Hallo,

I have a problem with bmap_process.

I get, that I can only use bitmaps sizes that are a power of two.

But if I process a bitmap with 2048 (my calculator says it is 2^11 ergo a power of two) the texture coordinates are messed up.

I wrote this testshader:

Code:
float4 process_test(in float2 tex: TEXCOORD0): COLOR 
{	
	return float4(tex.x, tex.y, tex.x, tex.y);
}


And wrote the result per bmap_process into a 32bit floating point bitmap.

Code:
BMAP *test = bmap_createblack(2048, 2048, 14);
	
bmap_process(test, test, mtlProjectText);
	
bmap_to_format(test, 14);
	
bmap_lock(test, 0);
	
float *bits = test.finalbits;
	
str_for_float(deb1, bits[2047]);
	
bmap_unlock(test);


result is 0.8...

if the bitmap is 1024 result is 0.99....


So why?

Last edited by jenGs; 08/14/14 11:07.