Windoendo
Devlog 04 - One year later
I took a huge gap in the devlogs to focus on the actual development of the game. So this one is a bit longer to compensate.
Here is the current showcase:
Owl updates
Visuals
After having the prototype running, I decided to create a new mesh, which took ages since I am not a big artist. This time decided to go with individual feathers instead of a single plane. In Blender, I tested using the particles' hair to populate the bird which worked nicely, although exporting it was not an easy task since there is no easy exporter that would fit my needs. From what I understood there Unreal supports a type of hair particle, but it wouldn't import my feathers meshes, and textures. So instead I converted everything to mesh and exported it.
Because of the post-process of the game the textures that I had draw for Blender end up much darker, a solution was to add some emissive color, but the black colors ended up not looking so good
So far so good, so now the animations. A lot of rigging to make my job easier this time. Created a bunch of animations, up, down, left, right, and in the end I am not using half of them. To control the wings I had to break the movement into up and down animations and in the anim graph change between the two depending on the input.
To make it look better, a bunch of springs were added for the feather bones, so it looks more natural without animating everything.
Code
I am not going too deep into the code. Basically, a lot was done to calculate the rotations of the body, each input is associated with a lot more things than what I was expecting. Interping all the parameters to make it look smooth. Then there are a few edge cases to add, when the wing is closed creates more states, and lately I added the breaking, so breaking with both wings enters in the full break, which adds other states...
As a next step, I need to change the velocity also in the Z axis, to create some more fluid and funny movement. Currently, everything is calculated as forward velocity, and set on each frame to the velocity vector.
Camera
I wanted to follow the design of Shadow of Colossus, since it is great to read the input of the player and move the camera to capture mainly landscape and where the player wants to go. In the end, I found some restrains in the camera while doing this, mainly moving in all the 3D axes, plus the fast movement of the character created some conflicts in a lot of angles, so I had to reduce it a bit to not lose the character from view. Still need to tweek it more to achieve the character more on the sides, but I am happy with the current result.
Here is just a center camera to get as a reference.
Island updates
After a while, decided to move to a smaller scale, mainly for testing and visual representation. Instead of huge terrain that I would have to edit, and to be honest was not getting the fancy results that I wanted, I tested with some small hills creating some islands. Of course, I still wanted the fancy terrain for the level design, but my main focus was to make the owl look and feel good.
A lot of the work to make the island look good was the materials, and here, a big shout out to PrismaticaDev https://www.youtube.com/@PrismaticaDev A lot of the materials in the game are mostly from his tutorials.
Trees / Props
I took a while to think/create the assets of the world, starting with the trees, I thought about trees with holes that the owl could go through. Although in the game it doesn't look that appellative, the size and attention to them are quite reduced from what I was thinking initially, also some of the current trees are from an asset pack so it can give more life to the game.
Instead of giving my full attention to the trees I moved to the totems, these grab much more of the player's attention, and when the owl goes through it gives some speed boost as a reward, maybe I could also add that to the trees, but it needs more design thought on that. Currently, I have two totems but I only use one. They need some materials/VFX to give more feedback when the player goes through them. From a design perspective, I could do a game objective finding and going through all the totems, but still need to try other things first.
Here was the point when I decided that the bright orange would be my "yellow paint".
Clouds
At some point, and I don't have any videos or images, I tried to create clouds using the Niagara system. It was ok'ish. But it was not what I wanted for the game. I left the idea for quite a while but one day I saw this tutorial VDB Stylized Fluffy Clouds and it worked quite nicely, plus I really enjoy going inside the cloud.
A problem that I noticed is that the clouds don't get rendered after the limit of the post-process fog. So I used that "problem" to make a feature and used that limit to spawn the clouds and have a cool fade-in fade-out.
Tests in the dev meet
Recently I took the game to test it out with people.
My annotations from the tests:
- Visually is really interesting
- Controls are confusing and quite hard to handle
- People that were able to handle the controls took around 1-2 min to understand them
- The totems are hard to go through, they should have some help
Conclusion/ Next steps
It's getting there. looking back, a year is a lot! but without a routine or, mostly working when I want, there is not as much progress as I wish I had. Non the less, I am happy with the current results.
What didn't wen't well:
I lost a lot of time doing this new owl mesh, which I understand considering I am not used to do art, and what I wanted was a bit on top of my current skill level. I did a lot of animation that also consumed a lot of time that I probably won't use, but it was a test non the less.
Some lack of motivation and not getting that rotine probably slowed down the project.
I don't consider a negative thing, but a lot of struggle from the test players seem to be an issue, need a tutorial or a way to explain the inputs in a more friendly way.
What went well:
I really enjoy navigating the owl, even with a small map it feels nice. I feel easily drawn to the flow and spend some more time playing then the desired "just start and test this thing" time. Also saw it on some test players after getting the controls right.
Not a lot of bugs, or at least easy to solve. The code seems clean although very complex, but even if I don't remember where are the things it's easy to navigate.
So proud that the clouds looks so nice and took very little work.
Visualy looks great, a lot of people seem to enjoy to watch the game, it looks a cozy caual game until they grab the controls...
Next steps:
Release a build? I am taking more time with the game before releasing it here, I would like it to grab more people and to be polish enough for anyone without understanding the game to play it. So it might take a little more time.
Main steps to do that: Have a small tutorial or a way to explain the controls. And have at least a way to exit the game without alt+f4.
Probably once that's done I will publish it.
Would like to do: a barrel roll movement, do something with the horizontal thumbsticks input other than break (might need to change how does it work), help the owl to go through the totems when close, Z velocity movement.
Ok that's all.
Leave a comment
Log in with itch.io to leave a comment.