Biosys is a PC game released in 1999, where navigation through its environments is done by clicking an "exit" on a scene, taking the player to a new scene. The player can't move in a scene, and interaction is done by clicking a location in it. In some scenes the player can freely look around in 360 degrees by moving the mouse.

My question is, what technique could be used to render such a scene? I'm asking because that doesn't look like it's true 3D; it looks drawn as if a picture is wrapped around the player, but how would you produce and render a picture in such a way?

Here is a rare YouTube video of that game, beginning at 3:40 where the technique I'm curious about is shown.

And here's a picture of one of those environments: enter image description here

  • \$\begingroup\$ Please narrow the scope of the question. There are lots of ways to render pseudo 3d. Tell us what your requirements are that you can't use 3D and what you've tried already. \$\endgroup\$ – MichaelHouse Mar 14 '15 at 16:12
  • 1
    \$\begingroup\$ Well, I was going to answer this, but it looks like the question got closed. Anyway, it looks like they've pre-rendered the scenes and baked them into cube maps, allowing you to look around but not move your viewpoint. They also sort of fake movement by zooming in on the exit when you move between scenes. Incidentally, Google Street View basically uses the same trick (and even has the same zoom transition effect), except that, of course, they use real 360 degree panoramic photos. \$\endgroup\$ – Ilmari Karonen Mar 14 '15 at 16:15
  • \$\begingroup\$ @Byte56 About narrowing the scope due to "lots of ways to render pseudo 3d", the question states "like the game Biosys", which narrows down the question to the particular effect that I described. The "too broad" rule mentions the potential for "too many answers" or answers that are "too long for this format", but judging by the only answer and comment at the time of writing, there seems to be one standard way of achieving this effect, so the question is perfectly answerable in a concise way. \$\endgroup\$ – Smig Mar 14 '15 at 17:25
  • \$\begingroup\$ Right, can you tell us your requirements? Or is it just "I want to know how game X did Y", by attempting to reword it as something else? \$\endgroup\$ – MichaelHouse Mar 14 '15 at 17:37
  • \$\begingroup\$ @Byte56 I want to know how game X did Y, which can be against the rules but not because it's too broad a question. That's the wrong explanation for closing this question. Also, I didn't reword it, someone else did. \$\endgroup\$ – Smig Mar 14 '15 at 18:03

Its a guess but ...

Maybe they just have a skybox behaviour where they track clicks to particular locations on each "quad" of the box.

This basically means each "scene" is no more than 1 cubes worth of verts and indexes.

Puts a lot of load on your artist though.

You could test this by rendering a normal cube inside out and putting the camera inside it, by turning off the z buffer you would basically remove those square edges you might expect from a normal cube.


Not the answer you're looking for? Browse other questions tagged or ask your own question.