Simple sprite routines
For a lot of advanced drawing routines, we recommend you to download the ExtGraph library by the TI-Chess Team.
| void Sprite8 (short x, short y, short height, const unsigned char *sprite, void *vm_addr, short mode); |
Draws a sprite with a width of 8 pixels or less on the screen.
Sprite8 draws a sprite with a width of 8 pixels or less on the screen.
Use Sprite16 or Sprite32 for wider sprites.
This routine is many times faster than TIOS routines like
DrawIcon, BitmapPut, etc.
x and y are the coordinates of the upper left corner of the sprite.
height is the height of the sprite. sprite is a pointer to the array of
unsigned characters which define the shape of the sprite (line by line). vm_addr is the pointer
to the video plane; it should be LCD_MEM if you are not using
grayscale or the PortSet function. mode is the drawing
mode, and it may have one of the following values (these constants are defined in the enum
SprtModes):
| SPRT_XOR | XOR the sprite into a background. This is used only for non-masked sprites. XOR mode switches every pixel with a corresponding '1' bit in the sprite array from white to black and verse vica. |
| SPRT_OR | OR the sprite into a background. This is mainly used for masked sprites together with SPRT_AND. OR means that every pixel with a corresponding '1' bit in the sprite array will be set to black, but all other pixels stay the same. If you want to turn all other pixels to white instead, one way to achieve this would be to AND the sprite as well. |
| SPRT_AND | AND the sprite into a background. This is used for masked sprites together with SPRT_OR. AND turns off every pixel with a corresponding '0' bit in the sprite array, but all other pixels remain untouched. |
Non-masked sprites. Using this method, the sprites are simply XORed into the background. This technique was popular in many games on old 8-bit computers. To erase the sprite, it is enough to call the routine again on the same position.
Masked sprites. This is the more advanced method, which needs more programming practice, but produces much better results. Using this method, you first need to erase a part of the background which occupies a space where the sprite need to be displayed, then to draw the actual sprite shape. To do this, AND the inverse of the sprite mask to the background, then OR the sprite at the same location. However, it is not so easy to restore the background later. For solving this problem, a lot of advanced methods are developed (like "double-buffering", etc.). These methods are quite common in advanced ASM games, and they are explained in many ASM tutorials.
Here is a simple example (called "Masked Sprite"), which first draws a simple "background", then draws a "masked" sprite (which is a simple 8x8 square with solid edges and blank interior) at (30,30):
// Display a masked sprite over an arbitrary background
#define USE_TI89 // Compile for TI-89
#define USE_TI92PLUS // Compile for TI-92 Plus
#define USE_V200 // Compile for V200
#define OPTIMIZE_ROM_CALLS // Use ROM Call Optimization
#define MIN_AMS 100 // Compile for AMS 1.00 or higher
#define SAVE_SCREEN // Save/Restore LCD Contents
#include <tigcclib.h> // Include All Header Files
// Main Function
void _main(void)
{
static const unsigned char sprite[] = {0xFF,0x81,0x81,0x81,0x81,0x81,0x81,0xFF};
static const unsigned char imask[] = {(unsigned char)~0xFF,(unsigned char)~0xFF,
(unsigned char)~0xFF,(unsigned char)~0xFF,(unsigned char)~0xFF,
(unsigned char)~0xFF,(unsigned char)~0xFF,(unsigned char)~0xFF};
int i;
ClrScr ();
for (i = 0; i <= LCD_WIDTH; i++)
DrawLine (i, 0, i, LCD_HEIGHT, A_SHADE_NS); // A simple background
Sprite8 (30, 30, 8, imask, LCD_MEM, SPRT_AND);
Sprite8 (30, 30, 8, sprite, LCD_MEM, SPRT_OR);
ngetchx ();
}
Here the sprite mask is {0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF}, but it needs to
be inverted before passing it to the Sprite8 function. For this purpose, the operator '~' may be
very useful. Note that '~' is "bitwise NOT". Of course, ~0xFF is the same as 0x00, but this notation
makes the program more clear (and it does not increase the code size, because the inverting will
be performed at compilation time). And if you want to use Sprite16 or
Sprite32, the notation ~0xFF will still be valid in a short int array, or in a long int array
if you add the 'L' suffix (see the respective info about Sprite32).
Without this notation, you must use 0x00 in Sprite8, but 0xFF00 in Sprite16, and
0xFFFFFF00 in Sprite32. This is why a notation like ~0xFF is more universal.
Starting from TIGCC v0.91, you can use binary numbers to define your sprites. On the one hand,
it makes the program incompatible with other C dialects as well as previous (ancient) versions of
TIGCC. On the other hand, it makes designing a sprite much easier
and also allows for editing the sprite at a later time without having to reconvert it.
The binary representation of the sprite presented above would look like this:
static const unsigned char sprite[]={
0b11111111,
0b10000001,
0b10000001,
0b10000001,
0b10000001,
0b10000001,
0b10000001,
0b11111111};
| void Sprite16 (short x, short y, short height, const unsigned short *sprite, void *vm_addr, short mode); |
Draws a sprite with a width of 16 pixels or less on the screen.
Sprite16 works exactly like Sprite8, but it takes sprites with a width of 16 pixels. sprite is now a pointer to the array of unsigned short integers which defines the sprite. So, to define a sprite (or sprite mask), use something like
static const unsigned short sprite[] = {...};
See Sprite8 for more info about sprites.
| void Sprite32 (short x, short y, short height, const unsigned long *sprite, void *vm_addr, short mode); |
Draws a sprite with a width of 32 pixels or less on the screen.
Sprite32 works exactly like Sprite8, but it takes sprites with a width of 32 pixels. sprite is now a pointer to the array of unsigned long integers which defines the sprite. So, to define a sprite (or sprite mask), use something like
static const unsigned long sprite[] = {...};
Note: If you define a sprite mask and invert it with the '~' operator, you need to take
care of the possibility of the first two bytes being zero in one constant. In this case, the constant will
be assumed to be a 16 bit value, so only 16 bits will be inverted. Then the constant is promoted to an
unsigned long integer, and the first 16 bits are still zero as if they actually belonged to the mask.
To take care of this problem, add the 'L' (long) suffix to the end of the value.
If you want to use sprites wider than 32 pixels (which is not very likely), one solution is to
use DoorsOS and its "put_sprite" function (see Frequently Asked Questions
for more info about how to do it). If you don't want to use DoorsOS, the proposed method depends
on what the use of the sprite will be. If you don't need too fast action, the built-in TIOS function
BitmapPut may be good enough. If you need a very fast sprite
routine for very large sprites, then you must write it yourself.
See Sprite8 for more info about sprites.
| enum SprtModes {SPRT_XOR, SPRT_OR, SPRT_AND}; |
An enumeration to describe possible modes of sprite drawing.
This type is used by all sprite functions.