xref: /third_party/openGLES/extensions/NV/NV_texture_compression_s3tc_update.txt

Name

    NV_texture_compression_s3tc_update

Name Strings

    GL_NV_texture_compression_s3tc_update

Contact

    Ian Stewart, NVIDIA Corporation (istewart 'at' nvidia.com)

Contributors

    Nicholas Haemel, NVIDIA Corporation
    Acorn Pooley, NVIDIA Corporation
    Antti Rasmus, NVIDIA Corporation
    Musawir Shah, NVIDIA Corporation

Status

    Complete.

Version

    Last Modified Date: Sep 27, 2012
    NVIDIA Revision: 2

Number

    OpenGL ES Extension #95

Dependencies

    This specification is written against the OpenGL ES 2.0.25
    specification.

    Depends on the GL_NV_texture_compression_s3tc extension.

Overview

    This extension allows for full or partial image updates to a
    compressed 2D texture from an uncompressed texel data buffer using
    TexImage2D and TexSubImage2D. Consquently, if a compressed internal
    format is used, all the restrictions associated with compressed
    textures will apply. These include sub-image updates aligned to 4x4
    pixel blocks and the restriction on usage as render targets.

IP Status

    NVIDIA Proprietary

New Procedures and Functions

    None

New Tokens

    None

Additions to Chapter 3 of the OpenGL ES 2.0.25 Specification (Rasterization)

    Modify Section 3.7.1, Texture Image Specification

    (change last paragraph on Page 67 as follows)

    Components are then selected from the resulting R, G, B, or A values
    to obtain a texture with the base internal format specified by
    <internalformat>, which must match <format> except when <target> is
    TEXTURE_2D and <internalformat> is one of the following compressed
    formats: COMPRESSED_RGB_S3TC_DXT1_NV,
    COMPRESSED_RGBA_S3TC_DXT1_NV, COMPRESSED_RGBA_S3TC_DXT3_NV, or
    COMPRESSED_RGBA_S3TC_DXT5_NV. In this case, conversion from only
    RGB and RGBA formats are supported during texture image processing.
    <format> values other than RBA or RGBA will result in the
    INVALID_OPERATION error. In all other cases where <internalformat>
    does not match <format>, the error INVALID_OPERATION is generated.
    Table 3.8 summarizes the mapping of R, G, B, and A values to texture
    components, as a function of the base internal format of the texture
    image. <internalformat> may be one of the five internal format
    symbolic constants listed in table 3.8 or the four compressed
    formats: COMPRESSED_RGB_S3TC_DXT1_NV,
    COMPRESSED_RGBA_S3TC_DXT1_NV, COMPRESSED_RGBA_S3TC_DXT3_NV, or
    COMPRESSED_RGBA_S3TC_DXT5_NV. Specifying a value for
    <internalformat> that is not one of the above values generates the
    error INVALID_VALUE. When a compressed <internalformat> is
    specified, a compressed texture is created and all the associated
    restrictions mentioned in Section 3.7.3 are imposed.

    Note that when encoding an RGBA image into a format using 1-bit
    alpha, any texels with an alpha component less than 0.5 end up
    with an alpha of 0.0 and any texels with an alpha component
    greater than or equal to 0.5 end up with an alpha of 1.0. When
    encoding an RGBA image into the COMPRESSED_RGBA_S3TC_DXT1_NV
    format, the resulting red, green, and blue components of any
    texels with a final alpha of 0.0 will automatically be zero
    (black).  If this behavior is not desired by an application, it
    should not use COMPRESSED_RGBA_S3TC_DXT1_NV.

    Modify Section 3.7.2, Alternate Texture Image Specification Commands

    (add to the end of section)

    When the internal format of the texture object is
    COMPRESSED_RGB_S3TC_DXT1_NV, COMPRESSED_RGBA_S3TC_DXT1_NV,
    COMPRESSED_RGBA_S3TC_DXT3_NV, or COMPRESSED_RGBA_S3TC_DXT5_NV, the
    update region specified in TexSubImage2D must be aligned to 4x4
    pixel blocks. If <xoffset> or <yoffset> are not multiples of 4 an
    INVALID_OPERATION error is generated. If <width> is not a multiple
    of 4 and <xoffset> + <width> is not equal to the width of the LOD
    then an INVALID_OPERATION error is generated.  If <height> is not
    a multiple of 4 and <yoffset> + <height> is not equal to the
    height of the LOD then an INVALID_OPERATION error is generated.

Additions to Chapter 4 of the OpenGL ES 2.0.25 Specification (Per-Fragment
Operations and the Framebuffer)

    Modify Section 4.4.3, Attaching Texture Images to a Framebuffer

    (add after last paragraph on Page 113)

    If <texture> is not zero and the internal format of the
    corresponding texture object is a compressed format, an
    INVALID_OPERATION error is generated.

Errors

    INVALID_OPERATION is generated by TexImage2D and TexSubImage2D if
    TEXTURE_INTERNAL_FORMAT is COMPRESSED_RGB_S3TC_DXT1_NV,
    COMPRESSED_RGBA_S3TC_DXT1_NV, COMPRESSED_RGBA_S3TC_DXT3_NV, or
    COMPRESSED_RGBA_S3TC_DXT5_NV and <format> is not RGB or RGBA.

    If <xoffset> or <yoffset> are not multiples of 4 an
    INVALID_OPERATION error is generated.

    If <width> is not a multiple of 4 and <xoffset> + <width> is not
    equal to the width of the LOD then an INVALID_OPERATION error is
    generated.

    If <height> is not a multiple of 4 and <yoffset> + <height> is not
    equal to the height of the LOD then an INVALID_OPERATION error is
    generated.

Issues

    1) Should sub-image updates be aligned to 4x4 pixel blocks?

    RESOLUTION: Yes. Relaxing this restriction would otherwise require
    CPU read-back of pixels at the 4x4 compression block boundary.

    2) Should we support binding of compressed textures as render targets?

    RESOLUTION: No. Writing to directly to compressed formats not
    possible using hardware.  Supporting this feature would involve
    read-back of the texture and re-compression on the CPU, which would
    be prohibitively slow.

    3) Should we support mip map generation?

    RESOLUTION: No. Mip map generation is not supported for compressed
    textures. However, uploading to specific mip level is supported.

Revision History

    Rev.    Date      Author       Changes
    ----   --------   ---------    -------------------------------------
     2     09/27/12   istewart     Changed all EXT references to NV.
                                   Added note about DXT1 alpha encoding.
                                   Fixed TexSubImage2D error conditions.
     1     04/06/11   mshah        First revision.