OH CRAB - Devlog 1

This week we have been exploring how we want the game to play and look. We are currently in the process of creating our art bible and tech document, while also prototyping mechanics and art assets to see what is possible and what isn't. 

During the prototyping phase we aim to answer several questions: 

1. Is our game fun to play? Why/ Why not?
2. How do we make the sand look believable? 
3. How will we morph the ground when the crabs shovel the sand?
4. How are we going to make the water look believable? 
5. How are we doing to destroy the castles?
6. What camera view, style, and lighting do we want for the game?

 ART DEVLOG:

We decided to go with a mid-poly, stylized hand painted game. We are planning to mostly use non-localized, reusable textures to optimize texture space, while opting to model in most of the detail and baking a light map on a separate channel. After some concepts, a frontal, 135° view seemed to be the most readable angle for our game - to emphasize readability even more, we will be using a slightly post-midday lighting for the scene so the shadows will accentuate the shapes while not distracting the player. 

While deciding the style of the game, we were looking into making a 'cartoon' styled game, but felt like it didn't fit with the environment we wanted to make nor created the look we envisioned. We started off looking into castles that had the style of a typical sandcastle mold, but after some concepting decided that we would go with castles that fit more into the 'Albion Online' style, where the castles would not solely be from molds but also hold some stylistic wood and details that no crab would logically be able to create.

We also ran into problems when looking into the camera angle, we wanted to have a more top down angle but also wanted to show an environment behind the beach. After some brainstorming, we decided to have the beach on a plane with a warped back, and place props on an angle so that we can both implement atmospheric fog and background detail while still having a readable camera angle.

     SAND:

Yin tried to mimic the sand by making a simulation of it in Houdini, however, it has not worked so far to import it into a game engine. I have tried to export it using a vertex animation texture to Unreal, while it moved in Unreal, it did not move correctly. I think it is because the little balls have very optimized UV space and hence the texture is not read and exported correctly. I have made the simulated model into a 'solid' so it does not have small balls, but is one mass instead. This will be more optimized for the game and uses UV space better, I will try to export it again next week.

    WATER:

We also brainstormed how to make a wave and have several ideas that we want to try out. If we decide to go through with Unreal we can use parallax occlusion using an animated texture, and if we decide to go with unity we can try out using blendshapes or an animated height map texture - which we will prototype during the next week. 

PROGRAMMING DEVLOG:

- Which kind of problems did we run into?

Problems arose when we found out that unreal doesn’t have an implemented way of changing heightmap data on runtime, since we want to be able and change the terrain while playing, this was a bummer. Unity provides this on the other hand, but if we want to change larger parts of the map around the person and/or use a brush to make it easier, it dropped our framerate. The quest for the best way of changing the terrain continues, or ‘faking’ the changes to the terrain.

- What worked or didn’t work?

Unity provided a nice way of changing the terrain data, which was a blessing, generating a map based on tiles and interpolating between them to change the map that way in unreal might have worked, if unreal accepted a nested for-loop. 

- What is your plan for the next coming week?

Gameplay:

Work out the actual gameplay so it turns from the literal sandbox it is now , into an actual game:

This means:

• being able to place more defined structures of sand like towers with the use of moulds you can find in the sand by digging
• the sea that influences the beach
• a pearl that you need to protect from the sea, but also from each other

Technical:

The sea part might prove a challenge. We need to figure out a way for it to interact with the beach in a way that’s both realistic and fun. Because we use a heightmap for the terrain. Interactions with the sea are not that straightforward. We could use some kind of sea-brush. But that might not be the best solution because of the framedrops whenever the terrain changes too much.  

Because of these framedrops, we will also be looking further to make the terrain manipulation as smooth as possible.

We can do this in unity with the heightmap by subdividing the beach into smaller terrains and connect those. This way, we only have to update a few smaller terrains, instead of the entire terrain when someone digs.

If the framedrops can’t be simply fixed or we don’t find a solution for the sea, we might look a bit more into other terrain methods like tiles/voxels/procedural mesh, or maybe some combination of these.