ImageGear versions prior to 14.5 used the standard Windows DIB format for internal storage of images. This format is made up of three components which are stored sequentially:

1. Header. This is a BITMAPINFOHEADER / AT_DIB struct containing the following information about a DIB:

struct tagAT_DIB
{
    AT_DWORD    biSize;
    AT_INT32    biWidth;
    AT_INT32    biHeight;
    AT_WORD     biPlanes;
    AT_WORD     biBitCount;
    AT_DWORD    biCompression;
    AT_DWORD    biSizeImage;
    AT_INT32    biXPelsPerMeter;
    AT_INT32    biYPelsPerMeter;
    AT_DWORD    biClrUsed;
    AT_DWORD    biClrImportant;
};
typedef struct tagAT_DIB AT_DIB;
typedef struct tagAT_DIB FAR *LPAT_DIB;

The removed function IG_image_DIB_ptr_get retrieved a pointer to this struct.

1. Palette. This is an array of RGBQUAD entries containing the colors that correspond to pixel values in an indexed image. The number of entries in this array is determined by the bit depth of the image (biBitCount). A palette is only stored for indexed images.

The function IG_image_DIB_palette_pntr_get retrieves a pointer to this array. This function still exists in ImageGear 14.5, but the DIB header and pixel data no longer surround the palette in memory.

1. Pixel data. This is the pixel data that actually makes up the image. The image rows are stored bottom-up and each row is DWORD-aligned. In the case of a 24-bit RGB image, the pixels are stored in BGR channel order.