Now that the project is over I can discuss aspects of it in more detail. As the project has gone on I tried to keep my blog updated, but during the pandemic I was merely doing a lot of testing. Until I was 100% sure I was able to use assets from Wind Waker (WW) and get them to work properly in Unity, so there wasn't a lot to talk about.
Admittedly I did somewhat forget about the project near the end of lockdown, I hadn't seen my immediate family in nearly 5 months, so I ended up having them over for well over a month. So as you could imagine, little to no work got done during this period!
When I got back into it though, I really got stuck in. I went from a tech demo to a fully playable project in about 2 weeks. This was possible because I'd already done any testing I needed, so I knew models would work and because of the work on my 2 previous projects, my skills were up to date and I had scripts I could migrate over to smooth the process.
Initially though, testing was not a fun process. I detailed a walkthrough further into the blog that goes through the process of extracting WW models and getting them to work nicely in Unity. This was a long process to nail down, the guide looks simple now but in reality, it took months. In fact, I vividly remember testing with the red post box, this took weeks! At first it wouldn’t extract, then when it did, its textures were missing, then its mapping coordinates were always obscured, then animations would break it even further and then scaling became an issue, until finally I got it working.
That was just the post box, Link had many more issues of his own. For example, his rig I had to export him many times and the plugin would mess up his bone structure randomly, so when i got him working then next time i exported him he was broke again. I had to manually re-rig him in Unity every time, until I found out I can just copy this Unity rig to each following import of him.
Link's Movement
After that, the next major issue was translating his animations from WW and getting them to work with the movement system I had set up. I did manage to get him walking and running, though anything else he would just refuse.
I did finally manage to export all the animations I would need for my vertical slice and after a lot of tweaking got them working with his movements. Turns out most of his animations needed specific tweaking to ensure they lined up correctly. But still his movement is not 100% I would have liked to make him feel just that bit more fluid, I think my biggest gripe is his jumping. He uses rays to detect ledges, but these are not very accurate, or at least they’re not for me, as Link won’t always jump, in fact he can fall off a little ledge and take damage (which is infuriating).
Also, to save time I had to weigh my options in regards to his overall interact-ability. For example, I set up Links entire movement system, but because he has multiple states there is much more than what I made, this is his movement system currently:
There should be at least another one like this because his movement and actions are slightly different for when he has his sword drawn. Switching between the two depending on context. It is at this point I decided to have more quality that implementing unneeded aspects. To get around this I attached his sword to his body and simply enabled its visual render when the player attacks. It is simple but effective.
Area Transitions
So, that was just Link and his movement. Another interesting obstacle I came across was interiors and area transitions. I had never done anything like this, connecting other scenes to a central hub so to speak. I had no idea how to code it or how its typically done.
I approached the idea using mini-cutscenes in the follow way; each time the player is in a trigger zone by a door and uses the interact button, it loads the attached timeline (cutscene) and plays it. Because of the work on his movement previously I made sure he had no root motion, so that means when I move him in cutscenes he actually moves and stays / continues from any movement.
So, with that in mind, each door is a cutscene. I found his open / close door animation, I trigger this and add a fade on each end to make it look much more cinematic. The logic behind this though is a bit more complex, I scripted a means to determine where Link has come from, where he is going and what should be loaded.
To do this, when the door is interacted with I run a specific routine, so for this example I will load the first house interior:
This tells my script to load said level and sets a variable to remember the current scene before the new one is loaded. It then loads the interior House A, which because it’s an enclosed area has its own cutscene upon loading. Now when exiting and returning back to the main island / hub my script detects this. Because it’s an important object I make it persistent throughout the game, so once it’s loaded it will never be deleted, even upon entering the same hub where its first initialised.
Anyway, every time the main island scene is loaded this runs automatically:
So, this checks that Link is returning the "Outset" (which is the name of the island) and if he is, it checks the variable mentioned previously to see what interior he's currently in. Because we know he travelled to "House A" you can see in the code above it notices this. So its checking that the level being loaded is in fact "Outset", if it is, what level did we just come from? If it's "HouseA", it finds a cutscene I made of Link opening a door back into Outset and then plays it immediately.
Now usually when reloading the same level, the player will always respawn where he is set in the level. But because I'm searching for a cutscene every time this scene is loaded, it plays said cutscene at the location I set it. So for House A, I placed this cutscene outside the hut, so when the code catches this event, it places him there so it appears as if he came back, when in fact it’s just re-loaded the level as if it was freshly loaded at the beginning of the game!
It may seem like a lot to take in, but this was the only way I could think of achieving this and it works and looks great too!
Grass / Breakables
This again was another aspect that caused me issues, but I managed to overcome. When I was testing enemies, I had a script to govern them, it allows me to control their health, spawn stuff and effect Link in a number of ways.
I realised that grass and pots were a type of enemy. They might be static, but they can be "destroyed" in a similar way, ultimately removing them from the world. So, I figured I would place all the grass in the level using a prefab, so that I can modify them all at once with ease. I placed around 400+ grass objects into my scene, all running this enemy code, I did not anticipate how un-optimised it would be. Playing the project caused my computer to lag for half an hour before I could return to editor (I had not saved and didn’t want to lose all my hard work).
The overall framerate for the project went from 100+ to 1 to 2, sometimes 5 depending on the camera angle!
Because cutting grass is a pivotable aspect of any Zelda game it was crucial I got it working. I ended up removing it for a few days because it was too unpractical to use, I even searched online for methods of instancing objects to reduce draw calls.
It then dawned on me... Because the grass prefabs are spaced all over the island, they don’t need to be constantly checking the player. With over 400 grass objects all simultaneously checking the player for input it was bound to make the project run slow. I fixed this by adding a distance check between the two for every object instead, just 3 lines of code improved my framerate massively:
So instead of checking the full enemy code many times a frame for each individual object it now checks the players distance to itself once a frame for every single grass piece, which is much more efficient. So, when the player is about 5 feet away, they then enable themselves to be interacted with and voila! Massive speed increase and I got to keep one of the most important elements to the game.
Sounds
I discussed this a few posts back, but I'd like to reiterate. Without them the project was bland, I cannot believe how much of an impact they make on the project. Even little things such as footsteps or the slicing of trees. Nintendo know this, that is why they put emphasis on sound through gameplay. In WW, they have chimed attack sounds for fighting enemies, the tone changes to create a melody every time Link attacks. Same for Mario in the New Mario Bros games; enemies and props dance in rhythm to the music!
I would have liked to include this but I did not have enemies so it would have been redundant.
Sounds however are something I will endeavour to include straight away in future projects instead of trying to retrofit them in later. As sound design is just as an important element as anything else in a game.
Improvements
I think the only think I'm sad about not getting chance to implement is a more dynamic interaction system. One where I could set contextual parameters such as: camera positions, player / NPC animations, sounds or even give items mid-conversation, all per message / dialogue box. It would have taken too long and not really required for this small vertical slice, but would had looked really nice.
Conclusion
So those above were just a few of the issues I encountered, they were probably the biggest though. I have learnt a lot throughout each of my projects for this course. I've learnt how to prioritise workload, not to create too ambitious milestones, coding efficiency and so much more. All of which I will continue to use in future projects.
I like to think that my work does highlight the issues of remakes and remasters in the industry. How I've managed to create 3 - albeit small - games to a pretty reasonable standard in such a small amount of time on my own. Then why are other studios, not putting as much effort into theirs with much larger teams.
If anything, I'm even more spurred on to create a fully functional game of my own. Not a remake or remaster, but something I can pour all that I have learnt into and call my own!
I hope that anyone who encounters my work will appreciate what I have done and that I inspire others to realise and act on the subject I highlighted:
How are Remasters / Remakes Effecting the Industry?
