This is mainly a theoretical question and I'm asking purely to find things out and raise any important points.

Let's say I wanted to make an animation overlay of rain falling down or something to that effect, and let's just say its running in 640 x 400 resolution. Hypothetically, I'll be using SFML or SDL to get this done. There are two ways that I can think of getting this done, though I'm sure there are more:

1. Create a sprite animation of rainfall on top of some weird purple to act as a color key/mask and then render that on top of the rest of the game's graphics every frame, omitting the purple
2. Create an animation to overlay using gometric primitives, probably modularized into a function for ease of use, which would entail coding their animations, as well.

Which of these two means would be more effective and memory-saving? Thank you for all the help :]

Colton

It depends on implementation. You can make both efficiently or inefficiently. I would go with the second option for two reasons:

1: Everything is stored in code - less data required. This has one pitfall though: not moddable by end-user. Is this a problem though?

2: Since everything is in code, it is trivial to implement fancy effects(event based for example, e.g. rain water turning into blood and into bloody snow?) to this kind of particle-based effect.

For the sake of this problem, we can assume that being able to be modded by the end user doesn't matter too much. I appreciate the response.... it seems like coding would be more difficult to implement but less data-intensive, so maybe some hard work here would result in a payoff. Thank you much :] I also enjoy the fact that, like you said, it wouldn't be too difficult to change the effect to my liking for other results.

Colton