DEVELOPMENT

11

2,945

Island Tileset, Resonar Slashes

4 years ago - Edited 4 years ago2,197 words

I'm trying to get to a point with... whatever this Divine Dreams prelude thing is going to be called where I can show off snazzy gifs of gameplay etc to generate interest in the project as I work towards completion. So this week, I've built an island tileset for Savitr and Collie to run around on, and animated their physical attacks so they can fight monsters.

I learned from doing research a few weeks ago that I need to start promoting what I'm making sooner rather than later, so then there can be a buildup of interest over months. So although I've still not fully planned the story - or even the title! - of this prelude thing I'm making, I've spent this week on asset creation and implementation so then I'll have something shiny to show off.

I've built (the beginnings of) an external tileset, and I've made it so that both Savitr and Collie can attack with their weapons. Here's a short video showing both:



There's also a short exchange between Savitr and Collie at the start, mainly to show the conversation system to people who might not have seen it, though I'm still not decided about her personality (how obsessive or sexual she should be).

(The placeholder music's an island theme I composed for Belief; I want to compose something new for this.)

Tileset

I built the basics of Divine Dreams' field mechanics ∞ way back in late January ∞. I started with an external environment, which was tile-based and extremely blocky:



Areas were also divided into rooms depicted as 'dioramas', with all but the current one hidden, because the alternative would be making something to fill up all the non-playable space.

In February, I experimented with ways to build these tile-based maps in a way that looked more organic. I was quite happy with what I had for the Dreamcave:



I've not done anything else with the graphics since then as I've been focused on other stuff, though when I've played games - or watched videos of other people playing them - I've been paying careful attention to graphical details that I might be able to make use of myself.

Animal Crossing: New Horizons was big recently, and I got that a few months ago since I'd been curious to play a game from the series for years. I was pleasantly surprised to see that it used a tile-based system for its island map, and I found it particularly interesting how it handled cliff edges:



(I'd use a screenshot of my own playthrough, but it's such a pain getting them off the Switch! So that one's from Google.)

I found it interesting because it's something that I kind of discovered by myself when making the Dreamcave tileset. To make things more organic, instead of sticking with just cubes, triangular 'half tiles' can serve as effective corners without needing to make too much additional geometry.

Animal Crossing's cliffs also seem to be mostly defined by a simple, stylised texture rather than geometry. They're not flat (look at the bottom), but little would change if they were. That saves on geometry and makes them easier to tile, too. With the Dreamcave tileset, I was trying to get a rocky look with geometry instead, but never felt happy about it. Seeing this made me feel maybe I could get away with just using a texture.

What I made in a couple of days as one person obviously can't compare to a professional game like Animal Crossing made by a team of professionals, but I think what I do have is maybe passable:



I like this combination of square and 'triangular' tiles for cliff edges, as it looks simplistically stylised without being so blocky it can't pass as organic. Animal Crossing's tiles are more complex than what I have here - there's subtle rounding on what looks like multiple tiles at corners, and they tile together seamlessly with added variation - but I'll stick with the simplicity for many practical reasons.

I've been wondering how best to tackle grass in an easily tiled way for years now. What I had in MARDEK was extremely basic:



I remember experimenting with it heavily back in 2011 with this game called "Alora Fane: Regression" (the beginnings of the Alora Fane world, though I did little with this particular game):





Obviously that second screenshot's more detailed and complex than anything in MARDEK, though I found it an ordeal to make the tiles and didn't care for the overly-busy result either.

For Taming Dreams, which came several years after that, I returned to the simpler, more abstract approach:



I'm mentioning these because I keep these old solutions in mind when tackling new problems.

I essentially tried to replicate that in 3D when I first made the Divine Dreams' grassy external area in January. The 'grass' was just a flat green plane with a few modelled tufts in different shades of green:



I noticed that the 'grass' in Animal Crossing was very abstract - literally coloured triangles - but it really worked despite that. It's not simply tiled, though; the texture extends over many different tiles (again, a team of professionals made this).

I was curious to experiment with a similar thing myself, drawing upon both this as inspiration, and what I learned years ago from experimenting with Alora Fane: Regression's grass tiles. I drew this texture:



Obviously that's not 'realistic', but nor is it as extremely stylised as Animal Crossing's. I also like that it incorporates the aster motif to some degree; having some order to it like this also makes it look better than if it was just a random mess of grass blades. It's still chaotic enough to be organic, but it's an ordered sort of chaos.

Speaking of grass, I've seen animated grass tufts in other games, but was never sure how it was done. Turns out it was easier than I expected to do with a simple shader effect:



(My gif recorder has been producing some very un-smooth results lately...)

Those grass tiles look like obvious tiles, though (that is, their edges are too clear), and the angle they're at looks quite ugly at the angle the camera's at (if they 'leaned back' more, it might look better), so I'll probably play around with that a bit. Maybe the blades don't need to be so dense...

Finally, area borders are a pain, and the reason I originally went with the dioramas look is because I wasn't sure what purely decorative geometry I could make around the playable area. ∞ In Belief, I surrounded the gameplay area with an ocean ∞, which I've adapted and reused here (I wrote the shader myself, unlike the sea in Sindrel Song). The world the Bold live in is a series of archipelagos anyway, which is convenient. This game will be set on a small island, and I might divide Divine Dreams' world into a series of small, self-contained islands too.

I've wondered quite a lot how to handle shorelines in a convincing way. Animal Crossing's shorelines look amazing, but I don't have the technical or artistic skill to pull of anything nearly as good, so maybe this fairly mediocre thing I've got will have to do. Most people won't care anyway, and getting caught up on details like that only leads to burnout and never finishing.

Building these battle 'arena' models is something I've been thinking about a lot, too. They're not the same kind of challenge as building a tileset, though some skills and ideas definitely carry over.



How to handle the distance is a problem. Here, I've used flat cutout images, which I think works okay and which is easier than making a bunch of 3D geometry.

When making games, these are the sorts of details that take up so much of the development time but which mean little to the average player in the end. Maybe you found it somewhat interesting to read about though.

Battle animations

I've decided to stick with the resonars for both Savitr and Collie. I think the similarities and contrasts between their attack animations shows they're both part of the same group (Seraphim), but he's obviously at a different ability level to her.

Like with everything I make, my animations aren't at a professional level, but I feel I'm getting better (I like Collie's victory animation), and I hope that these would at least be satisfying for the player. I feel that they are for me, at least. There are details like kicking up sand while moving, and the coloured weapon trails, which really help it feel more polished.

I felt that having a distinct hit sound effect for the resonars would go a long way towards adding that sense of visceral appeal. I'm not used to making my own sound effects, but I did make this one by combining several different sounds together, which seems to be what actual sound designers do. I wasn't sure about the result at first, though the more I hear it, the more it grows on me. It's definitely a skill I need to practice though.

I've recoded the previous reaction system to allow for multiple hits, as seen with Savitr here. There are a few concerns with this, though.



The way the stats are set up, pieces of equipment each have a number of properties which contribute to the character's stats, or which activate on certain types of reaction. The resonars, for example, each boost the attack stat by 10, and add 5 to that on a successful Attack reaction.

But then does that mean that each of Savitr's hits is as powerful as both resonars together?

Or would his attack be denoted as, say, "10/10", and each hit in this combo would only use one of those values?

What of reactions, then? Would that be a "5/5"? What if one resonar had a chance to grant a status effect on a successful reaction? How would that be shown on this status screen (currently, the REACTIONS list at the bottom combines all effects that'd trigger on successfully activating that reaction type)?

Or maybe Savitr's double attack would trigger a Light and a Dark reaction instead of two Attack ones? They're normally used for 'magic offence' and 'magic defence', sort of, but they could be different for different characters, perhaps?

The way I've written the code, characters - or monsters - could potentially use multi-hit attacks which could chain together several different reactions. I'm not sure how they'd work either. It's coded so that each hit has a 'hand' assigned, so those hands could use the stats of just the weapon in that hand. Hmm.

I'm unsure!

I'll be building on this system to work for musical magic; that's one of next week's tasks, so I'll talk about it then.



With how things are coming along, I might be able to make at least a video demo by the end of next week - or the week after that - which I can use to start promoting the game... though I still need to decide on what it's actually called, so I should do some more story planning next week.

Now that I'm working on this smaller project instead of trying to figure out what to do with some vast trilogy, I feel like progress might go somewhat smoothly from here and a release might only be a few months away. But I've never had much luck predicting this kind of stuff, so let's just see how it goes.



There are other things I need to see to, but haven't because the depression which is usually hanging over me had a fairly severe flare-up this week. I managed to get this productive stuff done despite it, but I've been in a lot of mental pain and I've been avoiding interacting with other people even more than I usually do because I feel I've inflicted my insanity on others enough and feel bad about that.

Mental issues are perhaps the price creators pay when they make their deals with the Devil for their creative abilities, though my life situation's certainly not helping at all. Money woes in particular weigh me down, though I need to actually finish and release something again if I want to have any hope of that changing. So I'm trying to get to that point.

I would have posted more about this stuff on ∞ my Patreon ∞, as I've posted a lot of other development stuff on there; these mental issues are why I didn't. Annoying.

Anyway, I didn't feel that was worth writing a separate personal post about, but it's got in the way of some things, so I wanted to at least mention it.

Hopefully this direction I'm going in seems okay.