4 replies to this topic

[loverock125]

Heya everyone,

This is something that, even though it may be simple, I can't get it clear enough inside my head. As I was trying to create a huge 3D world while playing around with different dimensions of a 3D Terrain Model, this question just hit me. Is there really any difference between having a huge world with huge models, and a small world with smaller models?

To better illustrate my point, take a look at the following screenshots (Inside the spoilers).

Spoiler

Spoiler

They look very identical and pretty much the same. The only difference is that the terrain in the second image is by 10 times bigger than the first one.
Even when I created a tiny terrain (16x16 pixels), when I zoomed in pretty close nobody would notice its small size since you can't tell the analogies.

Considering of course that the number of polygons of the model are exactly the same, does the size of the models affect performance? Or does it affect the quality of the texture of the terrain? (Since I assume the texture gets stretched according to the size of the model)

This may sound very simple, but I would really like some clarification.

Thank you very much, I appreciate your help

[darkai]

The difference is a matter of performance vs quality.
Zooming in with a small texture will lag less but look more fuzzy.
Zoomed out high quality textures will look great but possibly cause lag.

[icuurd12b42]

stick with 1 pixel = 1 meter. Or 1 pixel = 1 feet. (it's not a pixel BTW)


The smaller the size I think the more difficult time the 3d system will have the with zbuffer and you would oddities where models intersect.

[loverock125]

stick with 1 pixel = 1 meter. Or 1 pixel = 1 feet. (it's not a pixel BTW)


The smaller the size I think the more difficult time the 3d system will have the with zbuffer and you would oddities where models intersect.

That's exactly the answer I was looking for. Thank you : )

[Yourself]

stick with 1 pixel = 1 meter. Or 1 pixel = 1 feet. (it's not a pixel BTW)


The smaller the size I think the more difficult time the 3d system will have the with zbuffer and you would oddities where models intersect.

That's exactly the answer I was looking for. Thank you : )

Mathematically speaking there is no difference, but since the GPU uses floating point types to represent positions, using units on the order of 1 is recommended, since 1 is exactly in the middle of the floating point types. Half of all values that can be represented are between 0-1. The other half is greater than 1.