>Psychtoolbox>PsychDemos>PsychTutorials

AdditiveBlendingForLinearSuperpositionTutorial([outputdevice=’None’][, overlay=1][, colorclut=0][, doGainCorrection=0][, useVulkan=0]);

Illustrates use of floating point textures in combination with
source-weighted additive alpha blending to create linear superpositions
of image patches, in this case of two superimposed gratings. Due to the
use of additive alpha-blending, you’ll see that the two gratings will
superimpose onto each other in a mathematically correct way – pixel-wise
addition/subtraction of luminance values. The demo uses a 32 bit floating
point framebuffer on the latest hardware. This allows for an effective 23
bits of precision in all math done and in the final stimuli - more than
any display device in existence could resolve. On previous generation
hardware (older than NVidia Geforce 8800 or ATI Radeon HD2000), alpha
blending isn’t supported in 32 bpc float precision. Therefore the demo
will select 16 bpc floating point precision, where alpha blending works.
This way the effective precision is 11 bits, a bit less than what special
display devices can resolve. However, final gamma correction is done in
full 23 bits precision, so you can make use of the extra bits for proper
display linearization.

The demo also demonstrates Psychtoolbox support for high precision
display output devices, ie., display extenders, graphics cards and
display modes that allow for luminance display with more than the
standard 8 bits precision. By default the standard 8 bit framebuffer of
standard graphics cards is demonstrated.

However, by providing the first optional parameter ‘outputdevice’, you
can select amongst all devices supported by PTB:

‘NoneNoGamma’ - Same as ‘None’: Use standard 8 bit framebuffer, but
disable the gamma correction provided by PTB’s imaging pipeline. This is
usually not what you want, but it allows to test how much faster the
display runs without gamma correction.

‘PseudoGray’ - PseudoGray display, also known as “Bit stealing”. This
technique allows to create the perception of up to 1786 different
luminance levels on standard 8 bit graphics hardware by use of some
clever color rendering trick. See “help CreatePseudoGrayLUT” for
references and details.

‘Native10Bit’ - Enables the native 10 bpc framebuffer support on all supported
GPUs. All gpus from AMD since ~2006, Intel since around ~2010 and NVidia since
around ~2008 support this on Linux. This also works similar modern with NVidia
Quadro and AMD Fire professional gpu’s ounder MS-Windows.

‘Native11Bit’ - Enables the native ~11 bpc framebuffer support on some ATI
Radeon X1xxx / HDxxx GPU’s with DCE-8 to DCE-12 display engine when used under
Linux. These GPU’s do support ~11 bits per color channel when this special mode
is used (11 bits red, 11 bits green, 10 bits blue).

‘Native16Bit’ - Enables the native up to 16 bpc framebuffer support on AMD
GPU’s when used under Linux with Vulkan display backend. While this activates
a 16 bpc framebuffer, the precision of the video output signal depends on the
specific gpu, connection and display. As of 2023, the “Sea Islands” AMD gpu
family and later can output at most 12 bpc precision to suitable displays over
HDMI or DisplayPort.

‘Native16BitFloat’ - Enable native 16 bit floating point (~11 bit linear)
framebuffer support on suitable operating systems and graphics cards.

‘VideoSwitcher’ - Enable the Xiangrui Li et al. VideoSwitcher, a special
type of video attenuator (see “help PsychVideoSwitcher”) in standard
“simple” mode.

‘VideoSwitcherCalibrated’ - Enable the Xiangrui Li et al. VideoSwitcher,
but use the more complex (and more accurate?) mode with calibrated lookup
tables (see “help PsychVideoSwitcher”).

‘Attenuator’ - Enable support for standard Pelli & Zhang style video
attenuators by use of lookup tables.

Then we have support for the different modes of operation of the
Cambridge Research Systems Bits++ box:

‘Mono++’ - Use 14 bit mono output mode, either with color index overlay
(if the optional 2nd ‘overlay’ flag is set to 1, which is the default),
or without color index overlay.

‘Color++’ - User 14 bits per color component mode.

Then we have support for the different modes of operation of the
VPixx Technologies DPixx (DataPixx) box:

‘M16’ - Use 16 bit mono output mode, either with color index overlay
(if the optional 2nd ‘overlay’ flag is set to 1, which is the default),
or without color index overlay.

‘C48’ - User 16 bits per color component mode.

‘DualPipeHDR’ - Use experimental output to dual-pipeline HDR display
device.

The third optional parameter ‘colorclut’ if provided, will use color
lookup table based color correction / mapping, ie., mapping of luminance
values to RGB values in the framebuffer - or intensity values if
applicable. Any non-zero number will select a different CLUT. Look into
the code for description. This just to demonstrate that you can use CLUT
based color/intensity correction instead of simple power-law gamma
correction if you want.

The fourth optional parameter ‘doGainCorrection’ if provided and set to
1, will demonstrate use of display per-pixel gain correction, aka
vignetting correction. It will modulate the brightness of each pixel with
a gain factor, the gains increasing linearly from the left border to the
right border of the display. See “help VignettingCorrectionDemo” for more
details of this feature.

The fifth optional parameter ‘useVulkan’ if provided and set to 1, will try
to use a Vulkan based display backend, instead of the standard OpenGL based
display backend. See “help PsychVulkan” for system requirements and caveats.
You need to set useVulkan to 1 on Linux to get Native16Bit or Native16BitFloat
mode on AMD graphics under Linux.

The demo shows two superimposed sine wave gratings in the center of the
screen. You can shift the 2nd grating up and down in subpixel steps by
use of the cursor up-/down keys. You can change the contrast of the 2nd
grating by use of the cursor left-/right keys. You can move the 2nd
grating with the mouse, and rotate it clockwise or counterclockwise by
pressing the mouse buttons. The keys ‘i’ and ‘d’ allow to change the
“encoding gamma” factor used for the gamma correction algorithm. The ESC
ape key ends the demo. Actually the demo performs some benchmark run for
a few seconds after you’ve pressed ESC key, just to measure the speed of
your graphics card in stimulus conversion.

In ‘VideoSwitcher’ mode, it also draws some vertically moving greenish
sync line just to show how to generate trigger signals on the
VideoSwitcher device.

Needs hardware with support for imaging pipeline (GLSL shaders and
floating point framebuffers). Should work well on ATI Radeon X1000 and
later, Geforce 6000 and later and even better on DirectX10 hardware like
Radeon HD series and NVidia Geforce 8 / 9 series and later.