Search Unity

  1. Megacity Metro Demo now available. Download now.
    Dismiss Notice
  2. Unity support for visionOS is now available. Learn more in our blog post.
    Dismiss Notice

Isometric RPG game with 2D and 3D assets mixed

Discussion in '2D' started by Aimed, Oct 22, 2019.

  1. Aimed

    Aimed

    Joined:
    Jan 7, 2014
    Posts:
    9
    Is there anyone out there doing something similiar? We're trying to make a game where we build the world/environment with hand drawn 2D sprite assets. These might sometimes have a sprite based animation. For example a windmill turning its vane. But character animations, like humans, monsters etc. are realtime rendererd 3D models.
    We decided to use sprites instead of building everything in 3D, which should be a lot easier, because hand drawn stuff looks more appealing. It reminds us of isometric games from our childhood. Late 90s and early 00s. While 3D models for character animations are an obvious choice. Drawing them by hand, 5 different directions, frame by frame is simply madness. Even in late 90s in games like Diablo, Fallout, Baldurs Gate etc. 3D models were used but they were not rendererd in realtime. Another option was to use Spine and have simpler 2D animations but we didn't like that.

    What are your thoughts on this so far?
    1) Are we working ourselves into trouble here and should we do pure 3D? If so, what about performance? We'd like to make this game work on mobile devices as well and that extra performance we can get from rendering sprites seems useful to us. This game is also online and we have no map on the client. The idea here is to create a real "exploration" experience. Getting sprite IDs from our server and then loading & rendering them is less CPU intensive than doing the same for 3D meshes & terrain. Having every asset pre-loaded is not an option.
    2) Ideally we'd like to render character animations into render textures and use these textures just like the rest of sprites. This way we could use mesh renderers with flag meshes for everything on screen and only use their sorting layer & sorting order, instead of messing with Z coordinates and mixing flat meshes with volumetric meshes. Working with light & shadows if you have only flat meshes should also be easier than having these mixed up. Is this a good idea or is there something I am unaware of?

    Thanks in advance. I hope someone with experience will read this and give me some useful advice :)
     
    Xepherys likes this.