Pseudo 3D parallax engine

Yeah, we don’t know how to call it…
Basically the ones that existed around 98s, where 2d sprites exist within 3d space.

Currently me and my friends will be developing our first, actuallyfinished, game.

To lessen the headache, we want to cut our production trouble by using less polygon, less sprites (maybe not because we have a 2d freak here) and less camerawork.
It will serve as a demo, and possibly the basis of our personal engine.

If you know our other thread, we want to have some jrpg life simulation game embedded in it.

Here are some of the older example.


Noticeably, the character sprites are tied into the camera, and follows it strictly regardless of the vanishing line.
we plan to have higher pixel count so it might look really ugly. we might consider using Persona 3 & 4 lowpoly character instead.

Although of course, some better example would be Capcom’s Breath of fire,
which has higher pixel count but still manages to pull amazing closeup shots.

Some other examples would be thousand arms,
which uses painted stills to accomodate for its incapacity to zoom in.



If anything, to my understanding, this feature is most recently implemented by Bastion and Transistor.

I am more leaning towards Bastion because of the looks, but both games uses orthographic camera and even though the character sprites are basically rendered 3D models in sequence, i saw the prototype and it follows the same idea of using a very simple placeholder as a collider detection.
it also uses no physics engine, which to our understanding is quite the headache to calculate.

while BGE seemingly has it all as a game engine, minus few bits here and there…
it doesn’t seem to natively support 2D GUi or graphical assets so most things are still in 3D.
this is our basic reasoning to consider this design choices as our alternative because it saves more time and space(maybe?) and
still conform to BGE’s work process.
Although having a native cross platform GUI like CEGUI or RPGmaker to support the engine would save even more space (i believe).

I’d like to discuss further on how to make a game like this in Blender engine,
and archive our development progress here.

I look forward to your opinions and inputs.

Supports are always welcome!

You might have to do some proof-of-concept work to see the best way to do this. BGE is setup as a 3D game engine and assumes physics with 3D effects. To get the effects you are looking for, you will have to go out of your way to configure BGE to do this.

As a suggestion, I you might experiment with keeping the 3D nature of the game world, but making it look 2D.

For example, you model all your scenes in 3D with an orthographic projection and your game character is a simple billboard that renders your sprite sheet. Pressing “up” on your controller actually makes your character walk “away” from the camera which will make them appear to be moving up the screen. They stop walking when they hit the 3D wall in your world and the physics engine makes them stop.

The advantage is that you are still working in a 3D world which will make BGE more corporative with you but getting some of the visual elements you want and keeping your total polygon count down.

At one point, I did some work on integrating CEGUI into the BGE. It requires modification of the C++ source code to hook it into the BGEs render pipeline, but I remember it was fairly straight forward. I’m not sure if I even got around to integrating the actual keyboard/mouse input or not. :-/

Hey, thanks Kastoria!

CEGUI on Blender is straightforward huh? I feel like trying them out now!

A proof of concept sounds good, but i fear it’ll take abit too long…
How much does a proof of concept considered …proof of concept?
I do know prototype but they are usually very rough and doesn’t

As stills are not enough, does…fake trailer (entirely animation to replicate game engine) constitute a proof of concept?
If so, i might just be able to produce some bits without getting into the engine at all. :slight_smile:

Currently i do have some videos of Xenogears, one of the game that i refer towards more for it’s controls.
The other games have their directional control locked into predefined angle for their own good, further limiting some its camera view at the cost of its artistic merit.

The video is fairly straight forward, and right into its battle, i hope it illustrates enough of the concept i want before generating further content.

Well, I found it to be fairly straight forward. But I’ve worked on the rendering pipeline for 3 different 3D engines and have years of experience with C++. If you don’t know C++ or know how BGE’s rendering pipeline works then it probably moves up to “very hard”.

A proof of concept sounds good, but i fear it’ll take abit too long…
How much does a proof of concept considered …proof of concept?
I do know prototype but they are usually very rough and doesn’t

A PoC for this would be to try out using the orthographic projections like I suggested with very simple place-holder graphics just to see if you can recreate the pseudo-3D effect easily in Blender. Really you just need to setup geometry for a simple test room, a simple character, and the ortho camera rig. You would want to do that before you make a long cut scene or do a bunch of artwork just to find out that your game feel is really hard to achieve.

The video is fairly straight forward, and right into its battle, i hope it illustrates enough of the concept i want before generating further content.

So the question is, can you recreate that scene but using very simple textures just to get the movement figured out?

Well…we have to i suppose.

Speaking of which i did just found out that building GUI/HUD in BGE could do with overlay scene or viewport.

Or at least according to some of this tutorial

Are you familiar with it?

Err, trying to shoehorn a ‘hacky faked 3D engine’ into a 3D-based engine instead of just using low-res textures and the orphographic camera :spin:

You could try to do that sort of thing with overlay scenes, but purely doing it that way would mean having tons of them. Another method is to have a single scene with a perspective camera and just comprise the scene out of 2D cards that face the camera.

Or, if you want authentic 2D development that requires you to fake 3D, you might want to give an actual 2D engine a shot here (or one that can do 3D, but also has a genuine 2D side to it like Godot).

Yes, trying to shoehorn this proves to be hard.

Our team honestly had the idea just because we would like to minimize graphic constrain and focus on gameplay, but our limitation lies just in middle between going for super pixelated, and being super detailed with normal maps etc.

This genre in the middle suits us best but it seems the concern now is more on the gameplay, than graphic first. ;/

yeah, might actually try ourselves XNA development pipeline…but it’s a shame since i just got here and trying to learn Blender and its engine… :confused:

the 2d card scene sounds like it though, actually tying the angle to the camera and while its position being decided by a simple placeholder.

I think this thread is solved. Found ourselves some solution on extending (quite a new word for us) blender engine with Visual Studio Express for UI development.

It is pretty supportive and we’ve a team that’s die hard supporter of it.

Actually saw some games uses Quake engine and extends it with VS.

thanks for the responses!