it would be good to have both options, suiting PC and console players. I had chosen my font by this point and could alter the illustration. 

Start Screen

Sound effects and music



Play testing

Due to blueprinting taking longer than expected I have play tested later than planned. For future, I need to have a few working mechanics early on and get play-testing as soon as possible, and edit according to feedback as I go along.
Feedback I have initially received was that the jump was hard to control and to jolting. I altered the blueprints so the jump is now controlled and you're able to jump on the downfall. Especially with platforms and constant jumping, a smooth movement is crucial. I have also got the game functioning on controller as well as keyboard, as it's slightly more suited to my audience and nicer to play with a controller.
I changed my blueprint to this.

It was pointed out to my that my game did not have a map and that due to the mass amount of space covered, I should have one to help the player locate themselves and collectables. I initially wasn't going to have a map, so the play must explore themselves but this feedback was useful and relevant so I began designing a map for the player to access during play.

Using a screenshot of the level in Maya, I painted over in block colour, so the silhouette was an exact match to the level. Ideally the map would have a tracker so the player would be a pulsing dot, however this is too complicated for the little time I have and would be a function in the overall game play, not the prototype. I have given them the location of the first 2 keys, but left the underground second zone blank. This was intended as the second part of the level is supposed to increase in difficulty and the player will also know more about how to complete the level by now. 
Within my character blueprint I added the widget and a function to toggle it during play. The transaction was quite abrupt, so I created a fade in like the start screen and diagram have.

Feedback table
I devised a table with the small amount of feedback I got from my late play testing. I had feedback from non-players also, so noted these underneath as points to look into.

Sell Sheet

Tutor feedback
During this play test, there was a lot of lag, due to the mass amount of lighting in the underground area (around 1000 point lights). I was shown a more efficient way of creating light that will dramatically decrease the lag, which was to use emissive panels. I created a blueprint with a cube mesh inside, attached an emissive material to it and added a point light in the blueprint also. Once sized up in the blueprint to fit the window, I could place them in all the windows of that building, replacing the multiple lights originally there. I could then make material instances of the emissive material, saving more memory and just duplicate the blueprint and alter the size of the mesh to fit another window. Although time consuming, this is definitely the most efficient way of adding light through the windows and still getting the aesthetic intended. I can alter the light intensity and colour at ease as changing something in the blueprint will update all of them in my project.

I found placing the lighting and using it effectively challenging. Lighting is an important part of atmosphere, it can help lead the player through a level or point out items of importance and much more. In time I hope to learn how to use the different light source more effectively in UE4 and when to use emissive lighting to save memory.

Another issue raised was that on the largest building, both above ground and underground, some ledges could not be reached. With some nuts and bolts being placed there, the level would be impossible to complete fully. I have added cube meshes near the platforms to aid the jump up to them, so they are now accessible. In future, having the platforms separate would make this a lot easier, as I could simply just move it down. More planning whilst building the level and time spent test playing each aspect would ensure it doesn't get this far through play testing next time.