3ds Max C++ API Reference
Bitmap Types

See also: BitmapStorage, Color, AColor, AColor, LogLUV32Pixel, LogLUV24Pixel, RealPixel. More...

Macros

#define BMM_NO_TYPE   0
 Not allocated yet. More...
 
#define BMM_LINE_ART   1
 1-bit monochrome image More...
 
#define BMM_PALETTED   2
 8-bit paletted image. More...
 
#define BMM_GRAY_8   3
 8-bit grayscale bitmap. More...
 
#define BMM_GRAY_16   4
 16-bit grayscale bitmap. More...
 
#define BMM_TRUE_16   5
 16-bit true color image. More...
 
#define BMM_TRUE_32   6
 32-bit color: 8 bits each for Red, Green, Blue, and Alpha. More...
 
#define BMM_TRUE_64   7
 64-bit color: 16 bits each for Red, Green, Blue, and Alpha. More...
 
#define BMM_LOGLUV_32   13
 This format uses a logarithmic encoding of luminance and U' and V' in the CIE perceptively uniform space. More...
 
#define BMM_LOGLUV_24   14
 This format is similar to BMM_LOGLUV_32 except is uses smaller values to give a span of 5 order of magnitude from 1/4096 to 16 in 1.1% luminance steps. More...
 
#define BMM_LOGLUV_24A   15
 This format is similar to BMM_LOGLUV_24, except the 8 bit alpha value is kept with the 24 bit color value in a single 32 bit word. More...
 
#define BMM_REALPIX_32   16
 The "Real Pixel" format. More...
 
#define BMM_FLOAT_RGBA_32   17
 32-bit floating-point per component (non-compressed), RGB with or without alpha More...
 
#define BMM_FLOAT_GRAY_32   18
 32-bit floating-point (non-compressed), monochrome/grayscale More...
 
#define BMM_TRUE_24   8
 24-bit color: 8 bits each for Red, Green, and Blue. More...
 
#define BMM_TRUE_48   9
 48-bit color: 16 bits each for Red, Green, and Blue. More...
 
#define BMM_YUV_422   10
 This is the YUV format - CCIR 601. More...
 
#define BMM_BMP_4   11
 Windows BMP 16-bit color bitmap. More...
 
#define BMM_PAD_24   12
 Padded 24-bit (in a 32 bit register). More...
 
#define BMM_FLOAT_RGB_32   19
 ONLY returned by the GetStoragePtr() method of BMM_FLOAT_RGBA_32 storage, NOT an actual storage type! When GetStoragePtr() returns this type, the data should be interpreted as three floating-point values, corresponding to Red, Green, and Blue (in this order). More...
 
#define BMM_FLOAT_A_32   20
 ONLY returned by the GetAlphaStoragePtr() method of BMM_FLOAT_RGBA_32 or BMM_FLOAT_GRAY_32 storage, NOT an actual storage type! When GetStorageAlphaPtr() returns this type, the data should be interpreted as floating-point values one value per pixel, corresponding to Alpha. More...
 

Detailed Description

See also: BitmapStorage, Color, AColor, AColor, LogLUV32Pixel, LogLUV24Pixel, RealPixel.

Macro Definition Documentation

◆ BMM_NO_TYPE

#define BMM_NO_TYPE   0

Not allocated yet.

◆ BMM_LINE_ART

#define BMM_LINE_ART   1

1-bit monochrome image

◆ BMM_PALETTED

#define BMM_PALETTED   2

8-bit paletted image.

Each pixel value is an index into the color table.

◆ BMM_GRAY_8

#define BMM_GRAY_8   3

8-bit grayscale bitmap.

◆ BMM_GRAY_16

#define BMM_GRAY_16   4

16-bit grayscale bitmap.

◆ BMM_TRUE_16

#define BMM_TRUE_16   5

16-bit true color image.

◆ BMM_TRUE_32

#define BMM_TRUE_32   6

32-bit color: 8 bits each for Red, Green, Blue, and Alpha.

◆ BMM_TRUE_64

#define BMM_TRUE_64   7

64-bit color: 16 bits each for Red, Green, Blue, and Alpha.

◆ BMM_LOGLUV_32

#define BMM_LOGLUV_32   13

This format uses a logarithmic encoding of luminance and U' and V' in the CIE perceptively uniform space.

It spans 38 orders of magnitude from 5.43571 to 1.84467 in steps of about 0.3% luminance steps. It includes both positive and negative colors. A separate 16 bit channel is kept for alpha values.

◆ BMM_LOGLUV_24

#define BMM_LOGLUV_24   14

This format is similar to BMM_LOGLUV_32 except is uses smaller values to give a span of 5 order of magnitude from 1/4096 to 16 in 1.1% luminance steps.

A separate 8 bit channel is kept for alpha values.

◆ BMM_LOGLUV_24A

#define BMM_LOGLUV_24A   15

This format is similar to BMM_LOGLUV_24, except the 8 bit alpha value is kept with the 24 bit color value in a single 32 bit word.

◆ BMM_REALPIX_32

#define BMM_REALPIX_32   16

The "Real Pixel" format.

◆ BMM_FLOAT_RGBA_32

#define BMM_FLOAT_RGBA_32   17

32-bit floating-point per component (non-compressed), RGB with or without alpha

◆ BMM_FLOAT_GRAY_32

#define BMM_FLOAT_GRAY_32   18

32-bit floating-point (non-compressed), monochrome/grayscale

◆ BMM_TRUE_24

#define BMM_TRUE_24   8

24-bit color: 8 bits each for Red, Green, and Blue.

Cannot be written to.

◆ BMM_TRUE_48

#define BMM_TRUE_48   9

48-bit color: 16 bits each for Red, Green, and Blue.

Cannot be written to.

◆ BMM_YUV_422

#define BMM_YUV_422   10

This is the YUV format - CCIR 601.

Cannot be written to.

◆ BMM_BMP_4

#define BMM_BMP_4   11

Windows BMP 16-bit color bitmap.

Cannot be written to.

◆ BMM_PAD_24

#define BMM_PAD_24   12

Padded 24-bit (in a 32 bit register).

Cannot be written to.

◆ BMM_FLOAT_RGB_32

#define BMM_FLOAT_RGB_32   19

ONLY returned by the GetStoragePtr() method of BMM_FLOAT_RGBA_32 storage, NOT an actual storage type! When GetStoragePtr() returns this type, the data should be interpreted as three floating-point values, corresponding to Red, Green, and Blue (in this order).

◆ BMM_FLOAT_A_32

#define BMM_FLOAT_A_32   20

ONLY returned by the GetAlphaStoragePtr() method of BMM_FLOAT_RGBA_32 or BMM_FLOAT_GRAY_32 storage, NOT an actual storage type! When GetStorageAlphaPtr() returns this type, the data should be interpreted as floating-point values one value per pixel, corresponding to Alpha.