Perception Vision | Auric Artisan

Source

The card carries a grey ramp, because a grey ramp is what catches a simulation tinting neutrals. The Snellen chart is here so the acuity mode has something to be judged on.

Simulation

Eight, where the badge said nine — the ninth was “none”, which is a real entry in the list and not a simulation. The CSF overlay is gone: it multiplied the image by a spatial-frequency curve fed the distance from the centre, and returned a black frame at its own defaults. See Modes.

Type

Daltonise instead of simulate

Severity 1.00

At 1.00 this is dichromacy — the cone class is absent. Between 0 and 1 it is a linear blend toward it, which is a convention this page uses and not a model of anomalous trichromacy.

Compare

Divider 50%

Split is the default because the question is always “against what”. Both halves get the same adjustments, so the only thing that differs across the divider is the simulation. Drag on the image to move it.

Renderer

Fixation cross

The same maths either way: the colour transform composes into one matrix, which the CPU builds and the shader is handed. There was a complete CPU implementation of every mode in the file, used only by the export table — and no fallback at all when WebGL2 was missing.

Before the eye

Adaptation 1.00

Bradford, applied to the stimulus before the visual system, because that is the order light arrives in. Under illuminant A the scene leaves sRGB by a wide margin — that is the correct answer, and the readout below says how much.

Adjustments

Contrast 0.00 Gamma 1.00 Blur 0.0 Glare 0.00 Noise 0.000

These are not simulations and they are not counted as ones. Blur and glare had uniforms no shader read, so both sliders moved nothing at all.

Presets

Every value a preset sets reaches something that draws. The old low-light preset set glare to 0.30 against a uniform no shader read, which made it identical to plain scotopic.

The frame, and the same frame seen otherwise

GPU · —

Original Simulated

The grey ramp, which is the test that needs no reference

Neutral drift —

Colours merged —

Out of sRGB —

GPU and CPU agree —

Distinct colours left —

Largest channel shift —

Frame luminance, mean —

min and max — – —

“Colours merged” is the one this page exists to show: a dichromacy projection collapses a hue axis, so colours that were distinct arrive at the same place, and the figure counts how many of the frame’s own colours did. The other two are worth having. A dichromacy projection lands outside sRGB for saturated colours — the figure says how much of the frame, rather than letting the clamp hide it. And the two renderers are checked against each other on a grid of 216 colours after every change, because they used to be two different implementations of the same feature and nothing on the page said which one you were looking at.

Where a colour goes, and what each step takes

The same colours, three ways

The tell: a neutral must stay neutral

Every entry, and what it does to the card

Controls that were read and never used

Every dataset and transform this tool rests on

The colours, through the simulation on screen

0 rows

Input	Simulated	L	M	S	ΔE₀₀	In sRGB

The settings file, and what it carries

Reference