Evaluation.

Looking back at all of my work I feel that the past four months have been a successful time, where I have challenged my skills and developed them. Not only have my skills such as modelling and texturing developed, but I have picked up coding in java script relatively easily. Some parts were quite challenging, but I have been able to code the majority of the game by myself, because of that I am very proud of myself for that.

Originally when I set of on this project, I produced a specification of what this game must be. To evaluate this project directly, I feel I should test whether this product fits against my own specification, this will test whether or not I have achieved what I set out to do in the first place.

 

This game must have:

-This game will not have any sort of heads up display which shows on screen; instead all of this information will be shown inside the vehicle which the user is travelling in.

 

This game does not have any sort of heads up display. This was a design choice I made early in the games development because I wanted the user to be challenge when they need to see the information such as your points or score. It acts just like a dashboard in a car, which gives a realistic twist, like you would have when driving a car. Importantly though I have achieved this.

 

-It will use a camera element to take photos

 

The game does use a camera element, which acts like a first person shooting gun, but without actually firing anything.

 

-The subject of the photos are animals.

 

I have made a number of animals as subjects to take photos, and this is emphasised in the fact that you have build up a score by taking photos of these animals. The tick and cross system also establishes what to be taking a photo of.

 

-The design of these animals will be my own.

 

I designed, researched, and created all of the animals in this game and that is shown in the development of this.

 

-The setting for this piece is a jungle/river

 

I have created, and researched the Amazon rainforest and based the environment around it. The concept of layers in a rainforest was added to the design of the track and route that the car takes, this was done to subtly emphasise the layers of a rainforest. There are also two sections based in a river, there is a mix between the car being on track and in the river.

 

-There will be a vehicle designed to go on both land and in the river.

 

The car does not travel on land, but does on water and in the air on tracks. It is appropriately designed for this.

 

-You will use the gyroscope of the oculus rift as the camera.

 

This game takes full control of the oculus rift, and uses it as the aim for the camera.

 

-There will be a limit in the amount of photos you can take.

 

You are limited to 35 shots. That’s enough for the game to be challenging for more hardcore games, while being easy for first time, or younger players.

 

-The animals will animate

 

All animals have been hand animate, and tweened.

 

-There will be a scoring system.

 

There is a scoring system that is balanced and easy for the player to understand.

 

-Photos will be posted on to facebook or twitter while the user plays the ride

 

This element was not put into the game in the end as I ran out of time, however the photo you take is displayed on a screen inside the game.

 

-There will be elements in the game which emulate a rollercoaster experience

 

There are on rail elements but when designing the layout I chose to move away from a faster experience due to the fact that that would take away from the gameplay significantly, and it should be a relaxed sort of game rather than an intensive experience.

 

-There will be a balance between the gameplay and the rollercoaster side of it.

 

The on rail elements were balanced out so that the user did not get too disorientated throughout the experience.

 

-The game will have a beginning, middle and end.

 

In terms of any sort of narrative, this was written out of the game due to time constraints, however there is an opening menu of the game which takes you to the game or quits the application. The game will always take you back to the menu once it has finished it routine around the environment.

 

-It will last from around 2 minutes to 3 minutes.

 

The game actually last about 5 minutes. Mainly to get around the environment. You don’t exactly notice this though.

 

 

From this evaluation the majority of my specification was achieved, the only parts I missed out were minor, except for the timing being 2-3 minutes. This is quite a large issue which I forgot about throughout the development. There really is nothing I can do about this now, arguably I feel the timing to the finished product inst too much of an issue due to the fact that the game resets after it plays through the environment. So if the user does get bored, they can just put the game down and it will eventually revert back to the menu screen.

 

 

 

There are one or two improvements that could be made, but they’re improvements I have failed to make before the end of this project. I couldn’t make these improvements by the hand in because of the time restraints, and anything else that I would add to the project might cause it to be in an unfinished state by the time for my critique. Right at this point in time I feel that the game is in a finished and polished state.

The improvements that could be made would be:

 

Working on the texturing. Right now the textures on each of the animals really need a lot of improvements. Some of the animals need to be furry and they are not. Throughout the process I found methods for improving these textures but chose not to change any of the animals because they were in a state that, at the time I felt was fine. Again it is too late to change anything, but also if I were to spend more time on this, I would refine all of the models, but also reanimate them too.

 

Making the animations more refined and smooth would also be another improvement I made. I know that throughout animating each of the animals I did not touch the graph editor, and this was a big mistake. I know this was done due to my time restraints to get this project done, but it’s something I wish I could go back a redo, just to smooth out all the movement and refine them so they were absolutely perfect.

 

Adding more levels, but this would require going through the whole process again, and I think only one level to show is much more suitable and realistic in the time given.

 

The output for screenshots, I looked into this function within the coding section of this game, and it was possible, but was littered with problems throughout, and I needed to get the other function in the game to work. The problems I had was that it lagged the game whenever it took the photo, and it saved on top of each of the previous photos (in other words I could only save 1 photo per game) and getting those photos to automatically save out on to facebook, which would have taken the project even further in terms of technical skills was beyond me in the time given. I decided to avoid this function, but if I was able to go back and work on this, I would.

 

These were the main improvements that I could make to the game.

 

 

Sustainability of this product

 

This product can just about sustain itself due to the fact that it will be released on to the oculus site, and the idea of Oculus Snap should attract a lot of people due to the thought and idea of the concept in the first place (due to it being used in pokemon snap). When something is released on to the Oculus Share site, they seem to continuously get downloads, even the rollercoaster I made last year is still getting downloads.

 

What have I learnt?

 

Throughout this process despite not picking up any completely new skills, I did learn a lot.

Firstly I learn a good chunk of JavaScript for coding with, this was incredibly useful, and will be for future use in game creation. I am able to read the code and to some extent understand what it means, but also work out what it means if I don’t understand. This for me is an incredible breakthrough because coding the game was one of my main worries. I have leant about functions, the syntax of the language, timers, health, ammo, gravity, and just all of the basics for all of the JavaScript.

I have learnt more about texturing and the use of normal and specular maps. I have learnt the appropriate time for using each of them along along with how I can create low poly models and bring out their detail with normal maps. I feel this is a very useful skill to know and to use within my games.

I have worked on my modelling techniques, I have learnt extra function in Maya such as bevelling you corners and chamfering edges to give your assets certain styles, and make them render in certain ways. Since I learned these skills throughout, I wish I could go back, start again with the newly learn skills and create the whole project again. (That would lead to a cycle of learning even more skills and then using them and so on!)

That’s the majority of everything that I have learnt. Although it doesn’t seem all that much, it really is a huge amount specially i the coding side.

 

If I could do this game again what would I do?

 

 

If I could do this whole game again, I would pick a different location and apply all of the skills I had previously learnt.  The coding would be incredibly easy to replicate because I could just drag it out of that game and just dump it into another environment. It would then just be a matter of creating assets that fitted into each environment. In other words now that the game engine coding is covered, I can just keep creating environments again and again.

I think I would pick a cave like theme, or a desert, where the actual experience is a little more rough, and maybe the animals are a lot more animated than this previous environment. To some extent I would recreate one of the pokemon snap levels so see who close to the original I can get.

 

If I could do this project again what would I do?

 

If I could go back and start the whole project again from scratch, I would again create some sort of game, but possibly move away from the oculus, or create a game which uses the oculus but in some sort of way where you are taking a certain type of game and create it so it revolves around the Oculus Rift. I feel there is a huge potential with this product and with the idea of virtual reality, and I want to continue using it.

 

Where next?

 

Next I want to either create another oculus rift experience, or create a game and get it up on a  site/ system such as steam or apple, or getting something on the Wii U.  Only time will tell where I am going next.

 

I feel this project was defiantly my favourite, but also my most developed piece. I am happy with how this has turned out.

Showing it off.

So with everything done, I have held back a little with the photos mainly so I can show it all finished.

Here is all the photos of the game working, and a video that I took of it all working.

 

This is the splash video that plays at the beginning to mark that it is my work.

These are of the menu. Its quite an open area for a menu, it teaches how to play the game from the beginning so that users are ready to play the game straight off. No tutorials or anything!

 

All the next photos are of the environment and game that I have created.

And as you see at the end of the game you revert back to the menu screen ready for the next person, or for the same player to have another game.

The game will always reset upon opening in the gaming environment.

 

 

Here is the video.

 

 

One extra bit was sadly this got rejected from the oculus rift site once more, but there was some good news in the feedback.

 

I shall take another look at the in regards to the mac system version, but for now there is nothing I can do until after the crit. Luckily it will go up online as soon as possible though, I have been trying for about a month to get it up, it has still developed throughout though.

Coding 4 Cross and tick

The cross checker.

 

This was the most irritating and time consuming function, but importantly it allows for a checking system, which checks whether or not a photo has been taken.

 

Here it is

 

Static  variable crossPlay  = false is set as a Boolean.

Static  variable crossStop = false is set as a Boolean.

Static  variable creatureHit  = false is set as a Boolean.

 

Var anim is set as an animator

 

In the function update

Anim = GetComponent (Animator)

 

Then there is an if statement

If both the crossPlay and creatureHit change

 

Then play the animation for the cross

 

If the crossStop is true then playCross and creatureHit are false.

 

To simply put this, it is a checker, which checks whether the sphere that is created hits an objects. There are references to crossPlay and creatureHit (as they’re static variables) within other mention scripts which set the bools to true, therefore playing the animation of the cross.

 

That is the basis of all of the coding for all of the functions.

 

On top of that there is a small code which changes the scene back to the main menu after the whole trip around the rainforest.

Other than that, that is all of the coding.

Coding 3 Score system

 

 

The score system establishes however many points you get and a total that is shown on the dashboard.

 

There is a Scoremaster which receives all messages from each of the colliders. Whenever the sphere hits the collider (created from the source) a +X amount of points is added to the score master.

There is a static variable named ‘currentScore’  which is an integer.

There is also a variable named test which is also an integer.

 

By default the score and test are set to 0 in the function start.

In function update currentScore is test.  (this means that test = currentScore)

 

There is then a Getcomponent which refers to a textmesh (being the text mesh for the score) .text is then equal to text . To string(); (the to string part translates the integer to a string value, which is what the text mesh understands)

 

Coding 2 The receiving system

 

 

The receiving system is put on to each of the colliders which are parented to each of the animals. There is a script of each of these that does this:

 

otherObject 1 is set as a variable

otherObject 2 is set as a variable

tick is set as a variable

cross is set as a variable

 

In the function OnTriggerEnter  there is:

 

An if statement, which say if an object collides and is tagged as “bullet” then

 

The cross becomes false (this is a little complicated the cross part)

Creaturehit (which is a static variable which is a checker for the cross) becomes true.

 

otherObject1 plays the audio file attached to it. (this is a ding noise to initiate that you have score points)

otherObject2 plays the animation attached to text (score) on the dashboard which also initiates that you have score points.

Scoremaster. Current score gets an additional X+ points added on to the score (depending on the animal you get different points)

It then yields for two seconds

Ticks renderer becomes true, so it is viewable

Ticks animation plays to initiate that you have taken a photo.

Coding 1 Shooting.

With all of the coding finished, I am now going to explain exactly how all of the function properly work. There are gonig to be four sections to all of this, and some parts may refer to other later or earlier post, Since they all refer to different parts at the same time, it might be best to read these parts twice, since then you will understand everything.  I have also ordered these post to be in the most useful order to understand.

 

The shooting mechanic is the first function. This function just like a gun, so when you click the left mouse button, this happens.

Firstly a number of variables are set. These are done inside the editor. These variables are:

-The projectile that the object will shoot.

-The fire rate, or how many shots are fired at one time.

-An internal float value for the nextFire

-A variable for the speed.

-A variable for the current amount of shots

-A variable for the total amount of shots

-A variable for sound 1

-A variable for sound 2

-A variable for sound 3

-A variable for shutter 1

-A variable for shutter 2

-A variable for cam.

 

Originally this function was done through the update function, but was changed to Co-update because there was a yield statement I had to put into it. I will state where this is. This had to happen because the update function cannot have any sort of yield inside of it because of the way that the function works, as it is always updating and it cannot stop at any given time.

 

But this is what happens.

 

In the function Co-update, there is an if statement.

This if statement checks if the user has clicked the mouse button.

If this statement becomes true then this happens

1 will be deducted from the total bullets.

Sound 1 is played

Then a sphere (which has no mesh renderer) is instantiated into the scene from the source. The variables of how fast it travels in defined by the variable of the speed.

The script then yields (why its function co-update) for 0.1 of a second.

The cam.camera (which is for the camera screen on the dashboard) then turns off freezing the scene of the shot you just took.

Shutter 1 and Shutter 2 play their animations (which give a shutter effect on the camera)

Sound 3 is then played

The script then yields for 4 seconds

The camera is turned back to true.

A message is then sent to another script with a static variable (named current ammo) on it which tells the current ammo is the same as current bullets, deducting 1 from it.

Then there is an else statement, which acts like a checker.

This else statement checks if the ammo is at 0. To make this simple, the code is saying this;

If the ammo is above 35 do this. Else if it is 0, then…

 

Play a sound 2

And do nothing.

 

This means that when you run out of ammo, you can no longer shoot.

 

Solving the skybox issue

 

 

I found out by solving this all I had to do was download a skybox from the given link, add the script on to it and turn up the draw distance on the oculus cameras. This solved all my issues.

Tweaks, balancing and changes.

So considering the criticisms recently I have made some changes to the game, I feel that these changes have improved it significantly, even if it was just small changes.

Firstly I have added a small animation and sound effect to the scoring system, every time it registers that you have taken a shot with an animal in it, it will make a quick rewarding noise, and the text will animate. It is possible to play different animations or sound for each of the different animals so I may review this addition just to give it a few different sound effects depending on the animal.

Next I hid some of the other animals the you see straight at the start, I did this to balance out the gameplay as I found there was far too much going on early into the game, and by the time you got to the later parts you ran out of shots. It needed changing because there was too much to take photos of, I felt the players were being overwhelmed with stuff to look at, and really it needed to be a little more relaxing.

Finally my third change was rounding off the points system so it only worked in 5s and 10s instead of odd 1s, 2, and 7s. This made the numbers much easier to understand for the viewers.

My next changes are going to be a little more significant and technical.

Firstly, I am going to get a photo system working for this game. So when you’re playing the photo that you take appears somewhere on the dashboard. It will show for an amount of time then start playing back to the camera. This may be a bit of a technical extra that I will need some help with, will keep posted on this.

 

Importantly though, If I am to add such a mechanic I have one main problem with this. There really is not enough space on the dashboard for the picture, and at that, the ‘screens’ which I had designed in the first place come out far too small in the game, so my though is to redesign the dashboard completely, and possible the car if I really have to, just add that part into it, will keep you posted on the matter.

Instead of having the four little screens, I have now changed it to two screens the points and shots left will be displayed on one screen, and the photo you previously took will be displayed on the other. I have also changed the angle of these two, so they are directly in front of the player at a readable and viewable angle. I have removed the extra details as they no longer needed to be there. I also gave the screens a border around the edge just to make them look as if they’re an actual screen rather than just a rectangle.

I recreated the information shown on the dashboard, with titles a new background image. The titles are part of the image, so only the numbers are actually dynamic.

For the next part, I want to create a tick and cross mechanic so it acts as if it recognises whether you have taken a good photo or not.

Getting the tick to work was an easy job as I could implement the animation straight into the projectile script, however the cross side of the script was much more complicated. So complicated that I did need some help with getting it to function properly.

Originally I wanted it to play as part of the script where the bullet is instantiated into the scene. But since the object did not exist in the scene till the scene was actually active, it was impossible to assign it to anything within the scene, only in the project window. This meant that adding it to part of the script and getting it to actively work was going to be a hard and confusing task. I will be getting help from our technician for this part. Once it is solved I will update about it.

 

Another mechanic I wanted to add, which goes hand in hand with the tick cross function is the screenshot system.  There were two ways that I have tried with this.

One is actually taking a screenshot inside the game, but that pauses the game while you take the photo, and the photo taken can only been seen from the rift, as it takes an image of the screen, rather than just one of the cameras. Furthermore I would have no idea of implementing a facebook system, and also a way of uploading the photo which did not cut away from the experience.  And finally, there is no way I can think of that will save multiple screenshots, unity only takes one, and then saves over the previous photo. If I get time I shall look into this a little further to see if there is some sort of function that would allow to take multiple photos.

Secondly is using a render texture. This system is easier to set up and get working, all I had to do was set up a camera that was parented so it is controlled by the rifts rotation, then use a render texture to out put that into a quad, (which fitted nicely into the second rectangle space I created for it) then code the camera to turn on and off whenever the mouse is pressed, which conveniently became part of the projectile script.

 

It also seems that my game has again been rejected by the oculus site. This is quite annoying. Luckily though they suggested a fix which I have now applied to all of the environments to make sure that the skyboxes are now all working properly. Once the coding issues with the other mechanics are sorted, I will upload the game once again. They should approve of it next time.

Pegg’s Crit.

One of my tutors, who focuses on games had a go at oculus snap, and had some criticisms about the game and where I can improve. I at the time felt quite content with how the game had turned out, but from here I believe it really does need polishing a little bit more. Hopefully I will have some extra input from reviews online once it has been posted to the oculus share site, so once that is up, I will work on it to improve it further. But for now, here is what Pegg said, and what I am going to do.

 

The animations on the animals needed to be much more subtle, right now each of the animations are very much in your face. And as if they are reacting to the player travelling through the environment, rather than acting naturally, the movements should be more subtle and there should be less animals in each of the sections, or more spread out.

The animals should be further away from the camera so that they cannot be easily seen, so that taking a photo becomes more rewarding. In other words, make the viewer look for the animals, rather than bringing the animals to the viewer. To some extent it is too easy.

Make the viewfinder just a dot rather than the square. Or get rid of it all together.

Change the point system to 5 point for one animal and 10 points for another; it will make the numbers easier to work with.

The amount of animals in the environment is too much, tone it down- less is more.

Improve on the feedback system so when you take a photo you feel gratified that you have taken one. Right now next to nothing happens.

 

I feel that there are improvements that need to be made. I feel that for the first half of the game, the amount of animals that animate and that you can take photos of needs to be decreased, mainly because I found players take too many photos early on in the game, and that there is way too much going on at that part to really be taking a proper photo. On reflection of my target placement I feel that I did not think about it enough, and looking back now I really think I should re animate some of the animals, or take away a chunk of them to dub down the experience. It should be more about what they discover, rather than what is revealed it seems. This subtlety will bring up the difficulty, and I might create some sort of mechanic that stops players from taking more than one or two photos of certain animals.

Changing the point system will be simple to do and easy so I will get on that soon.

 

Another suggestion Pegg made was improving the textures and the look of the models. I am quite against the idea of having to remodel one of the animals and have to re texture, animate and place them back into the environment. I understand though that some of the models may not be to a high standard. Instead though, I am going to consider using different shaders to see if they can improve the look of these animals. One thought I have in mind is to use a fur shader, which can be bought off the unity asset store. This would defiantly improve the look of my models; it’s just whether or not it is worth spending the money on it. Thinking about it now, it is probably worth it.

 

One thing I have also noticed when people are playing this game is that they pay no attention to the scoring, or the amount of ammo left, and even though they are preoccupied with the surroundings, I would like some sort of system that might grab their attention that they’re actually doing something that is influencing a score.

 

 

Well to top off this damn criticism, it seems that my game was rejected from the oculus share site; luckily it’s an easy fix, so I shall get on that asaps.

It seems that there is a problem with the skybox in the main game and that it is “splitting” I am unsure what that means, but I think what I am going to do is just replace it with another one, which is what it said on the message, so I might as well do that. To follow up I will reset the menu skybox too, just in case there is anything wrong there too.

 

Testing

Testing at the arts bar night for DMP!

 

My game is finally in a working state, and after countless days of working to the max, I have produced this beast of a game. And it is finally time to do some testing.

Here are some videos from the night.

 

And here is what I learnt.

 

The game works, but there is still a little tweaking needed.

1- The visual blur on the game needed to be eradicated, it is horrible and renders badly. It needs to be excluded from the look of the game, it is awful. This blur was caused by a depth of field blur that I had applied earlier in the development of the game. This needed to be turned off. It was also blocking out the sun rays.

2- Consider creating a camera shutter like effect whenever you take a photo. This was suggested by one of the ex students who visited, which I thought was a very nifty idea I had not thought of. If I can get this function working I will defiantly consider trying to implement something that did this. I will probably do it by creating two quads and giving them both an animation where they animate like a shutter. It will also add some more reward value to taking that photo.

3-add a sight to the game menu just so that people can properly target the menu buttons. This needed to be there just to help people navigate the menu.

 

I also felt that while I introduced the game each time, I had to explain the game each time explaining they only had 35 shots, and only the animals gave points. I might make a hand page to go with the download just to explain how to play the game, when it comes to releasing it. I may do this in the trailer for the game, so it explains it there in a funky way.

 

 

 

The next day I made all of these improvements and gave it a another test run to make sure that it was full proof and ready to be played.

Which it was, there did not seem to be any problems as of yet

With all of these improvements made it is ready to be uploaded online to the oculus share site.