Mirrorlandia_minetest/minetest
1
0
Fork 1
mirror of https://github.com/minetest/minetest.git synced 2025-01-08 22:37:32 +01:00
Code Issues Packages Projects Releases Wiki Activity

Correctly render transparency in TGA type 1 with color format A1R5G5B5

Browse Source 
The branch removed in this patch handled color format A1R5G5B5 specially
when creating a texture from a TGA type 1 file, i.e. an image that has a
colormap. It did not handle the 1-bit alpha channel correctly, rendering
transparent pixels black instead.

Since the colormap is converted to A8R8G8B8 earlier anyways, the code
for the general case is able to handle this scenario already – at the
expense of making the created texture use twice as much GPU memory as
necessary.
This commit is contained in:
Ælla Chiana Moskopp 2024-11-05 22:07:24 +01:00 committed by sfan5
parent 1fa4ca7c59
commit 60cd1e4529
1 changed files with 10 additions and 21 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

31
irr/src/CImageLoaderTGA.cpp
View File

@ -181,27 +181,16 @@ IImage *CImageLoaderTGA::loadImage(io::IReadFile *file) const
header.ImageWidth, header.ImageHeight,
0, 0, (header.ImageDescriptor & 0x20) == 0);
} else {
switch (header.ColorMapEntrySize) {
case 16:
image = new CImage(ECF_A1R5G5B5, core::dimension2d<u32>(header.ImageWidth, header.ImageHeight));
if (image)
CColorConverter::convert8BitTo16Bit((u8 *)data,
(s16 *)image->getData(),
header.ImageWidth, header.ImageHeight,
(s32 *)palette, 0,
(header.ImageDescriptor & 0x20) == 0);
break;
// Note: 24 bit with palette would need a 24 bit palette, too lazy doing that now (textures will prefer 32-bit later anyway)
default:
image = new CImage(ECF_A8R8G8B8, core::dimension2d<u32>(header.ImageWidth, header.ImageHeight));
if (image)
CColorConverter::convert8BitTo32Bit((u8 *)data,
(u8 *)image->getData(),
header.ImageWidth, header.ImageHeight,
(u8 *)palette, 0,
(header.ImageDescriptor & 0x20) == 0);
break;
}
// Colormap is converted to A8R8G8B8 at this point – thus the code can handle all color formats.
// This wastes some texture memory, but is less of a third of the code that does this optimally.
// If you want to refactor this: The possible color formats here are A1R5G5B5, B8G8R8, B8G8R8A8.
image = new CImage(ECF_A8R8G8B8, core::dimension2d<u32>(header.ImageWidth, header.ImageHeight));
if (image)
CColorConverter::convert8BitTo32Bit((u8 *)data,
(u8 *)image->getData(),
header.ImageWidth, header.ImageHeight,
(u8 *)palette, 0,
(header.ImageDescriptor & 0x20) == 0);
}
} break;
case 16: