glDrawPixels(3G) OpenGL Reference glDrawPixels(3G)
NAME
glDrawPixels - write a block of pixels to the frame buffer
C SPECIFICATION
void glDrawPixels( GLsizei width,
GLsizei height,
GLenum format,
GLenum type,
const GLvoid *pixels )
PARAMETERS
width, height Specify the dimensions of the pixel rectangle that will be
written into the frame buffer.
format Specifies the format of the pixel data. Symbolic constants
GLCOLORINDEX, GLSTENCILINDEX, GLDEPTHCOMPONENT,
GLRGBA, GLABGREXT, GLRED, GLGREEN, GLBLUE, GLALPHA,
GLRGB, GLLUMINANCE, GLLUMINANCEALPHA and
GLYCRCB422SGIX are accepted.
type Specifies the data type for pixels. Symbolic constants
GLUNSIGNEDBYTE, GLBYTE, GLBITMAP, GLUNSIGNEDSHORT,
GLSHORT, GLUNSIGNEDINT, GLINT, GLFLOAT,
GLUNSIGNEDBYTE332EXT, GLUNSIGNEDSHORT4444EXT,
GLUNSIGNEDSHORT5551EXT, GLUNSIGNEDINT8888EXT,
and GLUNSIGNEDINT1010102EXT are accepted.
pixels Specifies a pointer to the pixel data.
DESCRIPTION
glDrawPixels reads pixel data from memory and writes it into the frame
buffer relative to the current raster position. Use glRasterPos to set
the current raster position, and use glGet with argument
GLCURRENTRASTERPOSITION to query the raster position.
Several parameters define the encoding of pixel data in memory and
control the processing of the pixel data before it is placed in the frame
buffer. These parameters are set with the commands glPixelStore,
glPixelTransfer, glPixelMap, and glPixelZoom. The state of
GLINTERLACESGIX (controlled by glEnable and glDisable) and the state
for the EXTconvolution, SGIcolormatrix, EXThistogram, and
SGIcolortable extensions (see the reference pages for glPixelTransfer,
glConvolutionFilter2DEXT, glColorTableSGI, glHistogramEXT, and
glMinmaxEXT) also affect the results of glDrawPixels. This reference
page describes the effects on glDrawPixels of some, but not all, of these
parameters.
Page 1
glDrawPixels(3G) OpenGL Reference glDrawPixels(3G)
type specifies the data type for pixels. The following table summarizes
the meaning of the valid constants for type:
type corresponding type
_______________________________________________________________________
GLUNSIGNEDBYTE unsigned 8-bit integer
GLBYTE signed 8-bit integer
GLBITMAP single bits in unsigned 8-bit integers
GLUNSIGNEDSHORT unsigned 16-bit integer
GLSHORT signed 16-bit integer
GLUNSIGNEDINT unsigned 32-bit integer
GLINT 32-bit integer
GLFLOAT single-precision floating-point
GLUNSIGNEDBYTE332EXT unsigned 8-bit integer
GLUNSIGNEDSHORT4444EXT unsigned 16-bit integer
GLUNSIGNEDSHORT5551EXT unsigned 16-bit integer
GLUNSIGNEDINT8888EXT unsigned 32-bit integer
GLUNSIGNEDINT1010102EXT unsigned 32-bit integer
If type is GLUNSIGNEDBYTE, GLBYTE, GLUNSIGNEDSHORT, GLSHORT,
GLUNSIGNEDINT, GLINT, or GLFLOAT, then data is read as a sequence of
signed or unsigned bytes, shorts, or integers, or single-precision
floating-point values. Each of these bytes, shorts, integers, or
floating-point values is interpreted as one color component, one depth
component, or one index, depending on format. Indices are always treated
individually. Colors are treated as groups of one, two, three, or four
elements, again based on format. Both individual indices and groups of
components are referred to as pixels.
If type is GLUNSIGNEDBYTE332, GLUNSIGNEDSHORT4444,
GLUNSIGNEDSHORT5551, GLUNSIGNEDINT8888, or
GLUNSIGNEDINT1010102 then all the elements of each group are read
from a single unsigned byte, unsigned short, or unsigned int. The number
of elements per packed pixel is fixed by type, and must match the number
of elements per group indicated by format.
The following table shows which values of format are valid for each of
the packed pixel types:
type number elements format
______________________________________________________________________
GLUNSIGNEDBYTE332EXT 3 GLRGB
GLUNSIGNEDSHORT4444EXT 4 GLRGBA,GLABGREXT
GLUNSIGNEDSHORT5551EXT 4 GLRGBA,GLABGREXT
GLUNSIGNEDINT8888EXT 4 GLRGBA,GLABGREXT
GLUNSIGNEDINT1010102EXT 4 GLRGBA,GLABGREXT
The elements in a packed pixel are ordered such that the first element is
in the most significant bits, followed by the second element, etc. For
example, if type is set to GLUNSIGNEDSHORT4444EXT then element 1
is read from bits 15-12, element 2 is read from bits 11-8, element 3 is
Page 2
glDrawPixels(3G) OpenGL Reference glDrawPixels(3G)
read from bits 7-4 and element 4 is read from bits 3-0.
The assignment of elements to fields in the packed pixel is as described
in the table below:
format element 1 element 2 element 3 element 4
___________________________________________________________
GLRGB red green blue
GLRGBA red green blue alpha
GLABGREXT alpha blue green red
If type is GLBITMAP, the data must be unsigned bytes, and format must be
either GLCOLORINDEX or GLSTENCILINDEX. Each unsigned byte is treated
as eight 1-bit pixels, with bit ordering determined by
GLUNPACKLSBFIRST. If GLUNPACKLSBFIRST is enabled then the 8
single-bit elements are ordered from most significant to least
significant; otherwise the ordering is from least significant to most
significant. (see glPixelStore).
If GLUNPACKSWAPBYTES is enabled (see glPixelStore), byte swapping is
performed as the data is read from memory. For pixels that aren't packed,
the byte ordering for multibyte color components, depth components, color
indices, or stencil indices is reversed. That is, if a four-byte
component is made up of bytes b0, b1, b2, b3, it is taken from memory as
b3, b2, b1, b0. In the case of the packed pixel types, byte swapping is
performed before the elements are extracted from each pixel.
width*height pixels are read from memory, starting at location pixels.
By default, these pixels are taken from adjacent memory locations, except
that after all width pixels are read, the read pointer is advanced to the
next four-byte boundary. The four-byte row alignment is specified by
glPixelStore with argument GLUNPACKALIGNMENT, and it can be set to one,
two, four, or eight bytes. Other pixel store parameters specify
different read pointer advancements, both before the first pixel is read,
and after all width pixels are read. Refer to the glPixelStore reference
page for details on these options.
The width*height pixels that are read from memory are each operated on in
the same way, based on the values of several parameters specified by
glPixelTransfer and glPixelMap. The details of these operations, as well
as the target buffer into which the pixels are drawn, are specific to the
format of the pixels, as specified by format. These operations generate
various fragment values as described below for each format. The number
of fragments generated for each source pixel and the assignment of x and
y window coordinates to these fragments are common to all formats and are
as follows.
If the current raster position is invalid, no fragments are generated for
any of the source pixels. Else, let (x , y ) be the current raster
r r
position, and let n and m be the column and row of the source pixel,
respectively. For each source pixel, fragments with values computed
according to that pixel are generated for destination pixels whose
Page 3
glDrawPixels(3G) OpenGL Reference glDrawPixels(3G)
centers are in the rectangle with corners at
(x + n * zoom , y + m * interlace * zoom ) and
r x r y
(x + (n + 1) * zoom , y + (m * interlace + 1) * zoom ). zoom and
r x r y x
zoom are the values of GLZOOMX and GLZOOMY, respectively, and
y
interlace is 1 or 2 depending on whether GLINTERLACESGIX is GLFALSE or
GLTRUE, respectively. GLZOOMX and GLZOOMY, initially set to 1.0,
are set using glPixelZoom; GLINTERLACESGIX, initially disabled, is
controlled by glEnable and glDisable. See the NOTES section for
additional discussion of GLINTERLACESGIX.
Except for when format is GLSTENCILINDEX the generated fragments are
treated just like the fragments generated by rasterizing points, lines,
or polygons. Texture mapping, fog, and all the fragment operations are
applied before the fragments are written to the frame buffer. For
GLSTENCILINDEX only the pixel ownership test, the scissor test, and the
stencil writemask affect the writes to the stencil buffer.
format can assume the following symbolic values:
GLCOLORINDEX
Each pixel is a single value, a color index. It is converted to
fixed-point format, with an unspecified number of bits to the right
of the binary point, regardless of the memory data type. Floating-
point values convert to true fixed-point values. Signed and
unsigned integer data is converted with all fraction bits set to
zero. Bitmap data convert to either 0.0 or 1.0.
Each fixed-point index is then shifted left by GLINDEXSHIFT bits
and added to GLINDEXOFFSET. If GLINDEXSHIFT is negative, the
shift is to the right. In either case, zero bits fill otherwise
unspecified bit locations in the result.
If the GL is in RGBA mode, the resulting index is converted to an
RGBA pixel using the GLPIXELMAPITOR, GLPIXELMAPITOG,
GLPIXELMAPITOB, and GLPIXELMAPITOA tables. If the GL is
in color index mode, and if GLMAPCOLOR is true, the index is
replaced with the value that it references in lookup table
GLPIXELMAPITOI. Whether the lookup replacement of the index is
b
done or not, the integer part of the index is then ANDed with 2 -1,
where b is the number of bits in a color index buffer.
The resulting index or RGBA color components, and the current raster
position z and texture coordinates are assigned to each of the
fragments generated for the source pixel.
GLSTENCILINDEX
Each pixel is a single value, a stencil index. It is converted to
fixed-point format, with an unspecified number of bits to the right
of the binary point, regardless of the memory data type. Floating-
point values convert to true fixed-point values. Signed and
unsigned integer data is converted with all fraction bits set to
zero. Bitmap data convert to either 0.0 or 1.0.
Page 4
glDrawPixels(3G) OpenGL Reference glDrawPixels(3G)
Each fixed-point index is then shifted left by GLINDEXSHIFT bits,
and added to GLINDEXOFFSET. If GLINDEXSHIFT is negative, the
shift is to the right. In either case, zero bits fill otherwise
unspecified bit locations in the result. If GLMAPSTENCIL is true,
the index is replaced with the value that it references in lookup
table GLPIXELMAPSTOS. Whether the lookup replacement of the
index is done or not, the integer part of the index is then ANDed
b
with 2 -1, where b is the number of bits in the stencil buffer.
The resulting stencil value is assigned to each of the fragments
generated for the source pixel.
GLDEPTHCOMPONENT
Each pixel is a single-depth component. Floating-point data is
converted directly to an internal floating-point format with
unspecified precision. Signed integer data is mapped linearly to
the internal floating-point format such that the most positive
representable integer value maps to 1.0, and the most negative
representable value maps to -1.0. Unsigned integer data is mapped
similarly: the largest integer value maps to 1.0, and zero maps to
0.0. The resulting floating-point depth value is then multiplied by
GLDEPTHSCALE and added to GLDEPTHBIAS. The result is clamped to
the range [0,1].
The resulting depth component, and the current raster position color
or color index and texture coordinates are assigned to each of the
fragments generated for the source pixel.
GLRGBA
GLABGREXT
Each pixel is a four-component group; for GLRGBA, the red component
is first, followed by green, followed by blue, followed by alpha;
for GLABGREXT the order is alpha, blue, green, and then red.
Floating-point values are converted directly to an internal
floating-point format with unspecified precision. Signed integer
values are mapped linearly to the internal floating-point format
such that the most positive representable integer value maps to 1.0,
and the most negative representable value maps to -1.0. Unsigned
integer data is mapped similarly: the largest integer value maps to
1.0, and zero maps to 0.0. The resulting floating-point color
values are then multiplied by GLcSCALE and added to GLcBIAS,
where c is RED, GREEN, BLUE, or ALPHA for the respective color
components. The results are clamped to the range [0,1].
If GLMAPCOLOR is true, each color component is scaled by the size
of lookup table GLPIXELMAPcTOc, then replaced by the value that
it references in that table. c is R, G, B, or A, respectively.
The resulting RGBA color components, and the current raster position
z and texture coordinates are assigned to each of the fragments
generated for the source pixel.
Page 5
glDrawPixels(3G) OpenGL Reference glDrawPixels(3G)
GLRGBA may be faster or slower than GLABGREXT on a given platform
depending on the internal representation of the colors.
GLRED
Each pixel is a single red component. This component is converted
to the internal floating-point format in the same way as the red
component of an RGBA pixel is, then it is converted to an RGBA pixel
with green and blue set to 0.0, and alpha set to 1.0. After this
conversion, the pixel is treated just as if it had been read as an
RGBA pixel.
GLGREEN
Each pixel is a single green component. This component is converted
to the internal floating-point format in the same way as the green
component of an RGBA pixel is, then it is converted to an RGBA pixel
with red and blue set to 0.0, and alpha set to 1.0. After this
conversion, the pixel is treated just as if it had been read as an
RGBA pixel.
GLBLUE
Each pixel is a single blue component. This component is converted
to the internal floating-point format in the same way as the blue
component of an RGBA pixel is, then it is converted to an RGBA pixel
with red and green set to 0.0, and alpha set to 1.0. After this
conversion, the pixel is treated just as if it had been read as an
RGBA pixel.
GLALPHA
Each pixel is a single alpha component. This component is converted
to the internal floating-point format in the same way as the alpha
component of an RGBA pixel is, then it is converted to an RGBA pixel
with red, green, and blue set to 0.0. After this conversion, the
pixel is treated just as if it had been read as an RGBA pixel.
GLRGB
Each pixel is a three-component group: red first, followed by
green, followed by blue. Each component is converted to the
internal floating-point format in the same way as the red, green,
and blue components of an RGBA pixel are. The color triple is
converted to an RGBA pixel with alpha set to 1.0. After this
conversion, the pixel is treated just as if it had been read as an
RGBA pixel.
GLLUMINANCE
Each pixel is a single luminance component. This component is
converted to the internal floating-point format in the same way as
the red component of an RGBA pixel is, then it is converted to an
RGBA pixel with red, green, and blue set to the converted luminance
value, and alpha set to 1.0. After this conversion, the pixel is
treated just as if it had been read as an RGBA pixel.
Page 6
glDrawPixels(3G) OpenGL Reference glDrawPixels(3G)
GLLUMINANCEALPHA
Each pixel is a two-component group: luminance first, followed by
alpha. The two components are converted to the internal floating-
point format in the same way as the red component of an RGBA pixel
is, then they are converted to an RGBA pixel with red, green, and
blue set to the converted luminance value, and alpha set to the
converted alpha value. After this conversion, the pixel is treated
just as if it had been read as an RGBA pixel.
GLYCRCB422SGIX
Each pixel is a two-component group: chroma first, followed by
luminance. The two components are converted to the internal
floating-point format of the red, green, and blue components of an
RGBA pixel using an unspecified conversion algorithm, and the alpha
component of the pixel is set to 1.0. The chroma component of each
even-count group in a row represents the Cb chroma value for a pixel
pair. The chroma component of each odd-count group in a row
represents the Cr chroma value for a pixel pair. Rows of odd length
will produce undefined red, green, and blue values for the final
pixel in a row. After this conversion, the pixel is treated just as
if it had been read as an RGBA pixel.
NOTES
Enabling GLINTERLACESGIX is equivalent to having a source rectangle
with image height equal to 2 * height where every other row contains
"transparent" pixels that do not affect the corresponding destination
pixels in the frame buffer. For example:
glEnable(GLINTERLACESGIX);
set current raster position to (x y )
r, r
glDrawPixels(width, height, GLRGBA, GLUNSIGNEDBYTE, I );
0
set raster position to (x y +zoom )
r, r y
glDrawPixels(width, height, GLRGBA, GLUNSIGNEDBYTE, I );
1
is equivalent to
glDisable( GLINTERLACESGIX);
set current raster position to (x y )
r, r
glDrawPixels(width, 2 * height, GLRGBA, GLUNSIGNEDBYTE, I );
2
where pixel rows (0,2,4,...) of I are from image I , and rows
2 0
(1,3,5,...) are from image I .
1
If type is set to GLUNSIGNEDBYTE332EXT,
GLUNSIGNEDSHORT4444EXT, GLUNSIGNEDSHORT5551EXT,
GLUNSIGNEDINT8888EXT, or GLUNSIGNEDINT1010102EXT and the
EXTpackedpixels extension is not supported then a GLINVALIDENUM error
is generated.
Page 7
glDrawPixels(3G) OpenGL Reference glDrawPixels(3G)
ERRORS
GLINVALIDVALUE is generated if either width or height is negative.
GLINVALIDENUM is generated if format or type is not one of the accepted
values.
GLINVALIDOPERATION is generated if type is set to
GLUNSIGNEDBYTE332EXT, GLUNSIGNEDSHORT4444EXT,
GLUNSIGNEDSHORT5551EXT, GLUNSIGNEDINT8888EXT, or
GLUNSIGNEDINT1010102EXT and the number of elements indicated by
type does not match the number indicated by format.
GLINVALIDOPERATION is generated if format is GLRED, GLGREEN, GLBLUE,
GLALPHA, GLRGB, GLRGBA, GLABGREXT, GLLUMINANCE, GLLUMINANCEALPHA,
or GLYCRCB422SGIX, and the GL is in color index mode.
GLINVALIDENUM is generated if type is GLBITMAP and format is not
either GLCOLORINDEX or GLSTENCILINDEX.
GLINVALIDOPERATION is generated if format is GLSTENCILINDEX and there
is no stencil buffer.
GLINVALIDOPERATION is generated if format is GLDEPTHCOMPONENT and
there is no depth buffer.
GLINVALIDOPERATION is generated if glDrawPixels is executed between the
execution of glBegin and the corresponding execution of glEnd.
ASSOCIATED GETS
glGet with argument GLCURRENTRASTERPOSITION
glGet with argument GLCURRENTRASTERPOSITIONVALID
glGet with argument GLINTERLACESGIX
MACHINE DEPENDENCIES
On RealityEngine, RealityEngine2, and VTX systems convolution may not be
used in the following circumstances:
1. When rendering to pixmaps.
2. When fragment processing (texturing, depth buffering, alpha
testing, multisampling, fog) is enabled.
3. When histogramming or minmax is enabled.
4. When either of the pixel zoom factors has a value other than 1.0
or -1.0.
In these cases, glDrawPixels and glCopyPixels report a
GLINVALIDOPERATION error and do not transfer any pixels.
Page 8
glDrawPixels(3G) OpenGL Reference glDrawPixels(3G)
Performance note for RealityEngine, RealityEngine2, and VTX systems:
Unsigned color types use the fastest pixel-drawing path. Smaller types
(e.g., GLUNSIGNEDBYTE) require less host-to-graphics bandwidth, and are
therefore faster than larger types (e.g., GLUNSIGNEDINT). Signed and
float types use the significantly slower floating-point pixel-drawing
path. The slower pixel-drawing path is also used when the format is
GLDEPTHCOMPONENT and when fragment operations (i.e., depth or alpha
testing, texturing, fog, etc.) are enabled.
For best performance on XS, XZ, Elan, and Extreme systems set type to
GLUNSIGNEDBYTE and, when drawing to the color buffer, set format to
GLABGREXT.
On InfiniteReality systems, signed color-index pixels written to
drawables with dual-personality (luminance + color-index) visuals will be
sign-extended into the high-order bits of the framebuffer. For example,
writing a signed byte value of 0x88 would yield 0xF88 in a 12-bit
drawable.
The SGIXycrcb extension is supported only on O2 systems.
The SGIXinterlace extension is supported only on InfiniteReality
systems, on RealityEngine, RealityEngine2, and VTX systems, and on O2
systems.
The EXTpackedpixels extension is not supported on RealityEngine,
RealityEngine2, and VTX systems.
SEE ALSO
glAlphaFunc, glBlendFunc, glColorTableSGI, glCopyPixels,
glConvolutionFilter2DEXT, glDepthFunc, glEnable, glHistogramEXT,
glLogicOp, glMinmaxEXT, glPixelMap, glPixelStore, glPixelTransfer,
glPixelZoom, glRasterPos, glReadPixels, glScissor,
glSeparableFilter2DEXT, glStencilFunc.
Page 9