Name

    APPLE_ycbcr_422

Name Strings

    GL_APPLE_ycbcr_422

Contact

    Geoff Stahl, Apple (gstahl 'at' apple.com)

Status

    Shipping as of August 24, 2002 (Mac OS X v10.2)

Version

    $Date: 2002/09/19 00:01:25 $ $Revision: 1.6 $
	
Number

    275

Dependencies

    OpenGL 1.1 is required
    APPLE_packed_pixels or OpenGL 1.2 is required
    Written against OpenGL 1.2.1

Overview

    This extension provides a method for GL to read, store and optionally
    process textures that are defined in Y'CbCr 422 video formats.  This
    extension supports the two common Y'CbCr 422 video formats (known by
    QuickTime FourCC as '2vuy' and 'yuvs'). These formats represent one of the
    most common 16 bit Y'CbCr formats in both standard and reverse byte
    ordering. From a client stand point these can be assumed to be decoded
    immediately (even though the implementation is free to optimize the data
    storage and keep it in the native format) and otherwise function as any
    other texture format.  The texture command <internalformat> parameter
    normally be should be specified as RGB, since Y'CbCr is just a form of RGB
    data.  This extension can be supported with either hardware or software
    decoding and it is up to the specific implementation to determine which is
    used.

    A new <format> is added, YCBCR_422_APPLE.  Additionally, to handle the
    difference in pixel size and byte ordering for 422 video, the pixel storage
    operations treat YCBCR_422_APPLE as a 2 component format using
    the UNSIGNED_SHORT_8_8_APPLE or UNSIGNED_SHORT_8_8_REV_APPLE <type>.

    The '2vuy' or k2vuyPixelFormat pixel format is an 8-bit 4:2:2 Component
    Y'CbCr format. Each 16 bit pixel is represented by an unsigned eight bit
    luminance component and two unsigned eight bit chroma components. Each pair
    of pixels shares a common set of chroma values. The components are ordered
    in memory; Cb, Y0, Cr, Y1. The luminance components have a range of [16,
    235], while the chroma value has a range of [16, 240]. This is consistent
    with the CCIR601 spec. This format is fairly prevalent on both Mac and Win32
    platforms. The equivalent Microsoft fourCC is UYVY.  This format is
    supported with the UNSIGNED_SHORT_8_8_REV_APPLE type for pixel storage
    operations.

    The 'yuvs' or kYUVSPixelFormat is an 8-bit 4:2:2 Component Y'CbCr format.
    Identical to the k2vuyPixelFormat except each 16 bit word has been byte
    swapped. This results in a component ordering of; Y0, Cb, Y1, Cr. This is
    most prevalent yuv 4:2:2 format on both Mac and Win32 platforms. The
    equivalent Microsoft fourCC is 'YUY2'.  This format is supported with the
    UNSIGNED_SHORT_8_8_APPLE type for pixel storage operations.

Issues

	Why is YCRCR_422 not provided as an <internalformat>?

		The internalFormat parameter passes two distinct pieces of information:
		which one of the six texEnv equations to use *and* what the desired
		internal storage format is.  All of the existing internal format enums
		have one of {RGBA, RGB, L, LA, A, I} embedded in the name to specify
		which of the six texEnv equations.  Since the YUV data contains RGB
		information, only the RGB texEnv setting is meaningful.  Thus, if we did
		provide a new internal format enum it would have to be something of the
		form GL_RGB_YCRCR422 (weird, but has the right meaning).  Using
		YCRCR_422 as an internalFormat setting would be incorrect with respect
		to the texEnv equations.

New Procedures and Functions

	None

New Tokens

    Accepted by the <format> parameter of DrawPixels, ReadPixels, TexImage1D,
    TexImage2D, GetTexImage, TexImage3D, TexSubImage1D, TexSubImage2D,
    TexSubImage3D, GetHistogram, GetMinmax, ConvolutionFilter1D,
    ConvolutionFilter2D, ConvolutionFilter3D, GetConvolutionFilter,
    SeparableFilter2D, SeparableFilter3D, GetSeparableFilter, ColorTable,
    GetColorTable:
    
	  YCBCR_422_APPLE                 0x85B9

    Accepted by the <type> parameter of DrawPixels, ReadPixels, TexImage1D,
    TexImage2D, GetTexImage, TexImage3D, TexSubImage1D, TexSubImage2D,
    TexSubImage3D, GetHistogram, GetMinmax, ConvolutionFilter1D,
    ConvolutionFilter2D, ConvolutionFilter3D, GetConvolutionFilter,
    SeparableFilter2D, SeparableFilter3D, GetSeparableFilter, ColorTable,
    GetColorTable:
    
	  UNSIGNED_SHORT_8_8_APPLE        0x85BA
	  UNSIGNED_SHORT_8_8_REV_APPLE    0x85BB

Additions to Chapter 3 of the OpenGL 1.2.1 Specification (Rasterization)

    Two entries are added to table 3.5 (DrawPixels and ReadPixels type parameter
    values and the corresponding OpenGL data types):
    
	type Parameter                Corresponding         Special
	  Token Name                  GL Data Type       Interpretation
	--------------                -------------      --------------
	UNSIGNED_SHORT_8_8_APPLE         ushort               Yes
	UNSIGNED_SHORT_8_8_REV_APPLE     ushort               Yes


    One entry is added to table 3.6 (DrawPixels and ReadPixels formats):
									
	Format Name			Element Meaning and Order 		Target Buffer
	-----------		    -------------------------		-------------
	YCBCR_422_APPLE		   Y luminance value,               Color
						  [Cb,Cr] chroma value	           
						  	

    Two entries are added to table 3.8 (Packed pixel formats):
    
	type Parameter                 GL Data    Number of        Matching
	  Token Name                    Type      Components     Pixel Formats
	--------------                 -------    ----------     -------------
	UNSIGNED_SHORT_8_8_APPLE       ushort         3          YCBCR_422_APPLE
	UNSIGNED_SHORT_8_8_REV_APPLE   ushort         3          YCBCR_422_APPLE
	
	
    Two entries are added to table 3.10 (UNSIGNED SHORT formats):
    
	UNSIGNED_SHORT_8_8_APPLE:

          15  14  13  12  11  10  9   8   7   6   5   4   3   2   1   0
        +-------------------------------+-------------------------------+
        |              1st              |              2nd              |
        +-------------------------------+-------------------------------+
                        

	UNSIGNED_SHORT_8_8_REV_APPLE:

          15  14  13  12  11  10  9   8   7   6   5   4   3   2   1   0
        +-------------------------------+-------------------------------+
        |              2nd              |              1st              |
        +-------------------------------+-------------------------------+


    One entry is added to table 3.12 (Packed pixel field assignments):

                        First      Second       Third      Fourth
	Format             Element     Element     Element     Element
	------             -------     -------     -------     -------
	YCBCR_422_APPLE   luminance    chroma
    

	The new format YCBCR_422_APPLE is added to the discussion of Conversion to
	RGB: 

	If the format is YCBCR_422_APPLE, the chroma and luminance values in each
	group are converted to R, G, and B values using an undefined algorithm. This
	conversion does not necessarily occur immediately as implementations are
	free to pass Y'CbCr 422 formated pixels directly to hardware that is capable
	of processing it. From a client stand point it can be assumed any
	optimization will be transparently applied and not affect rendering results.
	Pixel transfer operations will likely force conversion to RGB and will
	likely negate hardware Y'CbCr acceleration.  Additionally, if the format is
	YCBCR_422_APPLE, the conversion algorithm may produce undefined RGB values
	for final pixel of any row where the row length is not a multiple of 2.

Additions to Chapter 4 of the OpenGL 1.2.1 Specification (Per-Fragment Operations
and the Framebuffer)

    Add after last paragraph of Final Conversion:
    
    For an RGBA color and if the <type> is UNSIGNED_SHORT_8_8_APPLE or
    UNSIGNED_SHORT_8_8_REV_APPLE, the conversion to Y'CbCr occurs via an
    undefined reverse component conversion.  The actual equation used may vary
    per implementation.  If the row length is odd the final pixel maybe defined
    otherwise the conversion to the requested Y'CbCr output <type> is in all
    ways the same and any other <type>.
        
Additions to the GLX Specification

    None

GLX Protocol

    None

Errors

    None

New State
	
	None

New Implementation Dependent State

    None

Revision History

	None

