Jump to content

Do sclaled-down objects take less processing to render?


volnaiskra

Recommended Posts

While working on my current mod, I sometimes scale things up or down slightly for aesthetic or practical reasons. I wonder whether this has any impact on fps.

 

I guess a scaled-up object has the same amount of polygons, but it still requires more pixels to be calculated. Also, I can imagine that lighting, shadowing and texturing on a bigger version of an object might often be more complex than on a smaller version.

 

Does anyone have any more educated answers/speculations about this? Would frequent up-scaling (SLIGHT upscaling - I'm not talking about making sky-scraper sized barrels) of objects have a noticeable reduction of FPS? Or, if I'm building a whole bunch of buildings in an outdoor area, if I make everything a little bit smaller would that make the mod run smoother?

Link to comment
Share on other sites

I'm not sure if it would make a difference as the object would still have the same amount of polygons, and would still have the same sized texture, it'd just be stretched a bit.

Polygons yes, and textures almost definitely yes (there perhaps exists a theoretical possibility that things like anisotropic filtering would become more taxing on larger models). But what about lighting and shadowing? A large object will cast a larger shadow, which will affect more objects below it. And its increased size might mean that there's more SSAO, lighting, and reflections to apply to the screenspace. And possibly more AA processing because of its bigger edges.

 

I'm sure in many cases the fps difference would be 0. But I'm also pretty sure that in some it would be more than 0. I guess what I'm wondering is how much more, and how often.

 

 

Nope. If that was true, games would just involve extremely tiny models, and there would be no need to buy new hardware as modellers would just halve the size of everything.

I don't think we're talking about quite the same thing. I also don't buy the notion that if what I'm saying is correct then suddenly all game designers are going to set their games in Lilliput. That's as ridiculous as saying "if AA decreased performance, then games would never use AA". Actually, it's a little bit more ridiculous than that.
Link to comment
Share on other sites

 

Nope. If that was true, games would just involve extremely tiny models, and there would be no need to buy new hardware as modellers would just halve the size of everything.

I don't think we're talking about quite the same thing. I also don't buy the notion that if what I'm saying is correct then suddenly all game designers are going to set their games in Lilliput. That's as ridiculous as saying "if AA decreased performance, then games would never use AA". Actually, it's a little bit more ridiculous than that.

 

It's ridiculous? If the smaller the polygonal objects are, the better the performance, then game designers would scale their 3D models way, way down, and the game designer would just have to place the camera closer to the models.

Two completely different things, bud.

Link to comment
Share on other sites

 

Nope. If that was true, games would just involve extremely tiny models, and there would be no need to buy new hardware as modellers would just halve the size of everything.

I don't think we're talking about quite the same thing. I also don't buy the notion that if what I'm saying is correct then suddenly all game designers are going to set their games in Lilliput. That's as ridiculous as saying "if AA decreased performance, then games would never use AA". Actually, it's a little bit more ridiculous than that.

 

It's ridiculous? If the smaller the polygonal objects are, the better the performance, then game designers would scale their 3D models way, way down, and the game designer would just have to place the camera closer to the models.

Two completely different things, bud.

What you're saying makes no sense. Unlike in the physical world, "size" in the virtual world is relative, not absolute. They are relative to the camera.

 

If all the game's polygonal objects become small, and then designers place the camera really close, then the polygonal objects *are no longer small*, because their "smallness" or "bigness" only has meaning in relation to the other elements in the engine (most notably, the camera).

 

Maybe an illustration will help:

 

*I'm* saying I have a game engine like this: XXXXXXXXXXXX

and I wonder whether there'll be an improvement if I do this to it: XXXXxXXXxxXx

 

*You're* saying, "no, because if you take XXXXXXXXX

and turn it into xxxxxxxxx

but then magnify it so that it becomes XXXXXXXXX

then you're back where you're started, and you've made no difference.

 

We agree there, but what I'm talking about and what you're talking about are actually quite different things, even though it might not initially seem that way. See my above points about SSAO, AA, shadowing and lighting to see why it's more complicated than you make it out to be.

 

Anyway, I'm not discussing this with you any more. If you're not convinced, let's just say that I'm wrong and you're right and move on.

Link to comment
Share on other sites

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...