Making a 2.5D game: Everything you always wanted to know, but were afraid to ask
Hi everyone,
Ernest here, CEO of Byte Barrel and Technical Director of Forgive Me Father.
You sometimes ask us: Can the game have this, or why does it have that other thing? Every decision we made was conscious and tailored to our vision, and we want to tell you a bit more about it.
When you make a boomer shooter, you can basically go two ways: either full 3D (like Quake) or 2.5D (like Duke or Doom). We took the latter, and this came with certain, let’s call them opportunities and challenges.
Why not 3D?
2.5D or 3D? This was the first and most important decision to make. But to us, the answer was obvious. We took off with the goal of making a game where every screenshot you take looks like a comic book panel. It would be possible in 3D, but with a much, MUCH larger budget (so the models are complex enough and it looks like an actual comic book).
Going 2.5D, with sprites, felt like the best choice. But also, when we talked about our favorite games, the most mentioned title was Doom, which was 2.5D. So there’s that.
However, 2.5D IS NOT easy
While the technology lets you create a good-looking game with low requirements, making it look good isn’t easy. You have two clashing styles to combine: 3D environment and 2D objects. To make them match, we needed to work with the colors, textures and lighting a lot.
There was also the question of animation. With 2D, you can make it very fluid, but then it just looks off. It’s like we’re conditioned or pre-disposed to expect a bit of choppiness from moving pictures. So, we had to establish how fluid our animation should be, how many frames it should use.
This was especially hard when it comes to death animations. They have to be punchy. With a 3D model, you can use a physics engine and just go crazy with ragdoll, or make limbs fly off, easy peasy. In 2.5D you have to draw everything, so your static drawings have to look dynamic. We added a punch to death animations by having enormous blood splashes and emphasized headshots with prompts above enemies' heads.
Another big challenge was creating huge, huge bosses. Sprites are just fine when it comes to small enemies, but large ones look off once you start moving to their sides. They start rotating and it simply breaks the perspective. Our solution was to divide the enemy into separate bits and, contrary to basic enemies, not make them rotate to match your viewpoint. We also added extra debris and attack animation effects to add more 3D-like depth to the entire thing.
Finally, perhaps the most difficult thing was achieving the desired goal of making a retro game in a modern engine, that would be visually referring to timeless shooter classics made in the 90's, not to mention keeping that old school vibe.
We had to establish a process
For anything in game development, there are processes. Some are universal, some are made at the studio level. We created our own way of making 2D art. It goes like this:
1. We look for references in real life or in art. (We referred mostly to Frank Miller’s and Mike Mignola’s graphic novels.)
2. Then we make some rough drafts of the object or enemy in its default pose.
3. The team chooses the best version.
4. Then we make a complete, colored version of this base sprite.
5. We start adding frames until we reach the desired effect.
Of course, there were compromises
Regardless of the kind of game you’re making, there’s always something you have to cut. Initially, we intended to have more frames of animation. But as the project grew and the demand for animated objects with it, we decided to reduce the base number of frames for every animated object to be able to actually draw everything and keep it consistent with each other.
This also includes the dynamics of animation. With a reduced frame count, we couldn’t go too crazy with how enemies move, because going from position A to B would look too jarring. The enemies became less "flexible" in this regard.
Still, it’s not as bad as it might look. ;)
Being able to draw a lot of elements as 2D objects lets you make them look really unique and detailed. This would be impossible to pull off in 3D on our budget. Also, it gives the game a distinctive look and feel, which we very much wanted to accomplish.
3D is cool and all, but if you don’t want to go low-poly, it’s expensive, both when it comes to money, time, and computing power. At the same time, you can draw anything you’d like. Sure, it’s not gonna move as fluidly, but it can look stunning. It’s a great way to realize crazy visions on a budget.
We hope we proved it in Forgive Me Father!
Got other questions? Don’t hesitate to ask here, in the comments or on our Discord or Twitter.
Ernest here, CEO of Byte Barrel and Technical Director of Forgive Me Father.
You sometimes ask us: Can the game have this, or why does it have that other thing? Every decision we made was conscious and tailored to our vision, and we want to tell you a bit more about it.
When you make a boomer shooter, you can basically go two ways: either full 3D (like Quake) or 2.5D (like Duke or Doom). We took the latter, and this came with certain, let’s call them opportunities and challenges.
Why not 3D?
2.5D or 3D? This was the first and most important decision to make. But to us, the answer was obvious. We took off with the goal of making a game where every screenshot you take looks like a comic book panel. It would be possible in 3D, but with a much, MUCH larger budget (so the models are complex enough and it looks like an actual comic book).
Going 2.5D, with sprites, felt like the best choice. But also, when we talked about our favorite games, the most mentioned title was Doom, which was 2.5D. So there’s that.
However, 2.5D IS NOT easy
While the technology lets you create a good-looking game with low requirements, making it look good isn’t easy. You have two clashing styles to combine: 3D environment and 2D objects. To make them match, we needed to work with the colors, textures and lighting a lot.
There was also the question of animation. With 2D, you can make it very fluid, but then it just looks off. It’s like we’re conditioned or pre-disposed to expect a bit of choppiness from moving pictures. So, we had to establish how fluid our animation should be, how many frames it should use.
This was especially hard when it comes to death animations. They have to be punchy. With a 3D model, you can use a physics engine and just go crazy with ragdoll, or make limbs fly off, easy peasy. In 2.5D you have to draw everything, so your static drawings have to look dynamic. We added a punch to death animations by having enormous blood splashes and emphasized headshots with prompts above enemies' heads.
Another big challenge was creating huge, huge bosses. Sprites are just fine when it comes to small enemies, but large ones look off once you start moving to their sides. They start rotating and it simply breaks the perspective. Our solution was to divide the enemy into separate bits and, contrary to basic enemies, not make them rotate to match your viewpoint. We also added extra debris and attack animation effects to add more 3D-like depth to the entire thing.
Finally, perhaps the most difficult thing was achieving the desired goal of making a retro game in a modern engine, that would be visually referring to timeless shooter classics made in the 90's, not to mention keeping that old school vibe.
We had to establish a process
For anything in game development, there are processes. Some are universal, some are made at the studio level. We created our own way of making 2D art. It goes like this:
1. We look for references in real life or in art. (We referred mostly to Frank Miller’s and Mike Mignola’s graphic novels.)
2. Then we make some rough drafts of the object or enemy in its default pose.
3. The team chooses the best version.
4. Then we make a complete, colored version of this base sprite.
5. We start adding frames until we reach the desired effect.
Of course, there were compromises
Regardless of the kind of game you’re making, there’s always something you have to cut. Initially, we intended to have more frames of animation. But as the project grew and the demand for animated objects with it, we decided to reduce the base number of frames for every animated object to be able to actually draw everything and keep it consistent with each other.
This also includes the dynamics of animation. With a reduced frame count, we couldn’t go too crazy with how enemies move, because going from position A to B would look too jarring. The enemies became less "flexible" in this regard.
Still, it’s not as bad as it might look. ;)
Being able to draw a lot of elements as 2D objects lets you make them look really unique and detailed. This would be impossible to pull off in 3D on our budget. Also, it gives the game a distinctive look and feel, which we very much wanted to accomplish.
3D is cool and all, but if you don’t want to go low-poly, it’s expensive, both when it comes to money, time, and computing power. At the same time, you can draw anything you’d like. Sure, it’s not gonna move as fluidly, but it can look stunning. It’s a great way to realize crazy visions on a budget.
We hope we proved it in Forgive Me Father!
Got other questions? Don’t hesitate to ask here, in the comments or on our Discord or Twitter.