m/grafX2
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Sped up fast 24bit color reduction (Caanoo specific)

Browse Source 
git-svn-id: svn://pulkomandy.tk/GrafX2/trunk@1635 416bcca6-2ee7-4201-b75f-2eb2f807beb1
This commit is contained in:
Yves Rizoud 2010-10-02 13:52:42 +00:00
parent e3e9be75b0
commit b78bae876b
2 changed files with 4 additions and 79 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
src/loadsave.c
View File

@ -980,27 +980,7 @@ void Load_SDL_Image(T_IO_Context *context)
{
{
// Hi/Trucolor
///Sorry, but i give up! It needs a hard debug. So, a simply workaround here. Unfortunately it doubling memory usage by conversion surface :(
#if defined(__WIZ__) || defined(__CAANOO__)
SDL_Surface *pNewSurface;
SDL_PixelFormat pixfmt;
#endif
Pre_load(context, surface->w, surface->h, file_size ,FORMAT_ALL_IMAGES, PIXEL_SIMPLE, 1);
#if defined(__WIZ__) || defined(__CAANOO__)
memcpy(&pixfmt, surface->format, sizeof(SDL_PixelFormat));
pixfmt.BitsPerPixel = 32;
pixfmt.BytesPerPixel = 4;
pNewSurface = SDL_ConvertSurface(surface, &pixfmt, SDL_SWSURFACE);
if (pNewSurface == NULL)
{
File_error=1;
return;
}
SDL_FreeSurface(surface);
surface = pNewSurface;
#endif
}
for (y_pos=0; y_pos<context->Height; y_pos++)

63
src/op_c.c
View File

@ -1396,75 +1396,20 @@ extern void Set_palette_fake_24b(T_Palette palette);
/// Really small, fast and dirty convertor(just for handhelds)
int Convert_24b_bitmap_to_256_fast(T_Bitmap256 dest,T_Bitmap24B source,int width,int height,T_Components * palette)
{
T_Components * palette_fake24b;
int px_pos, py_pos;
int size, size_percent, size_percent_progress;
int delta, delta_old;
int i, idx, progress_counter;
char progress_str[8];
SDL_Rect pw;
int size;
Set_palette_fake_24b(palette);
palette_fake24b = palette;
Open_window(80, 30, "Progress");
pw.x = Window_pos_X;
pw.y = Window_pos_Y;
pw.w = Window_width * Menu_factor_X;
pw.h = Window_height* Menu_factor_Y;
Window_rectangle(pw.x, pw.y, pw.w, pw.h, MC_Light);
px_pos = (pw.w / (2 * Menu_factor_X)) - 10;
py_pos = (pw.h / (2 * Menu_factor_Y)) + 1;
progress_counter = 0;
size = width*height;
size_percent = size / 100;
size_percent_progress = size - 1;
while(size--)
{
//Progress counter update
if(size_percent_progress == size)
{
size_percent_progress = size - size_percent;
progress_counter++;
//Because of integer division it less of resolution and need to be checked for small files
//(maybe you can do a more smart checking or use a float values)
progress_counter = progress_counter > 100 ? 100 : progress_counter;
snprintf(progress_str, sizeof(progress_str), "%d%%", progress_counter);
Print_in_window(px_pos, py_pos, (const char *)progress_str, MC_Black, MC_Light);
SDL_LockSurface(Screen_SDL);
SDL_UpdateRect(Screen_SDL, pw.x, pw.y, pw.w, pw.h);
SDL_UnlockSurface(Screen_SDL);
}
palette = palette_fake24b;
delta = delta_old = INT_MAX;
//Searching for most suitable colour in palette
for(i=idx=0; i<256; i++)
{
//A HUGE field for optimizations lies here
//Compute delta beetween current and palette colours.
//(calling function 3 times in a row is a lame, i know)
delta = abs(source->R - palette->R) + abs(source->G - palette->G) + abs(source->B - palette->B);
if(delta < delta_old)
{
delta_old = delta;
idx = i;
//Most suitable color found - step out from loop
if(delta == 0) break;
}
palette++;
}
//Set palette color index to destination bitmap
*dest = idx;
*dest = ((source->R >> 5) << 5) |
((source->G >> 5) << 2) |
((source->B >> 6));
source++;
dest++;
}
Close_window();
return 0;
}