Hi folks, I'm in the early stages of developing a retro style 2d game for fun (so I may have some work to throw your way in the future!) ;). As a programmer, I know next to nothing about graphics/projections and stuff like that and I have a question about the graphics in this game:
My question is whether or not this is mixing 2 different perspectives in one image. Since you can see into the goal mouth it is not perfectly top-down, the camera is tilted slightly. But if the camera is tilted slightly then the lines on the pitch shouldn't be parallel, right? But the pitch is a perfect, undestorted rectangle.
I know old games probably just did this to simplify things, but I'd be interested to know if I'm understanding it correctly.
Thanks!
Mixed perspective was a thing in old paintings, too. Heck, cubism was all about playing with perspectives and mixing them up. Then there are hugely influential games like Atic Atac and The Legend of Zelda. So you're in excellent company.
Thanks for the reply!
So if the lines on the pitch were to match perspective with the goals, they should diverge slightly as they progress down the screen?
I guess strictly speaking then, the actual goal posts should not be parallel too, right? They should diverge as they increase in height.
I still haven't decided which approach I want to take in my own game, I'm still trying to educate myself with questions like this so I can make an informed decision :) It's certainly easier this way, but a render system using proper perspective might allow for some nice camera effects (in one dimension, at least).
btw, knowing a bit about the history and development process of this particular game, I don't think it was a conscious design choice, just a matter of sticking some sprites from another game on top of a large soccer pitch background.
And then the stands would hape to appear smaller as they are farther away... the players and ball too... before you know it, you went 2.5 at the very least, if not full 3D.
I made a game like that, you know. One of the few I worked on professionally. The background, including the goal, was prerendered, while the ball was handled through sprite scaling. It was pretty damn tricky to get right especially as 9 years ago I had little experience making games.
You can do it. But you don't have to. Keep it simple, and you might be surprised how well it comes out anyway.
Yeap you're right. Especially since I'm a one man team I think it's best to keep this one simple.
I'm taking a similar approach to what you described actually. The philosophy of the game is to make the ball the star, so I've given it pretty solid physics whilst everything else is just pre-rendered with the "sensible soccer perspective."
Take a look if you're interested:
http://www.higgis.at/senseless-soccer/videos.php
P.S. here's a couple of interesting vids showing my attempt at an algorithm to make the ball bounce in line with the sprite / goals perspective:
https://youtu.be/LseDNpbQKiA
https://youtu.be/LdijBxaBer4
https://youtu.be/3Kv7Vj1pHCE
and here's the same bounce without the prespective correction:
https://youtu.be/EpEkJu_O-9s
In 2d games, the best philosophy for graphics is "gameplay first, logic second". First ask yourself what information the player needs in order to play the game (first and foremost, position and hitboxes) and make the art reflect that.
Nobody will care or even notice if the perspective makes no sense as long as they can play effectively, but if the hitboxes are hard to judge people will notice even if the game looks great. Since the height of the ball probably isn't as important as the x/y position, focus should be placed on making sure the player can easily judge the position of the ball and the goal hitboxes.
Make sure the shadow is directly below the ball, and the goal area is clearly marked with visible lines on the ground. The graphics should then be built around that.
Couldn't agree more! Have discovered exactly the same things myself throough experimentation, especially the point about the ball shadow representing where the ball actually is.
The point about the hit-box for a goal is precisely why the ball physics/graphical representation MUST take perspective into account. Otherwise the ball can look like it's in the goal (due to it's hight) but actually hasn't corssed the line yet in the simulation model.
Cheers!