Optimizing for Spritesheets bug?

Hi,

I read everything about this Topic (https://www.scirra.com/blog/87/under-th ... onstruct-2 and https://www.scirra.com/blog/66/image-co ... onstruct-2), yet there is something unexplained to me.

Why is it, that I have Sprites like 32x32 or 64x64, but construct 2 does not glue them together like that, rather adding 1 px extra around the picture, causing the sprite beeing to big for 2n. Only, if I have Sprites like 30x30 or 62x62 it puts spritesheets toghether in the intended size.

So basicly, construct seems to add a "bleed" of 1 px around the sprites. Why is that? Or how to deal with it properly, since my sprites ARE optimized.

Is this a bug? Or do I misunderstand something?

Thanks for any hint in advance!

This has been brought up a few times in relation to the "most efficient" size question. In the case of separate images right next to each other on the same sheet being drawn, it may be the case that the edges of one image bleed over onto the edges of an image being drawn. For this reason, a 1 pixel buffer is added to prevent these edge artifacts, and the most efficient size is thus 2 less in each dimension than the nearest power of two size.

its not a bug.

I do believe it would be neat if this was disabled if the sampling is set to point or, better yet, by choice. It's not always a good thing to have.

I understand, it is just strange, I never saw anything of that mentioned in all those documentations relating (optimal) sprite sizes (and I have read far more, than just those links mentioned above). I understood everywhere it would be best to use power-of-two-sized images. Now I have to resize a lot

Thanks for your feedback guys!