EN:NCER Format

These files are used to build an image using some part of a big image, like making a puzzle. It's usually used to obtain the sprites of NCGR file. In Tinke they have the file format cell.

They are recognized by their header NCER or RECN (Nitro CEll Resource) and they usually have the same extension.

Structure

The file has the following structure:

• Generic header
• Section CEBK
  • Banks
    • Cells
• Section LBAL
• Section UEXT

Generic header

Offset	Name	Type	Description
0x00	ID	char[4]	NCER or RECN
0x04	Endianess	ushort	If 0xFFFE the Magic ID must be reversed
0x06	Constant	ushort	Unknown
0x08	File size	uint	File size
0x0C	Header size	ushort	Always 0x10
0x0E	Number of sections	ushort	It can be 0x01, 0x02 or 0x03

Section CEBK

Offset	Name	Type	Description
0x00	ID	char[4]	CEBK
0x04	Size	uint
0x08	Nunmber of banks	ushort
0x0A	Type of banks	ushort
0x0C	Unknown	uint	It should be 0x18 (offset?)
0x10	Block shift	uint	Bit 0-7 -> "image offset" << "block shift"
0x14	Unknown	uint	If it's different to 0, you must add "tilepos" to the "image offset"
0x18	Unknown	ulong
0x20	Banks	Banks[]

Banks

Offset	Name	Type	Description
0x00	Number of cells	ushort
0x02	Unknown	ushort
0x04	Cells offset	ushort	Offset of the cell info

if the bank is type 1, there are the following extra information:

Offset	Name	Type	Description
0x06	X maximum	ushort	Position X maximum
0x08	Y maximum	ushort	Position Y maximum
0x0A	X minimum	ushort	Position X minimum
0x0C	Y minimum	ushort	Position Y minimum

Cells

Offset	Name	Type	Description
0x00	Attribute 0	ushort	See below
0x02	Attribute 1	ushort	See below
0x04	Attribute 2	ushort	See below

More info: gbatek

• Attribute 0:

SS D M OO P R YYYYYYYY
 
S -> (shape) 0=square, 1=horizontal, 2=vertical, 3=prohibited
D -> (colors/palettes) 0=16/16, 1=256/1
M -> (mosaic) 0=off, 1=on
O -> (mode) 0=normal, 1=semi-transparent, 2=obj window, 3=prohibited
P -> if R=1 => (double size flag) 0=normal, 1=double
     if R=0 => (obj disabled) 0=normal, 1=not displayed
R -> (Rotation or scale flag) 0=off, 1=on
Y -> Coordinate Y

• Attribute 1:

SS PPPPP XXXXXXXXX

S -> (size) see the table below
P -> if in the attribute 0:
     R=0 => 9-11 ->(not used)
	        12 -> (horizontal flip) 0=normal, 1=mirrored
			13 -> (vertical flip) 0=normal, 1=mirrored
	 R=1 => 9-13 -> rotation or scale parameters
X -> Coordinate X (if it's bigger or equal to 0x100, you must substract 0x200)

• Attribute 2:

IIII PP TTTTTTTTTT

I -> Number of the palette (not used in mode 256/1)
P -> Priority relative to background (0 high priority)
T -> Offset of the image tiles

Size table:

Size	Square	Horizontal	Vertical
0	8x8	16x8	8x16
1	16x16	32x8	8x32
2	32x32	32x16	16x32
3	64x64	64x32	32x64

Section LABL

Offset	Name	Type	Description
0x00	ID	char[4]	LABL
0x04	Size	uint	Section size
0x08	Pointers	uint[]	Offset to the names, relative to this section
....	Bank names	string[]

Section UEXT

Offset	Name	Type	Description
0x00	ID	char[4]	UEXT
0x04	Size	uint	Section size
0x08	Unknown	uint

Notes

1.- Offset 0x14 in CEBK If the variable is different to 0, then in each bank the first cell will start with 0 in the "tile offset", so in order to show get the correct offset, you must add the offset of the last bank (the last offset plus the size of this cell).

2.- Block shift in CEBK

• If it's less than 4, you must displace (bitwise operation) to the left the "block shift" value. For instance, if the offset is 0x2F and the "block shift" is 3, then the final offset will be: 0x2F << 3 = 0x178

• If it's greater or equal to 4 the image is already displayed correctly, so we only need to cut it and display this portion in the cell. In order to know the coordinate X and Y, where you must cut the image, you can use the following code (C#): Basically, the "tile offset" in this case contains the number of pixel horinzontally that you must read to get the correct position, that is, if the "tile offset" is bigger than the image width, then you must add 1 to the Y coordinate

tileOffset /= (blockSize / 2);
int imageWidth = tile.rahc.nTilesX;
int imageHeight = tile.rahc.nTilesY;
if (tile.order == TileOrder.Horizontal)
{
   imageWidth *= 8;
   imageHeight *= 8;
}

int posX = (int)(tileOffset % imageWidth);
int posY = (int)(tileOffset / imageWidth);

if (tile.rahc.depth == System.Windows.Forms.ColorDepth.Depth4Bit)
   posY *= (int)blockSize * 2;
else
   posY *= (int)blockSize;
if (posY >= imageHeight)
   posY = posY % imageHeight;

3.- Size If the image is "No tile" then you must multiply by 8 the size of each cell.